{"count":298,"items":[{"slug":"pip","term":"Pip","category":"basics","definition":"A pip (Percentage in Point) is the smallest price move in a forex quote. For most currency pairs, a pip equals 0.0001 (the fourth decimal place). For JPY pairs, a pip is 0.01 (the second decimal place). Pips are the standard unit for measuring price changes and calculating profit or loss in forex trading. For example, if EUR/USD moves from 1.1050 to 1.1055, that's a 5-pip move.","relatedTerms":["pipette","spread","lot-size"],"relatedGuide":null},{"slug":"pipette","term":"Pipette","category":"basics","definition":"A pipette is one-tenth of a pip, representing the fifth decimal place in most currency pairs (0.00001) or the third decimal place for JPY pairs (0.001). Many modern brokers quote prices in pipettes for more precise pricing. Also known as a fractional pip or point.","relatedTerms":["pip","spread"],"relatedGuide":null},{"slug":"lot-size","term":"Lot Size","category":"basics","definition":"A lot is a standardized unit of measurement for a forex transaction. A standard lot equals 100,000 units of the base currency. A mini lot is 10,000 units, a micro lot is 1,000 units, and a nano lot is 100 units. Lot size directly affects the pip value and therefore the potential profit or loss of a trade.","relatedTerms":["pip","margin","leverage","lot"],"relatedGuide":null},{"slug":"spread","term":"Spread","category":"basics","definition":"The spread is the difference between the bid price (sell) and the ask price (buy) of a currency pair. It represents the broker's fee and is measured in pips. Tighter spreads mean lower trading costs. Spreads can be fixed or variable depending on the broker and market conditions. Major pairs like EUR/USD typically have the tightest spreads.","relatedTerms":["pip","bid-price","ask-price","slippage"],"relatedGuide":null},{"slug":"bid-price","term":"Bid Price","category":"basics","definition":"The bid price is the highest price a buyer is willing to pay for a currency pair. It is the price at which you can sell the base currency. The bid is always lower than the ask price, and the difference between them is the spread.","relatedTerms":["ask-price","spread","bid"],"relatedGuide":null},{"slug":"ask-price","term":"Ask Price","category":"basics","definition":"The ask price (also called the offer price) is the lowest price at which a seller is willing to sell a currency pair. It is the price at which you buy the base currency. The ask is always higher than the bid price.","relatedTerms":["bid-price","spread","ask"],"relatedGuide":null},{"slug":"base-currency","term":"Base Currency","category":"basics","definition":"The base currency is the first currency listed in a forex pair. In EUR/USD, the euro (EUR) is the base currency. When you buy a currency pair, you are buying the base currency and selling the quote currency. The exchange rate shows how much of the quote currency is needed to buy one unit of the base currency.","relatedTerms":["quote-currency","currency-pair"],"relatedGuide":null},{"slug":"quote-currency","term":"Quote Currency","category":"basics","definition":"The quote currency (or counter currency) is the second currency in a forex pair. In EUR/USD, the US dollar (USD) is the quote currency. It represents the amount of that currency needed to purchase one unit of the base currency.","relatedTerms":["base-currency","currency-pair"],"relatedGuide":null},{"slug":"currency-pair","term":"Currency Pair","category":"basics","definition":"A currency pair is the quotation of two different currencies, with the value of one currency being quoted against the other. The first currency is the base currency and the second is the quote currency. Major pairs include EUR/USD, GBP/USD, USD/JPY. Cross pairs don't include USD, such as EUR/GBP or AUD/JPY.","relatedTerms":["base-currency","quote-currency","major-pairs","cross-pairs"],"relatedGuide":null},{"slug":"major-pairs","term":"Major Currency Pairs","category":"basics","definition":"Major currency pairs are the most traded forex pairs in the world, all containing the US dollar. The seven majors are: EUR/USD, USD/JPY, GBP/USD, USD/CHF, AUD/USD, USD/CAD, and NZD/USD. They offer the highest liquidity, tightest spreads, and most trading volume.","relatedTerms":["currency-pair","cross-pairs","exotic-pairs"],"relatedGuide":null},{"slug":"cross-pairs","term":"Cross Currency Pairs","category":"basics","definition":"Cross pairs (or crosses) are currency pairs that do not include the US dollar. Popular crosses include EUR/GBP, EUR/JPY, GBP/JPY, and AUD/NZD. They typically have wider spreads than major pairs but can offer good trading opportunities.","relatedTerms":["major-pairs","currency-pair"],"relatedGuide":null},{"slug":"exotic-pairs","term":"Exotic Currency Pairs","category":"basics","definition":"Exotic pairs consist of one major currency paired with a currency from a developing economy (e.g., USD/TRY, EUR/ZAR, GBP/SGD). They have wider spreads, lower liquidity, and higher volatility compared to majors and crosses.","relatedTerms":["major-pairs","cross-pairs"],"relatedGuide":null},{"slug":"leverage","term":"Leverage","category":"trading","definition":"Leverage allows traders to control a large position with a relatively small amount of capital. Expressed as a ratio (e.g., 1:100, 1:500), it means that for every $1 of margin, you can control $100 or $500 worth of currency. While leverage amplifies potential profits, it equally magnifies potential losses. Brokers set maximum leverage limits, and regulations vary by jurisdiction.","relatedTerms":["margin","margin-call","lot-size"],"relatedGuide":"risk-management-for-ea"},{"slug":"margin","term":"Margin","category":"trading","definition":"Margin is the amount of money required to open and maintain a leveraged trading position. It is not a fee but a portion of your account equity set aside as collateral. Free margin is the available equity that can be used to open new positions. Used margin is the amount locked in open positions.","relatedTerms":["leverage","margin-call","free-margin"],"relatedGuide":null},{"slug":"margin-call","term":"Margin Call","category":"trading","definition":"A margin call occurs when your account equity falls below the required margin level, typically at 100% or 50% depending on the broker. The broker will notify you to either deposit more funds or close positions to restore the margin level. If the equity continues to drop, the broker may automatically close positions (stop-out).","relatedTerms":["margin","stop-out","leverage"],"relatedGuide":"risk-management-for-ea"},{"slug":"stop-out","term":"Stop Out Level","category":"trading","definition":"The stop-out level is the margin level at which the broker automatically begins closing your open positions to prevent further losses. Typically set at 20-50% margin level. Positions are closed starting from the one with the largest loss.","relatedTerms":["margin-call","margin"],"relatedGuide":null},{"slug":"swap","term":"Swap (Rollover)","category":"trading","definition":"A swap or rollover is the interest rate differential between the two currencies in a pair, charged or credited to your account when you hold a position overnight. Swap can be positive (you earn) or negative (you pay). Swap-free accounts (Islamic accounts) are available for traders who cannot receive or pay interest.","relatedTerms":["carry-trade","swap-rollover"],"relatedGuide":null},{"slug":"slippage","term":"Slippage","category":"trading","definition":"Slippage occurs when a trade is executed at a different price than expected. It happens during periods of high volatility or low liquidity, especially during news events. Slippage can be positive (better price) or negative (worse price). Using limit orders instead of market orders can help reduce slippage.","relatedTerms":["spread","market-order","limit-order"],"relatedGuide":null},{"slug":"liquidity","term":"Liquidity","category":"trading","definition":"Liquidity refers to how easily a currency pair can be bought or sold without causing a significant price change. Major pairs like EUR/USD are highly liquid with tight spreads. Higher liquidity generally means faster execution and less slippage. Liquidity varies throughout the day based on market sessions.","relatedTerms":["spread","slippage","major-pairs"],"relatedGuide":null},{"slug":"volatility","term":"Volatility","category":"trading","definition":"Volatility measures the degree of price fluctuation over a given period. High volatility means larger price swings and potentially more profit opportunities but also higher risk. Volatility is often measured using indicators like ATR (Average True Range) or Bollinger Bands. News events, economic releases, and session overlaps increase volatility.","relatedTerms":["atr","bollinger-bands"],"relatedGuide":null},{"slug":"carry-trade","term":"Carry Trade","category":"trading","definition":"A carry trade involves borrowing a currency with a low interest rate and investing in a currency with a higher interest rate, profiting from the interest rate differential (swap). It is a long-term strategy that works best in stable market conditions.","relatedTerms":["swap","leverage"],"relatedGuide":null},{"slug":"long-position","term":"Long Position","category":"trading","definition":"Going long means buying a currency pair, expecting the base currency to appreciate against the quote currency. If EUR/USD is at 1.1000 and you go long, you profit when the price rises above 1.1000.","relatedTerms":["short-position","base-currency"],"relatedGuide":null},{"slug":"short-position","term":"Short Position","category":"trading","definition":"Going short means selling a currency pair, expecting the base currency to depreciate against the quote currency. If EUR/USD is at 1.1000 and you go short, you profit when the price falls below 1.1000.","relatedTerms":["long-position","base-currency"],"relatedGuide":null},{"slug":"technical-analysis","term":"Technical Analysis","category":"analysis","definition":"Technical analysis is a method of evaluating markets by analyzing price charts, patterns, and statistical indicators. It assumes that price movements follow trends and that history tends to repeat itself. Common tools include moving averages, RSI, MACD, and support/resistance levels.","relatedTerms":["fundamental-analysis","moving-average","rsi","macd"],"relatedGuide":"what-is-forex-indicator"},{"slug":"fundamental-analysis","term":"Fundamental Analysis","category":"analysis","definition":"Fundamental analysis evaluates a currency's value by examining economic, financial, and geopolitical factors. Key data includes GDP, interest rates, inflation, employment figures, and central bank policies. Fundamental traders focus on economic news releases and macroeconomic trends.","relatedTerms":["technical-analysis"],"relatedGuide":null},{"slug":"support-resistance","term":"Support and Resistance","category":"analysis","definition":"Support is a price level where buying pressure is strong enough to prevent further decline. Resistance is where selling pressure prevents further rise. These levels are identified using historical price data, trend lines, moving averages, and Fibonacci retracements. When price breaks through support or resistance, it often signals a trend continuation.","relatedTerms":["trend-line","fibonacci-retracement","breakout","support","resistance"],"relatedGuide":null},{"slug":"trend-line","term":"Trend Line","category":"analysis","definition":"A trend line is a straight line drawn on a chart connecting two or more price points, used to identify the direction and strength of a trend. An uptrend line connects higher lows; a downtrend line connects lower highs. The more times price touches a trend line without breaking it, the stronger it is.","relatedTerms":["support-resistance","trend"],"relatedGuide":null},{"slug":"trend","term":"Trend","category":"analysis","definition":"A trend is the general direction of a market or asset's price over time. An uptrend (bullish) has higher highs and higher lows. A downtrend (bearish) has lower highs and lower lows. A sideways trend (range) shows no clear direction. 'The trend is your friend' is a fundamental trading principle.","relatedTerms":["trend-line","trend-following"],"relatedGuide":null},{"slug":"moving-average","term":"Moving Average (MA)","category":"analysis","definition":"A moving average smooths out price data by calculating the average price over a specified period. The Simple Moving Average (SMA) gives equal weight to all prices; the Exponential Moving Average (EMA) gives more weight to recent prices. Common periods are 20, 50, 100, and 200. MAs are used to identify trends, support/resistance, and crossover signals.","relatedTerms":["ema","sma","golden-cross","death-cross"],"relatedGuide":null},{"slug":"ema","term":"Exponential Moving Average (EMA)","category":"analysis","definition":"The EMA is a type of moving average that places more weight on recent prices, making it more responsive to new information than the SMA. Commonly used periods include EMA 9, 21, 50, and 200. EMAs are popular for short-term trading signals and trend identification.","relatedTerms":["moving-average","sma"],"relatedGuide":null},{"slug":"sma","term":"Simple Moving Average (SMA)","category":"analysis","definition":"The SMA calculates the average price over a specific number of periods, giving equal weight to each price point. SMA 200 is widely watched as a long-term trend indicator. When price crosses above the SMA, it may signal an uptrend; below signals a downtrend.","relatedTerms":["moving-average","ema"],"relatedGuide":null},{"slug":"rsi","term":"Relative Strength Index (RSI)","category":"analysis","definition":"RSI is a momentum oscillator that measures the speed and magnitude of price changes on a scale of 0 to 100. RSI above 70 suggests overbought conditions (potential reversal down); below 30 suggests oversold (potential reversal up). The standard period is 14. RSI divergence with price can signal trend weakness.","relatedTerms":["oscillator","overbought","oversold","divergence"],"relatedGuide":null},{"slug":"macd","term":"MACD (Moving Average Convergence Divergence)","category":"analysis","definition":"MACD is a trend-following momentum indicator showing the relationship between two EMAs (typically 12 and 26 periods). The MACD line minus the signal line (9 EMA of MACD) generates buy/sell signals. When MACD crosses above the signal line, it's bullish; below is bearish. The histogram visualizes the difference between the two lines.","relatedTerms":["ema","divergence"],"relatedGuide":null},{"slug":"bollinger-bands","term":"Bollinger Bands","category":"analysis","definition":"Bollinger Bands consist of a middle band (SMA 20) and two outer bands placed 2 standard deviations above and below. They measure volatility -- bands widen during high volatility and narrow during low volatility (squeeze). Price touching the upper band may indicate overbought conditions; touching the lower band may indicate oversold.","relatedTerms":["sma","volatility","overbought","oversold"],"relatedGuide":null},{"slug":"fibonacci-retracement","term":"Fibonacci Retracement","category":"analysis","definition":"Fibonacci retracement uses horizontal lines at key Fibonacci levels (23.6%, 38.2%, 50%, 61.8%, 78.6%) to identify potential support and resistance areas. These levels are drawn between a significant high and low point on a chart. The 61.8% level (golden ratio) is considered the most significant.","relatedTerms":["support-resistance","golden-ratio"],"relatedGuide":null},{"slug":"atr","term":"Average True Range (ATR)","category":"analysis","definition":"ATR is a volatility indicator that shows the average range of price movement over a specified period (typically 14). It doesn't indicate direction, only the degree of price volatility. ATR is commonly used to set stop-loss levels, position sizing, and identifying volatility breakouts.","relatedTerms":["volatility","stop-loss"],"relatedGuide":null},{"slug":"oscillator","term":"Oscillator","category":"analysis","definition":"An oscillator is a technical indicator that moves between fixed boundaries (e.g., 0-100), helping identify overbought and oversold conditions. Common oscillators include RSI, Stochastic, and CCI. They are most effective in ranging markets.","relatedTerms":["rsi","stochastic","overbought","oversold"],"relatedGuide":null},{"slug":"stochastic","term":"Stochastic Oscillator","category":"analysis","definition":"The Stochastic oscillator compares a closing price to its price range over a given period. It produces two lines (%K and %D) ranging from 0-100. Readings above 80 indicate overbought, below 20 indicate oversold. Crossovers between %K and %D generate trading signals.","relatedTerms":["oscillator","rsi"],"relatedGuide":null},{"slug":"candlestick","term":"Candlestick Chart","category":"analysis","definition":"A candlestick chart displays price data using candle-shaped bars showing open, high, low, and close prices for a time period. A bullish (green/white) candle closes higher than it opens; a bearish (red/black) candle closes lower. Candlestick patterns like Doji, Hammer, and Engulfing are used to predict reversals.","relatedTerms":["doji","trend"],"relatedGuide":null},{"slug":"doji","term":"Doji","category":"analysis","definition":"A Doji is a candlestick pattern where the opening and closing prices are virtually equal, creating a cross or plus sign shape. It signals market indecision and potential reversal. Types include Standard Doji, Dragonfly Doji, Gravestone Doji, and Long-Legged Doji.","relatedTerms":["candlestick"],"relatedGuide":null},{"slug":"divergence","term":"Divergence","category":"analysis","definition":"Divergence occurs when price moves in the opposite direction of a technical indicator (e.g., RSI, MACD). Bullish divergence: price makes lower lows while the indicator makes higher lows (potential upward reversal). Bearish divergence: price makes higher highs while the indicator makes lower highs (potential downward reversal).","relatedTerms":["rsi","macd"],"relatedGuide":null},{"slug":"overbought","term":"Overbought","category":"analysis","definition":"A market condition where a security has risen significantly and rapidly, potentially beyond its fair value. Indicators like RSI above 70 or Stochastic above 80 suggest overbought conditions. It may signal an upcoming pullback or reversal, though strong trends can remain overbought for extended periods.","relatedTerms":["oversold","rsi"],"relatedGuide":null},{"slug":"oversold","term":"Oversold","category":"analysis","definition":"A market condition where a security has fallen significantly and rapidly, potentially below its fair value. Indicators like RSI below 30 or Stochastic below 20 suggest oversold conditions. It may signal an upcoming bounce or reversal.","relatedTerms":["overbought","rsi"],"relatedGuide":null},{"slug":"breakout","term":"Breakout","category":"analysis","definition":"A breakout occurs when price moves above resistance or below support with increased volume. Breakout trading involves entering a position when price breaks a key level, expecting the momentum to continue. False breakouts (fakeouts) occur when price briefly breaks a level but quickly reverses.","relatedTerms":["support-resistance","volume"],"relatedGuide":null},{"slug":"golden-cross","term":"Golden Cross","category":"analysis","definition":"A golden cross is a bullish signal that occurs when a short-term moving average (typically SMA 50) crosses above a long-term moving average (typically SMA 200). It suggests the beginning of a major uptrend and is widely followed by traders.","relatedTerms":["death-cross","moving-average"],"relatedGuide":null},{"slug":"death-cross","term":"Death Cross","category":"analysis","definition":"A death cross is a bearish signal that occurs when a short-term moving average (typically SMA 50) crosses below a long-term moving average (typically SMA 200). It suggests the beginning of a major downtrend.","relatedTerms":["golden-cross","moving-average"],"relatedGuide":null},{"slug":"volume","term":"Volume","category":"analysis","definition":"Volume represents the total number of shares or contracts traded during a given period. In forex, tick volume (number of price changes) is used as a proxy since there's no centralized exchange. High volume during a price move confirms the strength of the move.","relatedTerms":["breakout","liquidity"],"relatedGuide":null},{"slug":"expert-advisor","term":"Expert Advisor (EA)","category":"platform","definition":"An Expert Advisor is an automated trading program written in MQL4 or MQL5 that runs on the MetaTrader platform. EAs can analyze markets, place trades, manage positions, and execute complete trading strategies without manual intervention. They follow predefined rules and algorithms, eliminating emotional trading decisions.","relatedTerms":["metatrader-4","metatrader-5","mql","expert-advisor-ea"],"relatedGuide":"what-is-expert-advisor"},{"slug":"metatrader-4","term":"MetaTrader 4 (MT4)","category":"platform","definition":"MetaTrader 4 is the most widely used forex trading platform, developed by MetaQuotes Software. It supports Expert Advisors written in MQL4, custom indicators, scripts, and backtesting via the Strategy Tester. MT4 offers charting tools, multiple order types, and a built-in marketplace for trading tools.","relatedTerms":["metatrader-5","expert-advisor","mql"],"relatedGuide":"how-to-install-ea-mt4"},{"slug":"metatrader-5","term":"MetaTrader 5 (MT5)","category":"platform","definition":"MetaTrader 5 is the successor to MT4, supporting more asset classes (stocks, futures, options alongside forex), more timeframes, a built-in economic calendar, DOM (Depth of Market), and improved backtesting with multi-currency strategy testing. It uses MQL5 programming language.","relatedTerms":["metatrader-4","expert-advisor","mql","mt5-metatrader-5"],"relatedGuide":"how-to-install-ea-mt5"},{"slug":"mql","term":"MQL (MetaQuotes Language)","category":"platform","definition":"MQL is the programming language used to create Expert Advisors, custom indicators, scripts, and libraries for MetaTrader platforms. MQL4 is used for MT4 and MQL5 for MT5. It is a C++-like language specifically designed for developing trading algorithms and technical analysis tools.","relatedTerms":["expert-advisor","metatrader-4","metatrader-5"],"relatedGuide":null},{"slug":"backtesting","term":"Backtesting","category":"platform","definition":"Backtesting is the process of testing a trading strategy using historical market data to evaluate its performance. In MetaTrader, the Strategy Tester simulates trades based on past data. Key metrics include profit factor, max drawdown, win rate, and total profit. Backtesting helps validate strategies before risking real money, though past results don't guarantee future performance.","relatedTerms":["forward-testing","strategy-tester","profit-factor","backtest"],"relatedGuide":"forex-robot-backtesting"},{"slug":"strategy-tester","term":"Strategy Tester","category":"platform","definition":"The Strategy Tester is MetaTrader's built-in backtesting tool. It simulates an EA's performance using historical data with different modeling modes: every tick (most accurate), 1-minute OHLC, and open prices only (fastest). MT5's tester supports multi-currency and multi-timeframe testing.","relatedTerms":["backtesting","expert-advisor"],"relatedGuide":"forex-robot-backtesting"},{"slug":"forward-testing","term":"Forward Testing","category":"platform","definition":"Forward testing (or paper trading) runs a strategy on a demo account with live market data in real-time to validate backtesting results. It accounts for real-world factors like slippage, spread variations, and execution delays that backtesting may miss.","relatedTerms":["backtesting","demo-account","forward-test"],"relatedGuide":null},{"slug":"demo-account","term":"Demo Account","category":"platform","definition":"A demo account is a simulated trading account funded with virtual money, allowing traders to practice without financial risk. It uses real-time market data and replicates live trading conditions. Ideal for testing strategies, learning platforms, and evaluating Expert Advisors before live trading.","relatedTerms":["forward-testing","live-account","demo-account-parity"],"relatedGuide":null},{"slug":"vps","term":"VPS (Virtual Private Server)","category":"platform","definition":"A VPS is a remote server that runs your MetaTrader platform 24/7 without interruption. Essential for Expert Advisors that need continuous operation. Benefits include low latency to broker servers, uninterrupted power and internet, and no need to keep your computer running. Popular VPS providers offer MetaTrader-optimized plans.","relatedTerms":["expert-advisor","vps-virtual-private-server","persistent-vps"],"relatedGuide":"vps-for-forex-robots"},{"slug":"indicator","term":"Forex Indicator","category":"platform","definition":"A forex indicator is a mathematical calculation based on price, volume, or open interest data that generates visual signals on charts. Indicators help traders identify trends, momentum, volatility, and potential entry/exit points. They can be leading (predictive) or lagging (confirming). Custom indicators can be created in MQL for MetaTrader.","relatedTerms":["expert-advisor","technical-analysis"],"relatedGuide":"what-is-forex-indicator"},{"slug":"script","term":"Script (MetaTrader)","category":"platform","definition":"A script is a program that runs once to perform a specific task in MetaTrader, unlike an EA which runs continuously. Scripts can close all open orders, place pending orders, calculate lot sizes, or perform other one-time operations. They are written in MQL4/MQL5.","relatedTerms":["expert-advisor","mql"],"relatedGuide":null},{"slug":"timeframe","term":"Timeframe","category":"platform","definition":"A timeframe defines the period each candlestick or bar represents on a chart. Common timeframes include M1 (1 minute), M5, M15, M30, H1 (1 hour), H4, D1 (daily), W1 (weekly), and MN (monthly). Scalpers use lower timeframes (M1-M15); swing traders use H4-D1; position traders use D1-MN.","relatedTerms":["candlestick","scalping","swing-trading"],"relatedGuide":"best-timeframe-for-scalping"},{"slug":"drawdown","term":"Drawdown","category":"risk","definition":"Drawdown is the peak-to-trough decline in account equity, expressed as a percentage. Maximum drawdown is the largest percentage drop from a peak before a new peak is reached. It is a critical risk metric for evaluating trading strategies and Expert Advisors. A 50% drawdown requires a 100% gain to recover, making low-drawdown strategies preferable.","relatedTerms":["profit-factor","risk-reward-ratio","maximum-drawdown","peak-drawdown"],"relatedGuide":"what-is-drawdown"},{"slug":"profit-factor","term":"Profit Factor","category":"risk","definition":"Profit factor is the ratio of gross profit to gross loss. A profit factor above 1.0 means the strategy is profitable. Values above 1.5 are considered good; above 2.0 is excellent. It is one of the most important metrics for evaluating trading system performance. Profit Factor = Total Winning Trades / Total Losing Trades.","relatedTerms":["drawdown","win-rate"],"relatedGuide":"what-is-profit-factor"},{"slug":"sharpe-ratio","term":"Sharpe Ratio","category":"risk","definition":"The Sharpe ratio measures risk-adjusted return by dividing the excess return of an investment over the risk-free rate by its standard deviation. A higher Sharpe ratio indicates better risk-adjusted performance. Values above 1.0 are acceptable; above 2.0 is very good; above 3.0 is excellent.","relatedTerms":["profit-factor","drawdown"],"relatedGuide":null},{"slug":"risk-reward-ratio","term":"Risk-Reward Ratio","category":"risk","definition":"The risk-reward ratio compares the potential loss (risk) to the potential profit (reward) of a trade. A 1:2 ratio means risking $1 to potentially gain $2. Professional traders typically aim for at least 1:1.5 or 1:2. Combined with a win rate above 50%, favorable risk-reward ratios lead to long-term profitability.","relatedTerms":["stop-loss","take-profit"],"relatedGuide":"risk-management-for-ea"},{"slug":"win-rate","term":"Win Rate","category":"risk","definition":"Win rate is the percentage of trades that are profitable. A 60% win rate means 6 out of 10 trades are winners. However, win rate alone doesn't determine profitability -- it must be considered alongside the risk-reward ratio. A strategy with 40% win rate but 1:3 risk-reward can be more profitable than 70% win rate with 1:0.5.","relatedTerms":["profit-factor","risk-reward-ratio"],"relatedGuide":null},{"slug":"position-sizing","term":"Position Sizing","category":"risk","definition":"Position sizing determines how much capital to allocate to a single trade, usually based on risk percentage per trade (e.g., risking 1-2% of account balance). Proper position sizing protects against large losses and is fundamental to money management. The Kelly Criterion and fixed fractional methods are common sizing approaches.","relatedTerms":["lot-size","risk-reward-ratio","drawdown"],"relatedGuide":"risk-management-for-ea"},{"slug":"equity","term":"Equity","category":"risk","definition":"Equity is the current value of your trading account, calculated as Balance + Unrealized Profit/Loss. It fluctuates with open positions. Equity determines your available margin and margin level. When equity drops below the required margin, a margin call occurs.","relatedTerms":["margin","balance","drawdown"],"relatedGuide":null},{"slug":"balance","term":"Balance","category":"risk","definition":"Account balance is the total amount of money in your trading account after all closed trades but not accounting for open positions. It changes only when positions are closed. Equity = Balance + Floating P/L.","relatedTerms":["equity","margin"],"relatedGuide":null},{"slug":"free-margin","term":"Free Margin","category":"risk","definition":"Free margin is the amount of money in your account that is available to open new positions. Calculated as Equity minus Used Margin. If free margin drops to zero, no new positions can be opened.","relatedTerms":["margin","equity","leverage","drawdown"],"relatedGuide":null},{"slug":"market-order","term":"Market Order","category":"order","definition":"A market order is an instruction to buy or sell immediately at the current market price. It guarantees execution but not the exact price, as slippage may occur during volatile markets. Market orders are the simplest and most common order type.","relatedTerms":["limit-order","slippage"],"relatedGuide":null},{"slug":"limit-order","term":"Limit Order","category":"order","definition":"A limit order is set to buy below or sell above the current market price. A buy limit is placed below the current price (expecting a bounce); a sell limit is placed above (expecting a reversal). Limit orders guarantee the execution price but not execution itself.","relatedTerms":["market-order","stop-order"],"relatedGuide":null},{"slug":"stop-order","term":"Stop Order","category":"order","definition":"A stop order (or stop entry order) is placed above or below the current price. A buy stop is above the current price (expecting a breakout); a sell stop is below (expecting a breakdown). Once the price reaches the stop level, it becomes a market order.","relatedTerms":["limit-order","breakout"],"relatedGuide":null},{"slug":"stop-loss","term":"Stop Loss","category":"order","definition":"A stop loss is an order placed to automatically close a position at a predetermined loss level, limiting potential losses. It is the most important risk management tool. Trailing stop losses move automatically with favorable price action to lock in profits. Every trade should have a stop loss.","relatedTerms":["take-profit","trailing-stop","risk-reward-ratio","stop-loss-sl"],"relatedGuide":"risk-management-for-ea"},{"slug":"take-profit","term":"Take Profit","category":"order","definition":"A take profit order automatically closes a position when it reaches a specified profit level. It removes the emotional aspect of deciding when to exit a winning trade. Combined with stop loss, it defines the complete risk-reward profile of a trade.","relatedTerms":["stop-loss","risk-reward-ratio","take-profit-tp"],"relatedGuide":null},{"slug":"trailing-stop","term":"Trailing Stop","category":"order","definition":"A trailing stop is a dynamic stop loss that moves with the price in your favor, maintaining a fixed distance. If the price reverses by the trailing amount, the position is closed. It allows you to lock in profits while letting winners run. Available as a built-in MetaTrader feature or coded in EAs.","relatedTerms":["stop-loss","take-profit"],"relatedGuide":null},{"slug":"pending-order","term":"Pending Order","category":"order","definition":"A pending order is an instruction to open a position when the price reaches a specified level. Types include buy limit, sell limit, buy stop, and sell stop. MT5 also supports buy stop limit and sell stop limit. Pending orders allow you to enter the market at precise price levels without monitoring charts constantly.","relatedTerms":["limit-order","stop-order"],"relatedGuide":null},{"slug":"scalping","term":"Scalping","category":"strategy","definition":"Scalping is a high-frequency trading strategy that aims to profit from very small price movements. Scalpers open and close many trades within minutes, targeting 5-20 pips per trade. It requires fast execution, tight spreads, and high concentration. Scalping EAs automate this process with precise entry/exit rules.","relatedTerms":["day-trading","spread","pip"],"relatedGuide":"scalping-strategy-explained"},{"slug":"day-trading","term":"Day Trading","category":"strategy","definition":"Day trading involves opening and closing all positions within the same trading day, avoiding overnight risk and swap charges. Day traders analyze intraday charts (M5-H1) and make multiple trades daily. It requires active monitoring and disciplined risk management.","relatedTerms":["scalping","swing-trading"],"relatedGuide":null},{"slug":"swing-trading","term":"Swing Trading","category":"strategy","definition":"Swing trading aims to capture medium-term price moves over days to weeks. Swing traders use H4 and D1 timeframes, focusing on trend changes, pullbacks, and breakouts. It requires less screen time than scalping or day trading but involves overnight and weekend risk.","relatedTerms":["day-trading","trend-following"],"relatedGuide":null},{"slug":"trend-following","term":"Trend Following","category":"strategy","definition":"Trend following is a strategy that enters trades in the direction of the prevailing trend and exits when the trend reverses. It uses indicators like moving averages, MACD, and ADX to identify and confirm trends. Trend-following EAs are popular for their simplicity and effectiveness in trending markets.","relatedTerms":["trend","moving-average","macd","trend-following-ea"],"relatedGuide":"trend-following-strategy"},{"slug":"grid-trading","term":"Grid Trading","category":"strategy","definition":"Grid trading places multiple buy and sell orders at regular intervals above and below a set price, creating a grid. It profits from market oscillation within a range. Grid EAs automatically manage the grid of orders. Risk: in strongly trending markets, one side of the grid accumulates large losses requiring careful money management.","relatedTerms":["martingale","range-trading"],"relatedGuide":"grid-trading-explained"},{"slug":"martingale","term":"Martingale","category":"strategy","definition":"Martingale is a position-sizing strategy that doubles the lot size after every losing trade to recover all previous losses with one winning trade. While mathematically sound in theory, it carries extreme risk of account blow-up during extended losing streaks. Anti-martingale (reverse) increases size after wins instead.","relatedTerms":["grid-trading","position-sizing"],"relatedGuide":"martingale-strategy-risks"},{"slug":"hedging","term":"Hedging","category":"strategy","definition":"Hedging involves opening opposing positions to reduce risk exposure. In forex, you might go long EUR/USD and long USD/CHF (since they're negatively correlated). Direct hedging opens buy and sell on the same pair simultaneously (supported on MT5 hedge accounts). It limits both losses and profits.","relatedTerms":["correlation","hedging-account"],"relatedGuide":null},{"slug":"range-trading","term":"Range Trading","category":"strategy","definition":"Range trading identifies support and resistance levels in a sideways market and buys at support while selling at resistance. It works best when markets are not trending. Oscillators like RSI and Stochastic help confirm range-bound conditions.","relatedTerms":["support-resistance","rsi","oscillator"],"relatedGuide":null},{"slug":"breakout-strategy","term":"Breakout Strategy","category":"strategy","definition":"A breakout strategy enters trades when price breaks through established support or resistance levels with increased momentum. Traders look for consolidation patterns (triangles, channels, rectangles) and enter when the price breaks out. Volume confirmation and false breakout filters improve success rates.","relatedTerms":["breakout","support-resistance","volume"],"relatedGuide":null},{"slug":"mean-reversion","term":"Mean Reversion","category":"strategy","definition":"Mean reversion assumes that prices tend to return to their average over time. When price deviates significantly from its mean (measured by Bollinger Bands, RSI, etc.), traders enter positions expecting a return to the mean. It works best in ranging markets and for pairs with established trading ranges.","relatedTerms":["bollinger-bands","rsi","range-trading"],"relatedGuide":null},{"slug":"news-trading","term":"News Trading","category":"strategy","definition":"News trading involves taking positions based on economic news releases and events (NFP, interest rate decisions, GDP). Traders either straddle before news (placing pending orders in both directions) or react quickly after the release. High volatility and slippage are common during news events.","relatedTerms":["fundamental-analysis","volatility","slippage"],"relatedGuide":null},{"slug":"copy-trading","term":"Copy Trading","category":"strategy","definition":"Copy trading (or social trading) allows traders to automatically replicate the trades of experienced traders. Platforms like MQL5 Signals, ZuluTrade, and eToro offer copy trading services. Followers select signal providers based on their track record, risk profile, and trading style.","relatedTerms":["expert-advisor"],"relatedGuide":null},{"slug":"algorithmic-trading","term":"Algorithmic Trading","category":"strategy","definition":"Algorithmic trading uses computer programs to execute trades based on predefined rules and mathematical models. It eliminates emotional decisions, enables faster execution, and can process multiple markets simultaneously. Expert Advisors on MetaTrader are a form of algorithmic trading accessible to retail traders.","relatedTerms":["expert-advisor","backtesting","high-frequency-trading"],"relatedGuide":null},{"slug":"high-frequency-trading","term":"High-Frequency Trading (HFT)","category":"strategy","definition":"HFT is a subset of algorithmic trading that executes a very large number of orders at extremely high speeds (milliseconds). It requires specialized hardware, co-located servers, and direct market access. HFT is primarily used by institutional traders and is not practical for retail traders on MetaTrader.","relatedTerms":["algorithmic-trading","scalping"],"relatedGuide":null},{"slug":"price-action","term":"Price Action Trading","category":"strategy","definition":"Price action trading analyzes raw price movements without relying on indicators. Traders use candlestick patterns, chart formations, support/resistance, and trend lines to make decisions. It emphasizes reading the 'story' of the chart and understanding market psychology through price behavior alone.","relatedTerms":["candlestick","support-resistance","trend-line"],"relatedGuide":null},{"slug":"correlation","term":"Currency Correlation","category":"strategy","definition":"Currency correlation measures how two currency pairs move in relation to each other. Positive correlation (e.g., EUR/USD and GBP/USD) means they tend to move in the same direction. Negative correlation (e.g., EUR/USD and USD/CHF) means opposite directions. Understanding correlation helps diversify risk and avoid doubling exposure.","relatedTerms":["hedging","currency-pair","correlation-ea-portfolio"],"relatedGuide":null},{"slug":"order-block","term":"Order Block","category":"analysis","definition":"An order block is a specific candle (typically the last bullish candle before a strong bearish impulse, or vice versa) interpreted in ICT / Smart Money Concepts methodology as the price zone where institutional traders accumulated positions before driving price aggressively in the opposite direction. ICT practitioners use order blocks as anticipated reversal zones — if price returns to the order block after the impulse, they expect resumption of the impulse direction. Detection is rules-based (specific candle structure + impulse threshold), making order blocks one of the more codable ICT concepts for EA automation.","relatedTerms":["fair-value-gap","liquidity-sweep","breaker-block","imbalance","displacement"],"relatedGuide":"/guide/how-to/trade-smart-money-with-ea"},{"slug":"fair-value-gap","term":"Fair Value Gap (FVG)","category":"analysis","definition":"A Fair Value Gap (FVG), also called an imbalance or inefficiency, is a three-candle price-action pattern where the second candle's range produces a gap between the first candle's wick and the third candle's wick. The unfilled portion represents 'inefficient delivery' in ICT terminology — price moved through an area without producing two-sided transactions. ICT methodology anticipates that price will return to fill the FVG as institutional traders provide liquidity to the unfilled zone. FVG detection is mechanically codable; EAs commonly use FVGs as entry triggers with a multi-timeframe confluence filter.","relatedTerms":["order-block","imbalance","displacement","liquidity-sweep"],"relatedGuide":"/guide/how-to/trade-smart-money-with-ea"},{"slug":"liquidity-sweep","term":"Liquidity Sweep","category":"analysis","definition":"A liquidity sweep (also called a 'stop hunt' or 'liquidity grab') is a price-action move that pushes briefly past a notable swing high or swing low to trigger stop-loss orders clustered above/below those levels, then reverses sharply in the opposite direction. ICT methodology interprets liquidity sweeps as deliberate institutional behaviour to harvest retail stop-loss liquidity before initiating the larger directional move. Detection requires identifying significant liquidity pools and confirming the reversal velocity; the EA implementation depends on configurable thresholds for what counts as a 'sweep' vs a genuine breakout.","relatedTerms":["order-block","fair-value-gap","displacement","stop-loss"],"relatedGuide":"/guide/how-to/trade-smart-money-with-ea"},{"slug":"breaker-block","term":"Breaker Block","category":"analysis","definition":"A breaker block in ICT methodology is a previously-valid order block that has been violated by price (closed beyond it) but whose role then reverses — a former bullish order block that's broken downward becomes a bearish breaker, expected to act as resistance on subsequent revisits. The concept models how 'failed support becomes resistance' in classical technical analysis but with stricter ICT-specific rules around what qualifies. Breaker block detection requires tracking order block invalidation events and is moderately codable in EAs, though the reliability of breaker reactions is more debated than original order blocks.","relatedTerms":["order-block","fair-value-gap","displacement","liquidity-sweep"],"relatedGuide":"/guide/how-to/trade-smart-money-with-ea"},{"slug":"imbalance","term":"Imbalance","category":"analysis","definition":"Imbalance in ICT terminology refers to an inefficient price move — a sequence of candles where one side of the order flow dominates so completely that price moves through levels without producing matched bid-ask transactions on the way through. Imbalances overlap conceptually with Fair Value Gaps (FVG) and are sometimes used interchangeably; technically an imbalance is the broader concept and FVG is the specific three-candle visualisation. ICT methodology anticipates imbalances will be rebalanced — price will return to fill the inefficient zone as part of normal order-flow rebalancing. Many ICT EAs use imbalance detection as a primary entry signal.","relatedTerms":["fair-value-gap","order-block","displacement","liquidity-sweep"],"relatedGuide":"/guide/how-to/trade-smart-money-with-ea"},{"slug":"displacement","term":"Displacement","category":"analysis","definition":"Displacement in ICT methodology is a fast, momentum-driven price move (typically 3+ consecutive same-direction candles with above-average range) that signals strong institutional intent. Displacement is what creates order blocks (the last opposite-direction candle before the displacement) and fair value gaps (the imbalance candles within the displacement). ICT practitioners use displacement as confirmation that the trading direction has institutional backing. Displacement detection in EAs is straightforward: measure candle-range deviation from rolling average, confirm directional consistency across N consecutive candles. The challenge is configuring the threshold parameters without overfitting to backtest noise.","relatedTerms":["order-block","fair-value-gap","imbalance","liquidity-sweep"],"relatedGuide":"/guide/how-to/trade-smart-money-with-ea"},{"slug":"sharpe-ratio","term":"Sharpe Ratio","category":"metrics","definition":"The Sharpe ratio is a risk-adjusted return metric defined as (mean return − risk-free rate) ÷ standard deviation of returns. It quantifies how much excess return a strategy generates per unit of volatility. Annualised Sharpe ratios above 1.0 are typically considered acceptable for retail forex EAs, above 2.0 strong, and above 3.0 exceptional but rare in real live operation. The ratio is most informative when comparing strategies of similar style on the same data sample.","relatedTerms":["sortino-ratio","calmar-ratio","profit-factor","recovery-factor","expectancy"],"relatedGuide":null},{"slug":"sortino-ratio","term":"Sortino Ratio","category":"metrics","definition":"The Sortino ratio (Frank Sortino, 1980s) is a risk-adjusted return metric similar to Sharpe but using downside deviation — the standard deviation of negative returns only — in the denominator. The intuition is that upside volatility is desirable and should not be penalised. Sortino ratios above 1.5 are typically considered good for retail EAs, above 2.5 excellent. Sortino is preferred over Sharpe for asymmetric strategies such as trend-following or breakout systems where large upside moves are part of the edge.","relatedTerms":["sharpe-ratio","calmar-ratio","profit-factor","recovery-factor"],"relatedGuide":null},{"slug":"calmar-ratio","term":"Calmar Ratio","category":"metrics","definition":"The Calmar ratio is annualised compound return divided by maximum peak-to-trough drawdown over the measurement window (typically 3 years). It is one of the most intuitive risk-adjusted metrics because both numerator and denominator are quantities traders care about directly: 'how much I made' and 'how bad the worst run was'. A Calmar ratio of 2.0 means the strategy delivered 2× its worst drawdown in annual returns. Retail forex EAs typically run Calmar 1-3; above 5 is rare and should prompt scrutiny.","relatedTerms":["sharpe-ratio","sortino-ratio","drawdown","recovery-factor","profit-factor"],"relatedGuide":null},{"slug":"win-loss-ratio","term":"Win/Loss Ratio","category":"metrics","definition":"The win/loss ratio (also called the reward-to-risk ratio at the trade level) compares the average size of winning trades to the average size of losing trades. Mathematically: |avg winning trade $ amount| ÷ |avg losing trade $ amount|. A ratio of 2.0 means typical winners are twice the size of typical losers. This metric combined with win rate determines expectancy and is essential for evaluating any trading system. Scalpers often run win/loss < 1 (small wins, larger losses) but compensate with high win rate; trend-followers often run win/loss > 2 with lower win rates.","relatedTerms":["expectancy","r-multiple","profit-factor","win-rate"],"relatedGuide":null},{"slug":"recovery-factor","term":"Recovery Factor","category":"metrics","definition":"Recovery factor is a common backtest metric defined as total net profit ÷ maximum drawdown (both absolute amounts). It answers how completely a strategy has compensated for its worst-case loss with cumulative gains. Recovery factor above 5 on a multi-year track is strong; above 10 is excellent. The metric is particularly useful for evaluating long-running EAs because it scales with track length and rewards strategies that continue producing profit beyond their worst drawdown.","relatedTerms":["calmar-ratio","sharpe-ratio","profit-factor","drawdown"],"relatedGuide":null},{"slug":"expectancy","term":"Expectancy","category":"metrics","definition":"Expectancy is the mathematically expected profit (or loss) per trade for a given strategy. Formula: E = (P_win × avgWin) − (P_loss × avgLoss), where P denotes probability and avg denotes average amounts. Positive expectancy means the strategy mathematically expects to make money over a sufficient sample size. The expected R-multiple version expresses expectancy in units of initial risk: E_R = (P_win × avgWinR) − (P_loss × 1). Expectancy is the single most important metric for evaluating whether a strategy has an edge; everything else is execution detail.","relatedTerms":["r-multiple","win-loss-ratio","kelly-criterion","profit-factor"],"relatedGuide":null},{"slug":"r-multiple","term":"R-Multiple","category":"metrics","definition":"An R-multiple (introduced by Van K. Tharp) expresses the outcome of any trade as a multiple of its initial risk. The 'R' is the distance from entry to initial stop-loss in dollars; if entry is 1.1000 with stop at 1.0950 on 0.1 lots, R = $50. A trade that hits target at 1.1100 yields $100 profit, which equals +2R. R-multiples are essential for comparing trades across different position sizes and account sizes, and they normalise expectancy calculations.","relatedTerms":["expectancy","win-loss-ratio","stop-loss","kelly-criterion"],"relatedGuide":null},{"slug":"mae-max-adverse-excursion","term":"Maximum Adverse Excursion (MAE)","category":"metrics","definition":"Maximum Adverse Excursion (MAE) is the largest unrealised loss a trade reached at any point between entry and exit. Introduced by John Sweeney in the 1990s, MAE analysis is a diagnostic tool for stop-loss optimisation: by plotting MAE distributions for winning vs losing trades, traders can identify whether stops are too tight (cutting winners short) or too wide (giving back too much on losers). MAE is measured per trade and analysed in aggregate across the strategy's trade sample.","relatedTerms":["mfe-max-favourable-excursion","stop-loss","r-multiple","expectancy"],"relatedGuide":null},{"slug":"mfe-max-favourable-excursion","term":"Maximum Favourable Excursion (MFE)","category":"metrics","definition":"Maximum Favourable Excursion (MFE) is the largest unrealised profit a trade reached at any point during its life, before closing at the actual exit price. The counterpart to MAE, MFE measures how much profit was 'available' at the trade's peak vs how much was actually realised. The ratio realised_profit / MFE reveals exit-strategy efficiency: trades that realise close to their MFE have well-tuned take-profits or trailing logic; large gaps between MFE and realised indicate exit logic that gives back too much.","relatedTerms":["mae-max-adverse-excursion","take-profit","trailing-stop","r-multiple"],"relatedGuide":null},{"slug":"kelly-criterion","term":"Kelly Criterion","category":"metrics","definition":"The Kelly Criterion (John L. Kelly Jr., 1956) is a formula for optimal bet sizing that maximises the long-term geometric growth rate of capital. For a binary outcome: f* = (bp − q) / b, where f* is the optimal fraction of capital to risk, b is the win/loss payoff ratio, p is the win probability, and q = 1 − p is the loss probability. Full Kelly maximises geometric return but produces very high variance — most practitioners use fractional Kelly (typically 25% or 50% Kelly) to balance growth and drawdown.","relatedTerms":["r-multiple","expectancy","win-loss-ratio","position-sizing"],"relatedGuide":null},{"slug":"requote","term":"Requote","category":"execution","definition":"A requote happens when the broker rejects the original order price and presents a new price for the trader (or EA) to accept or decline. The cause is usually fast price movement during the order's transit: by the time the order reaches the dealer, the original quote is no longer available. Requotes are frequent at retail market-maker brokers (dealing desk model) and during high-volatility events. They are operationally damaging for scalping EAs because the delay between request and confirmation breaks the original entry rationale.","relatedTerms":["slippage","spread-spike","partial-fill","off-quote","a-book-vs-b-book"],"relatedGuide":null},{"slug":"partial-fill","term":"Partial Fill","category":"execution","definition":"A partial fill is the execution of less than the requested order volume at the requested price. For example, an order for 10 lots may be filled as 6 lots at the original price, with the remaining 4 lots either filled at a different price (slippage) or rejected. Partial fills are common in fast markets, low-liquidity windows, and large-volume orders. EAs that don't explicitly handle partial fills can end up with inconsistent position state — a stop-loss attached to 'the trade' may only protect the filled portion.","relatedTerms":["slippage","requote","off-quote","time-in-force-gtc-fok"],"relatedGuide":null},{"slug":"off-quote","term":"Off-Quote","category":"execution","definition":"Off-quote (MetaTrader error 136, `ERR_OFF_QUOTES`) occurs when the broker rejects an order because the requested price is no longer at the current market quote — typically because the price has moved significantly since the EA captured it. Common during fast markets, after weekend rollovers, and when EA logic uses cached prices that have become stale. Resolution is to refresh the symbol's quote with `SymbolInfoTick` and resubmit, or adjust the slippage tolerance to permit a wider acceptable range.","relatedTerms":["requote","slippage","spread-spike","virtual-hosting"],"relatedGuide":null},{"slug":"spread-spike","term":"Spread Spike","category":"execution","definition":"A spread spike is a sudden widening of the bid-ask spread, typically during low-liquidity periods (rollover at 22:00-23:00 GMT for many brokers), around major news releases, or during illiquid Asian sessions. EURUSD might normally trade at 0.1-0.3 pips spread but briefly see 3-8 pips during a spike. The danger for EAs: stop-loss orders execute at the bid (for longs) or ask (for shorts), and a spread spike can trigger stops that wouldn't otherwise be hit by 'real' price action.","relatedTerms":["slippage","stop-loss","requote","off-quote","swap"],"relatedGuide":null},{"slug":"last-look-execution","term":"Last-Look Execution","category":"execution","definition":"Last-look execution is a controversial market-making practice where the liquidity provider (LP) or broker can reject an incoming order after a brief 'last look' window (typically 50-200ms). During this window, the LP sees the order, can compare it against current market conditions, and can decline to fill if the trade would be unprofitable. Last-look advantages the broker but disadvantages the trader, especially for time-sensitive strategies. ECN brokers without last-look (no-LP rejection) are preferred for serious algo trading.","relatedTerms":["a-book-vs-b-book","stp-ndd-broker","fix-api","requote","latency-arbitrage"],"relatedGuide":null},{"slug":"latency-arbitrage","term":"Latency Arbitrage","category":"execution","definition":"Latency arbitrage exploits price feed delays between brokers — a faster trader sees the new market price before slower brokers' quotes update, allowing them to trade at the slower broker's stale (now mispriced) quote. The strategy was profitable in the early 2000s with substantial speed advantage; today, brokers actively detect and combat it (rejecting orders, widening spreads for fast traders, banning accounts). Almost impossible at retail scale; institutional latency arb requires co-location and proprietary infrastructure.","relatedTerms":["last-look-execution","a-book-vs-b-book","stp-ndd-broker","fix-api","slippage"],"relatedGuide":null},{"slug":"virtual-hosting","term":"Virtual Hosting (VPS)","category":"execution","definition":"A Virtual Private Server (VPS) is a remote server (essentially a Windows machine running in a data centre) that runs MetaTrader 4/5 and EAs continuously without dependency on the trader's home PC or internet connection. VPS provides three benefits: 24/7 uptime (no power outages or PC reboots), low latency to broker servers (especially when co-located in the same data centre), and isolation from local network issues. Typical retail VPS cost: $15-$40/month for a basic forex VPS.","relatedTerms":["fix-api","off-quote","latency-arbitrage","slippage"],"relatedGuide":null},{"slug":"a-book-vs-b-book","term":"A-Book vs B-Book Broker","category":"execution","definition":"A-book and B-book describe the two fundamental ways brokers handle client orders. A-book brokers (also called STP or ECN) pass orders through to external liquidity providers and earn from spreads/commissions regardless of client P&L. B-book brokers (market makers) take the opposite side of client trades internally — when the client loses, the broker profits; when the client wins, the broker loses. Most retail brokers use a hybrid model, B-booking small/unprofitable clients and A-booking large/profitable ones.","relatedTerms":["stp-ndd-broker","last-look-execution","fix-api","requote","slippage"],"relatedGuide":null},{"slug":"fix-api","term":"FIX API","category":"execution","definition":"FIX (Financial Information eXchange) is a messaging protocol used across financial markets for real-time order entry, trade execution, and market data delivery. Originally developed in 1992 by Fidelity Investments and Salomon Brothers, FIX is now the industry standard for institutional electronic trading. Forex-specific use includes connecting to ECNs (Currenex, Hotspot, Integral) and prime broker connections. Retail traders typically don't use FIX directly; the equivalent for retail is MT4/MT5 native APIs.","relatedTerms":["a-book-vs-b-book","stp-ndd-broker","latency-arbitrage","virtual-hosting"],"relatedGuide":null},{"slug":"stp-ndd-broker","term":"STP / NDD Broker","category":"execution","definition":"STP (Straight-Through Processing) and NDD (No Dealing Desk) are functionally similar broker-model descriptions: orders pass directly to liquidity providers (LPs) without internal market-making or human intervention. The broker earns from spread markup or per-lot commission rather than from being on the opposite side of client trades. STP/NDD overlaps substantially with ECN (Electronic Communication Network) brokers; the terms are often used interchangeably though technically distinct.","relatedTerms":["a-book-vs-b-book","fix-api","last-look-execution","spread","requote"],"relatedGuide":null},{"slug":"breakeven-stop","term":"Breakeven Stop","category":"order","definition":"A breakeven stop is a stop-loss adjustment rule that moves the original stop to the entry price (or slightly beyond, accounting for spread/commission) once the trade has reached a configurable profit threshold. The threshold is typically a multiple of risk (e.g. when price reaches +1R, move stop to entry +0.1R). After breakeven activates, the trade cannot become a loss — but small price retracement can stop out trades that would have continued favourably. Standard feature in trend-following EAs.","relatedTerms":["stop-loss","trailing-stop","take-profit","r-multiple","partial-close"],"relatedGuide":null},{"slug":"oco-order","term":"OCO Order (One-Cancels-Other)","category":"order","definition":"OCO (One-Cancels-Other) is a pair of pending orders linked such that the execution of either order automatically cancels the other. The most common form is the stop-loss + take-profit pair attached to an open position — when either is hit, the position closes and the other is cancelled. Standalone OCO setups can also be used for entry: pending buy-stop + pending sell-stop (cancellation on first fill) is a common breakout-entry pattern.","relatedTerms":["stop-loss","take-profit","stop-limit","trailing-stop"],"relatedGuide":null},{"slug":"stop-limit","term":"Stop-Limit Order","category":"order","definition":"A stop-limit order combines two parameters: a stop price (the trigger level) and a limit price (the worst acceptable execution price). When the market reaches the stop, a limit order is placed at the limit price. This caps adverse slippage on entry/exit at the cost of potentially missing fills if price moves through the limit without filling. Used in scenarios where execution-quality matters more than guaranteed fill, e.g. breakout entries during news where slippage can be extreme.","relatedTerms":["stop-loss","trailing-stop","oco-order","partial-fill","slippage"],"relatedGuide":null},{"slug":"market-on-open","term":"Market-on-Open Order","category":"order","definition":"Market-on-open (MOO) is technically a stock-market order type that executes at the day's opening auction. Forex markets trade continuously without exchange-style opening auctions, so MOO in forex context means a market order timed to session opens — London open (08:00 GMT), NY open (14:00 GMT), or Asian open (Tokyo open). EAs implementing 'open-of-session' entries typically check exact timing against session-specific UTC offsets and place market orders within a precise window (e.g. first 60 seconds after London open).","relatedTerms":["oco-order","stop-limit","spread-spike","swap"],"relatedGuide":null},{"slug":"partial-close","term":"Partial Close","category":"order","definition":"A partial close reduces the size of an open position without closing it entirely. For example, a position of 1.0 lot can be partially closed by 0.5 lots, leaving 0.5 lots open. This is a common trade-management technique: close half the position at a profit target to lock in gains, leaving the remainder to run for larger profit potential. Often combined with moving the stop to breakeven on the remaining position — creating a 'free trade' that can only win or break even.","relatedTerms":["take-profit","trailing-stop","breakeven-stop","r-multiple","partial-fill"],"relatedGuide":null},{"slug":"time-in-force-gtc-fok","term":"Time-in-Force (GTC, FOK, IOC, GTD)","category":"order","definition":"Time-in-force (TIF) is an order parameter specifying the lifetime and fill-policy of a pending order. Common values: GTC (Good Till Cancelled — stays active indefinitely until filled or manually cancelled), GTD (Good Till Date — active until a specified expiration), FOK (Fill or Kill — must fill the entire order immediately or be cancelled), IOC (Immediate or Cancel — fill what is immediately available and cancel the unfilled remainder), DAY (active only for the current trading day). Different TIF values produce different fill behaviour and slippage characteristics.","relatedTerms":["partial-fill","stop-limit","oco-order","off-quote"],"relatedGuide":null},{"slug":"set-file","term":".set File (MetaTrader EA Preset)","category":"files-config","definition":"A .set file is a plain-text MetaTrader preset file that stores EA input parameter values. The file is loaded via the EA's input dialog using the 'Load' button, applying all saved parameter values at once. Vendors typically distribute recommended .set files alongside their EAs (e.g. 'EURUSD_M5.set', 'XAUUSD_aggressive.set'), and traders create their own .set files after parameter optimisation. Files are human-readable text with `parameter_name=value` lines.","relatedTerms":["tpl-template","magic-number","mql5-symbol-mapping"],"relatedGuide":null},{"slug":"tpl-template","term":".tpl File (MetaTrader Chart Template)","category":"files-config","definition":"A .tpl file is a MetaTrader template that stores a complete chart configuration: applied indicators with their parameters, EA attachment, colour scheme, time scale, drawing objects, and chart-window properties. Save a setup via 'Charts → Template → Save Template'; apply via 'Template → Load'. Vendors distribute .tpl files alongside EAs to recreate the recommended chart layout (indicators, colours, EA settings). Different from .set files, which store only EA input parameters.","relatedTerms":["set-file","magic-number","mql5-symbol-mapping"],"relatedGuide":null},{"slug":"magic-number","term":"Magic Number (EA Identifier)","category":"files-config","definition":"A magic number is an integer (typically 4-8 digits) attached to every order placed by an EA via the MQL5 `magic` parameter in trade requests. The EA filters its own positions by magic number, ignoring trades from other EAs or manual trades. This is the standard mechanism for multi-EA coexistence on a single account: each EA uses a unique magic number, and each EA manages only positions with its own magic number. Conflicts between EAs sharing magic numbers cause one EA to modify or close another's positions.","relatedTerms":["set-file","tpl-template","mql5-symbol-mapping"],"relatedGuide":null},{"slug":"mql5-symbol-mapping","term":"MQL5 Symbol Mapping","category":"files-config","definition":"MQL5 symbol mapping is the practice of resolving instrument names across brokers that use different naming conventions. The same instrument (EUR/USD) may be called 'EURUSD' at one broker, 'EURUSD.r' at another, 'EURUSDpro' at a third, and 'EURUSD-T2' at an ECN broker offering tiered liquidity. EAs that hardcode symbol names break when run on different brokers; robust EAs detect and adapt to the broker's naming convention either via input parameter or pattern-matching against `SymbolsTotal()`.","relatedTerms":["set-file","tpl-template","magic-number"],"relatedGuide":null},{"slug":"pattern-recognition-trading","term":"Pattern Recognition (in Trading)","category":"ai-ml","definition":"Pattern recognition in algorithmic trading is the automated identification of recurring price-action or indicator structures that historically preceded specific market behaviour. Implementations range from rules-based detection (e.g. 'engulfing candle = 2nd candle's body fully contains 1st candle's body') to machine learning models (CNNs trained on chart images, LSTMs on price sequences). Pattern recognition is a foundational AI/ML application in trading, used in scalping, breakout, reversal, and structure-based strategies.","relatedTerms":["overfitting-curve-fitting","supervised-learning-trading","walk-forward-optimization","ensemble-model"],"relatedGuide":null},{"slug":"supervised-learning-trading","term":"Supervised Learning (in Trading)","category":"ai-ml","definition":"Supervised learning is the machine learning paradigm where models learn a mapping from input features (X) to output targets (y) by training on labelled examples. In trading: features X might include recent returns, volatility, volume, technical indicators, time-of-day; target y might be the binary outcome 'price up or down 30 minutes from now' or the continuous outcome 'return over next 30 minutes'. Common algorithms: gradient-boosted trees (XGBoost, LightGBM), neural networks, support vector machines, logistic regression.","relatedTerms":["pattern-recognition-trading","overfitting-curve-fitting","walk-forward-optimization","regime-detection","ensemble-model","reinforcement-learning-trading"],"relatedGuide":null},{"slug":"walk-forward-optimization","term":"Walk-Forward Optimization","category":"ai-ml","definition":"Walk-forward optimisation (WFO) is a backtesting methodology that simulates real-time strategy adaptation. The process: (1) optimise parameters on data from window T1, (2) test on out-of-sample window T2, (3) advance both windows forward, (4) optimise again on data from T1+step, test on T2+step, etc. The result is a series of out-of-sample test results that approximate how the strategy would have performed in live trading if parameters were periodically reoptimised. Far more reliable than single in-sample optimisation.","relatedTerms":["overfitting-curve-fitting","supervised-learning-trading","pattern-recognition-trading","regime-detection","walk-forward-test","walk-forward-analysis"],"relatedGuide":null},{"slug":"overfitting-curve-fitting","term":"Overfitting / Curve-Fitting","category":"ai-ml","definition":"Overfitting (in trading often called curve-fitting) is the failure mode where a model or strategy is so tightly tuned to historical data that it captures noise rather than genuine signal. The strategy looks excellent on the data used to develop it but fails on new data — because the patterns it 'learned' were random rather than recurring. Overfitting is the single most common failure mode of retail-developed EAs and the reason walk-forward validation is essential for credible strategy development.","relatedTerms":["walk-forward-optimization","supervised-learning-trading","regime-detection","overfitting","curve-fitting"],"relatedGuide":null},{"slug":"regime-detection","term":"Regime Detection","category":"ai-ml","definition":"Regime detection identifies the current 'market state' — a discrete or continuous characterisation of how the market is behaving. Common regime distinctions: trending vs ranging, high-volatility vs low-volatility, risk-on vs risk-off, mean-reverting vs momentum. Implementation methods range from simple indicator thresholds (ADX > 25 = trending) to statistical models (Hidden Markov Models) to ML classifiers. Strategies use regime detection to switch parameters, activate/deactivate, or rebalance across multiple sub-strategies.","relatedTerms":["supervised-learning-trading","overfitting-curve-fitting","ensemble-model","pattern-recognition-trading"],"relatedGuide":null},{"slug":"ensemble-model","term":"Ensemble Model","category":"ai-ml","definition":"An ensemble model combines predictions from multiple base models (called 'weak learners') to produce a final prediction that is typically more accurate and robust than any single base model. Common ensemble patterns: bagging (parallel training on bootstrap samples, e.g. Random Forest), boosting (sequential training with each model focusing on previous errors, e.g. XGBoost), and stacking (a meta-model learns to combine base-model outputs). Widely used in algorithmic trading because variance reduction and overfitting resistance are critical.","relatedTerms":["supervised-learning-trading","overfitting-curve-fitting","regime-detection","walk-forward-optimization"],"relatedGuide":null},{"slug":"reinforcement-learning-trading","term":"Reinforcement Learning (in Trading)","category":"ai-ml","definition":"Reinforcement learning (RL) is the machine learning paradigm where an agent learns optimal behaviour through trial and error in an environment, guided by a reward signal. In trading applications: the agent observes market state, takes actions (open long, open short, close, hold), and receives rewards (profits) or penalties (losses). RL frameworks include Q-learning, policy gradient methods, and actor-critic models. While theoretically attractive for trading, RL has produced limited live success vs supervised learning due to data inefficiency, non-stationarity, and reward signal noise.","relatedTerms":["supervised-learning-trading","pattern-recognition-trading","ensemble-model","regime-detection"],"relatedGuide":null},{"slug":"chargeback","term":"Chargeback","category":"compliance","definition":"A chargeback is a card-network mechanism that reverses a card payment by the buyer's issuing bank, debiting the merchant. In retail forex it is the principal recovery channel when a broker takes a deposit and then refuses or stalls a withdrawal, falsifies trades, or vanishes. The buyer files a dispute with their bank citing a specific Visa or Mastercard reason code (commonly 'services not provided' or 'not as described'). The bank investigates, requests evidence (deposit screenshots, broker correspondence, withdrawal requests), and on a successful claim returns the funds. Chargebacks have time limits — typically 120 days from the transaction or expected delivery — which is why early action matters. Bank-transfer deposits are not covered. Cryptocurrency deposits are not covered. Card-funded deposits to regulated brokers in scope of consumer protection are the strongest case.","relatedTerms":["regulator-complaint","withdrawal-trap","kyc-know-your-customer"],"relatedGuide":null},{"slug":"kyc-know-your-customer","term":"KYC (Know Your Customer)","category":"compliance","definition":"Know Your Customer (KYC) is the identification, verification, and ongoing-monitoring obligation that every regulated financial institution must apply to new and existing clients. In retail forex it manifests as a sign-up step that collects passport or ID, proof of address (utility bill, bank statement), and sometimes proof of source of funds. Regulators (FCA, CySEC, ASIC, FINRA-affiliated NFA, etc.) require this under anti-money-laundering law; failure to comply is a licence-revoking offence. For traders, KYC has two implications: legitimate brokers will always perform it before a withdrawal is processed, and the absence of KYC is a clear early signal that the operator is unregulated. A 'no-KYC' broker is not friendlier — it is a flag.","relatedTerms":["regulator-complaint","ubo-ultimate-beneficial-owner","sanctions-pep-screening"],"relatedGuide":null},{"slug":"ubo-ultimate-beneficial-owner","term":"UBO (Ultimate Beneficial Owner)","category":"compliance","definition":"Ultimate Beneficial Owner is the natural person — not the holding company — who exercises ultimate control over a legal entity. Anti-money-laundering law in most jurisdictions defines UBO via a 25%+ ownership or voting threshold, with carve-outs for indirect control. Regulators require firms to disclose UBOs to authorities; in the EU and UK the data is partly public via registries (Companies House, RegDataX). For a forex consumer this matters because if a broker collapses or commits fraud, recovery and legal action target the UBO, not an offshore shell. Scams routinely hide UBOs behind nominee directors and chain-of-shells across BVI, Seychelles, Vanuatu, Marshall Islands. Inability to identify a UBO is a credible fraud signal.","relatedTerms":["appointed-representative-ar","kyc-know-your-customer","regulator-complaint"],"relatedGuide":null},{"slug":"appointed-representative-ar","term":"Appointed Representative (AR)","category":"compliance","definition":"An Appointed Representative (AR), in UK FCA terminology and analogous structures in other jurisdictions, is a firm that conducts regulated activity under the umbrella of a 'principal' firm that holds the actual authorisation. The principal is responsible for the AR's conduct. The structure is legitimate for genuine introducer networks, but is also used by lower-quality operators to advertise an FCA reference number that actually belongs to an unrelated principal firm. For a trader checking credentials, the test is to look up the firm on the FCA register and confirm both the AR status and the principal firm's standing. A firm presenting itself as 'FCA authorised' while being an AR of a thinly-related principal is using ambiguity to imply more accountability than exists.","relatedTerms":["ubo-ultimate-beneficial-owner","regulator","regulator-complaint"],"relatedGuide":null},{"slug":"regulator-complaint","term":"Regulator Complaint","category":"compliance","definition":"A regulator complaint is a formal submission to a financial-services regulator alleging misconduct by a firm in its jurisdiction. Common grounds include refusal to process withdrawals, falsified trading conditions, mis-selling of leveraged products, and breach of conduct rules. Each regulator (FCA UK, CySEC Cyprus, ASIC Australia, NFA US, ESMA-coordinated EU regimes) publishes an online complaint form. The complaint triggers a regulatory review and, in cases of suspected fraud, an investigation. For unregulated firms the regulator typically has no jurisdiction but will often publish a warning notice that materially impacts the firm's reputation and search visibility. Filing is a free, formal escalation that runs in parallel to a chargeback and is documented evidence in subsequent civil or criminal proceedings.","relatedTerms":["chargeback","regulator","appointed-representative-ar"],"relatedGuide":null},{"slug":"sanctions-pep-screening","term":"Sanctions & PEP Screening","category":"compliance","definition":"Sanctions and Politically Exposed Persons (PEP) screening is the part of KYC that checks a new or existing customer against published sanctions lists (US OFAC SDN, UN consolidated, EU restrictive measures, UK HMT) and PEP databases. The objective is anti-money-laundering enforcement: a firm that accepts a sanctioned counterparty or a high-risk PEP without enhanced due diligence breaches both criminal law and licensing conditions. In a regulated forex onboarding the check is automated against third-party data vendors (Refinitiv World-Check, Dow Jones Risk Center, ComplyAdvantage) and runs both at account opening and periodically thereafter. For a trader the relevance is indirect — it is one of the operational controls that distinguishes a compliant broker from a shell. A broker that visibly skips KYC is also skipping sanctions screening.","relatedTerms":["kyc-know-your-customer","ubo-ultimate-beneficial-owner","regulator-complaint"],"relatedGuide":null},{"slug":"custody-in-trading","term":"Custody (in trading)","category":"compliance","definition":"Custody is the structural question of who holds the customer's assets — cash deposits, position collateral, settled profits — and under what legal protection. In a regulated retail-forex setup, client money must be held in segregated accounts at a third-party custodian bank, ring-fenced from the broker's own balance sheet, so that broker insolvency does not consume client funds. The FCA CASS rules, EU MiFID II Article 16, and analogous regimes prescribe the operational detail. In an unregulated setup, client deposits go to a general-purpose corporate bank account and are commingled with operating funds; if the firm fails, depositors become unsecured creditors and recover little or nothing. The custody question is more important than headline trading conditions because it governs what happens in the failure case rather than the steady state.","relatedTerms":["regulator","verifiable-performance","withdrawal-trap"],"relatedGuide":null},{"slug":"withdrawal-trap","term":"Withdrawal Trap","category":"compliance","definition":"A withdrawal trap is the operational pattern, common to forex and CFD scams, where deposits process instantly but withdrawals encounter a wall of friction or refusal. Variants include: bonus terms that retroactively void all profits if any withdrawal is requested; minimum-trade-volume requirements that make any cashout economically impossible; sudden 'enhanced KYC' demands for documents already supplied; manufactured account-security freezes that require a new deposit to 'unlock'; and silent non-response once a sufficiently large balance accumulates. The defining feature is asymmetric processing speed — deposits in minutes, withdrawals in months. Defence is preventive: only deposit by card (chargeback eligible) to firms with verified live track records of paid withdrawals, and test with a small withdrawal before scaling capital.","relatedTerms":["chargeback","verified-track-record","vendor-transparency"],"relatedGuide":null},{"slug":"verifiable-performance","term":"Verifiable Performance","category":"compliance","definition":"Verifiable performance is the principle that trading-results claims by a strategy provider, signal seller, EA vendor, or asset manager must be supported by independent, tamper-resistant evidence rather than self-published screenshots or backtests. Standard forms include MyFxBook or FX Blue live-account verification (read-only API access exposing every trade), broker-issued statements, and third-party audits by a recognised firm. The threshold quality is third-party data hosting plus broker-API authentication: the manager cannot edit or selectively delete trades. For commercial EAs the equivalent is publishing the MQL5.com Signals page with full history. Backtest reports alone are not verifiable performance — they describe what the rules would have done on historical data, not what they did with real capital. Decisions to fund a strategy should rest on verifiable performance, not on marketing visuals.","relatedTerms":["verified-track-record","live-track-record","vendor-transparency"],"relatedGuide":null},{"slug":"verified-track-record","term":"Verified Track Record","category":"compliance","definition":"A verified track record is a published history of live trades whose authenticity is guaranteed by an independent custodian holding read-only broker-API credentials. The dominant platforms in retail forex are MyFxBook, FX Blue, and the MQL5.com Signals marketplace. Each connects to the broker via an investor password or API key and ingests every trade in real time; the manager cannot delete losing trades or edit history retrospectively. Track-record length matters — a six-month record covers one regime, while a three-year record exposes the strategy to at least one volatility spike and one trend reversal. The combination of independent custody, broker-API authentication, sufficient duration, and a real account (not demo) constitutes the minimum credibility bar for any product seeking capital.","relatedTerms":["verifiable-performance","live-track-record","vendor-transparency"],"relatedGuide":null},{"slug":"live-track-record","term":"Live Track Record","category":"compliance","definition":"A live track record is a trading history accumulated on a real-money, broker-funded account — as opposed to a demo (paper) account or a backtest report. The distinction matters because demo accounts hide slippage, latency, requotes, partial fills, and broker-side risk management; backtest reports hide future market regimes entirely. Strategies that look profitable on demo or in backtest frequently fail live because the assumed fill quality does not survive contact with a live order book. A live record of meaningful duration (twelve months at minimum, ideally three years) at the broker the strategy will be deployed on is the only honest evidence that a commercial EA or signal service performs as advertised. Anything shorter or any demo/backtest substitute should be treated as marketing material, not data.","relatedTerms":["verified-track-record","verifiable-performance","demo-account-parity"],"relatedGuide":null},{"slug":"vendor-transparency","term":"Vendor Transparency","category":"compliance","definition":"Vendor transparency is the publication of the operational and accountability facts a buyer needs to assess a commercial trading product or service: the registered company name and jurisdiction, regulatory licences if any, names and verifiable identities of principals, an independently verified live track record, a documented drawdown profile across all relevant market regimes, refund and cancellation terms, and the substantive basis of any AI, machine-learning, or quantitative claims. Genuine vendors publish this proactively because they survive scrutiny. Scams and low-quality operators obscure each layer — anonymous principals, generic stock-photo team pages, undocumented offshore shells, backtest screenshots in place of live records, and AI marketing language without method. Vendor transparency is the integrated signal: a vendor that publishes everything tends to be a vendor with nothing to hide; selective disclosure is the dominant fraud pattern.","relatedTerms":["verifiable-performance","verified-track-record","regulator"],"relatedGuide":null},{"slug":"ponzi-scheme","term":"Ponzi Scheme","category":"scam-patterns","definition":"A Ponzi scheme is a fraud structure where claimed investment returns are funded by incoming new-investor capital rather than by any genuine economic activity. Named after Charles Ponzi (Boston, 1920), it is the dominant typology of large-scale forex and crypto investment fraud. The operator markets steady, high, low-volatility returns that are mathematically inconsistent with real market conditions (a 'guaranteed 2% per week' product cannot exist legitimately). Early investors receive payouts and recommend the scheme, generating organic recruitment. Inflows must grow geometrically to service withdrawals; when growth stalls or attracts regulator attention, withdrawals freeze and the scheme collapses. Documented retail-forex Ponzi cases include MTI ($1.7B, CFTC 2021), Forsage ($300M, SEC 2022), and numerous smaller managed-account frauds. The telltale signs are unrealistic smoothness of returns, opacity about strategy, and pressure to recruit.","relatedTerms":["pump-and-dump","guaranteed-returns","managed-trading-offer"],"relatedGuide":null},{"slug":"pump-and-dump","term":"Pump-and-Dump","category":"scam-patterns","definition":"Pump-and-dump is an asset-manipulation fraud where the organisers accumulate a position in a thinly-traded instrument (a micro-cap altcoin, an obscure stock, occasionally an exotic forex pair), then run a coordinated marketing campaign — social media calls, paid 'signals', fabricated news, influencer promotion — to attract retail buyers and inflate the price. Once the price has been pushed up enough, the organisers sell into the manufactured demand, the campaign stops, and the price collapses. Retail participants who bought during the pump are left holding losses. The structure works only on assets with low natural liquidity, where a coordinated small group can move price meaningfully. Regulator actions have targeted hundreds of crypto pump-and-dumps; the SEC's 2021 SafeMoon case ($200M) is a documented forex-adjacent example. Defence is to refuse 'urgent' calls to buy promoted assets, especially with stated price targets.","relatedTerms":["ponzi-scheme","guaranteed-returns","signal"],"relatedGuide":null},{"slug":"guaranteed-returns","term":"Guaranteed Returns","category":"scam-patterns","definition":"A 'guaranteed return' offer in retail forex or CFD trading is a scam by structural definition. Leveraged trading on liquid markets is a probabilistic activity with non-zero risk of loss; no strategy that genuinely participates in market price discovery can guarantee an outcome. The marketing claim therefore signals either a Ponzi structure (where 'guaranteed' returns are funded from new deposits), an outright theft (where no trading occurs and deposits are pocketed), or — at minimum — a regulator-prohibited mis-selling that exposes the buyer to recovery and tax problems. Every major regulator (FCA, SEC, CFTC, ESMA, ASIC, CySEC) explicitly bans the use of 'guaranteed', 'risk-free', or 'no-loss' language in regulated marketing. Encountering this phrase on a forex promotion is sufficient evidence to disengage and report; no due diligence into the underlying offer is required.","relatedTerms":["ponzi-scheme","verifiable-performance","regulator-complaint"],"relatedGuide":null},{"slug":"managed-trading-offer","term":"Managed Trading Offer","category":"scam-patterns","definition":"A managed trading offer is a structure where a customer hands over operational control of trading capital to a third party who is supposed to trade it under an agreed strategy in exchange for a share of profits. The legitimate version is regulated as asset management (MiFID portfolio management licence in EU, RIA registration in US) or as a structured PAMM/MAM account at a regulated broker — both involving segregated client money, audited reporting, and licence-suspended-on-fraud consequences. The unregulated version, marketed through Telegram, Instagram, Discord, and 'mentor' networks, is the dominant retail fraud vector: it allows the operator to take custody of funds, fabricate performance, and exit. Red flags include zero regulatory disclosure, refusal of independently-verified track records, vague 'trading desk' attribution, and pressure to deposit quickly. Legitimate managed offers are slow, paper-heavy, and KYC-intensive; scam offers are fast, friendly, and digital-only.","relatedTerms":["ponzi-scheme","guaranteed-returns","pamm"],"relatedGuide":null},{"slug":"broker","term":"Broker","category":"execution","definition":"A forex broker is the firm that intermediates between a retail trader and the global FX market. It provides the trading account, the platform (MetaTrader 4/5, cTrader, web/mobile clients), the price feed, leverage, and the deposit/withdrawal rails. Brokers differ along multiple dimensions: regulatory status (tier-1 like FCA/ASIC/CySEC vs offshore vs unregulated), execution model (A-book STP/ECN vs B-book market maker vs hybrid), spread and commission structure, available instruments, and operational quality (slippage, requote rate, withdrawal speed). For an automated trader the broker choice is at least as important as the strategy: a strategy with positive backtest expectancy can become unprofitable on a B-book broker with poor execution. Selection should follow strategy: scalping demands ECN with sub-1ms execution; trend-following tolerates wider conditions but still requires regulated custody.","relatedTerms":["a-book-broker","b-book-broker","broker-execution-model"],"relatedGuide":null},{"slug":"broker-execution-model","term":"Broker Execution Model","category":"execution","definition":"Broker execution model is the operational answer to 'what does the broker do with my order after I submit it'. The two pure forms are A-book (the broker forwards the order to one or more external liquidity providers and earns a markup or commission, with profit independent of client P&L) and B-book (the broker takes the opposite side of the trade as principal, profiting when the client loses and losing when the client profits). Most retail brokers operate hybrid models: profitable client cohorts are routed to A-book to offload risk, unprofitable cohorts are B-booked because their losses become broker revenue. The execution-model choice affects spread quality, slippage, requote rate, and structural alignment of incentives. Disclosure is required by FCA, ASIC, and CySEC but is often buried in client agreements. STP/NDD is a marketing label that usually implies A-book.","relatedTerms":["a-book-broker","b-book-broker","broker"],"relatedGuide":null},{"slug":"a-book-broker","term":"A-book Broker","category":"execution","definition":"An A-book broker is one whose business model passes client orders out to external liquidity providers — tier-1 banks, prime brokers, ECN venues — and earns from a fixed commission per lot or a markup added to the raw spread. Because revenue derives from volume, not from client losses, A-book broker incentives are structurally aligned with client trading success: a profitable client generates more volume than a busted one. A-book is typically marketed under STP (Straight-Through Processing), ECN (Electronic Communication Network), or NDD (No Dealing Desk) labels. Real A-book brokers are identifiable by raw or near-raw spreads with a commission, named LP relationships, and willingness to disclose execution-quality statistics. Smaller spreads with no commission usually indicate hidden B-book or hybrid handling. For automated strategies, A-book is generally preferred — especially for scalping and high-frequency systems.","relatedTerms":["b-book-broker","broker-execution-model","broker"],"relatedGuide":null},{"slug":"b-book-broker","term":"B-book Broker","category":"execution","definition":"A B-book broker is a market-maker model where the broker takes the opposite side of client trades on its own book rather than forwarding them to external liquidity. Revenue is the client's net loss. The arrangement is legal and disclosed under regulated regimes (FCA, ASIC, CySEC) and is the dominant model for retail forex globally — most clients lose money over time, so B-book is profitable. The structural conflict of interest is that the broker benefits from client losses and may engage in spread widening, requoting, last-look rejection, and stop-hunting at the margins. Honest B-book operators manage these temptations through internal controls; dishonest ones do not. For active traders, the practical implications are slippage on stop-losses, slow execution at fill events, and worse fills during high-volatility news. Strategies sensitive to execution quality should avoid known B-book operators.","relatedTerms":["a-book-broker","broker-execution-model","broker"],"relatedGuide":null},{"slug":"market-execution","term":"Market Execution","category":"execution","definition":"Market execution is the order-handling mode where the broker accepts a market order and fills it immediately at the best available price at execution time, without intermediate requoting. The trader does not see the actual fill price before submission — only the displayed quote, which may move during the round-trip to the broker. The trade-off is speed (immediate fill, no rejection on price movement) for price uncertainty (positive or negative slippage). Market execution is the default for ECN and most STP setups and is generally preferred for automated strategies that need deterministic fill behaviour. The alternative, 'instant execution', requotes the order if the price has moved beyond a tolerance, giving the trader a price-quality guarantee at the cost of possible rejection. Most retail forex brokers offer both modes; market execution is the modern default.","relatedTerms":["fill-rate","slippage","broker-execution-model"],"relatedGuide":null},{"slug":"fill-rate","term":"Fill Rate","category":"execution","definition":"Fill rate is the proportion of submitted orders that the broker executes at the requested or better price, as opposed to rejecting them, requoting, or filling only partial volume. It is one of the cleanest broker-execution-quality metrics because it cannot be obscured by spread marketing. Healthy ECN and tier-1 STP brokers report fill rates above 97%; rates below 90% indicate either insufficient liquidity, last-look interference by the underlying LP, or systematic B-book rejection. For scalping and other high-frequency strategies, fill rate compounds with slippage to determine real execution cost: a 5% rejection rate eliminates a small statistical edge entirely. Most regulators (FCA, ASIC, MiFID II RTS 27/28) require execution-quality disclosure including fill statistics, but few retail brokers publish them voluntarily — asking is itself a useful filter.","relatedTerms":["market-execution","slippage","last-look-execution"],"relatedGuide":null},{"slug":"co-location","term":"Co-location","category":"execution","definition":"Co-location is the practice of running the trader's algorithmic infrastructure on a server physically housed in the same data centre as the broker's matching engine and price-feed gateway. By eliminating wide-area network hops, round-trip latency drops from typical retail internet figures (30-100 ms) to sub-millisecond figures (often 0.1-1 ms). The economic value depends on strategy class: scalpers and arbitrage strategies extract edge from being first to act on a price change, and co-location is often the difference between profitable and unprofitable; trend-following and swing strategies make decisions on bar closes and gain little from co-location. Major forex venues (LD4 London, NY4 New York, TY3 Tokyo) offer co-location through Equinix and KVH; retail-tier providers (NewYorkCity Servers, BeeksFX) offer affordable proximity hosting that captures most of the benefit at lower cost.","relatedTerms":["persistent-vps","ping","latency-arbitrage"],"relatedGuide":null},{"slug":"ping","term":"Ping","category":"execution","definition":"Ping is the round-trip time in milliseconds for a network packet from the trader's machine (or VPS) to the broker's server and back. It is the primary network-side determinant of execution latency, and consequently of slippage on market orders during fast-moving conditions. Typical figures: retail home broadband to a tier-1 broker, 30-100 ms; consumer-grade VPS in the same continent, 10-40 ms; proximity-hosted VPS in the same data centre region (e.g. LD4 London for European brokers), 1-10 ms; full co-location, sub-1 ms. For scalping strategies, ping over ~20 ms can eliminate edge entirely. For swing or position strategies acting on H1+ timeframes, ping under 200 ms is generally adequate. The MetaTrader terminal reports current ping in the bottom-right status bar; running 'tracert' or 'mtr' to the broker's IP reveals which hop is responsible for latency.","relatedTerms":["co-location","persistent-vps","vps-virtual-private-server"],"relatedGuide":null},{"slug":"negative-balance-protection","term":"Negative-Balance Protection","category":"execution","definition":"Negative-balance protection is a regulatory or contractual guarantee that a retail trader's account balance cannot fall below zero, regardless of market gap, slippage on stop-loss, or any other event that would otherwise produce a debt to the broker. The Swiss-franc unpegging of January 2015 — when EUR/CHF gapped roughly 30% in seconds and many traders' equity went sharply negative — established it as a structural requirement; ESMA mandated it across the EU in 2018, and the FCA, ASIC, BaFin, and CySEC followed. Outside these regimes (offshore brokers, some US futures venues, some institutional accounts) traders remain legally liable for the negative balance. For retail clients the practical implication is asymmetric: a worst-case event with NBP loses the deposit; without NBP, the worst case is an open-ended debt. The protection should be confirmed in writing before any meaningful deposit.","relatedTerms":["margin-call","stop-out","regulator"],"relatedGuide":null},{"slug":"backtest","term":"Backtest","category":"metrics","definition":"A backtest is a simulation of a trading strategy applied to historical price data, producing a synthetic trade-by-trade record of what the strategy would have done over the test window. In MetaTrader the canonical implementation is the Strategy Tester, which replays tick or M1 data through the EA's logic and reports equity curve, profit factor, drawdown, and trade statistics. Backtesting is necessary for strategy development — it validates that the rules are coherent, the implementation is correct, and the broad shape of returns is plausible. It is also the most misused tool in retail trading: developers tune parameters until the backtest looks attractive, embedding selection bias that does not survive the move to live trading. A backtest is genuine evidence only when paired with out-of-sample validation, walk-forward analysis, and (eventually) live track-record confirmation.","relatedTerms":["backtesting","forward-test","walk-forward-test"],"relatedGuide":null},{"slug":"forward-test","term":"Forward Test","category":"metrics","definition":"A forward test is the validation step where a finalised, fully-parameterised strategy is run on incoming live or demo market data going forward in time, after development and backtesting are complete. Because the data is genuinely unseen at the time of model freeze, performance during a forward test is uncorrupted by the parameter selection that inflates backtest results. The standard forward-test duration is at least 2-3 months on a demo account at the broker the strategy will be deployed to, followed by a small-capital live phase before scaling. Forward testing also reveals execution issues that backtesting hides: actual slippage on the broker's quotes, fill rates, requote frequency, and behaviour during news events. A strategy that survives both a robust backtest and a credible forward test is meaningfully more likely to be profitable live than one that survives only the backtest.","relatedTerms":["backtest","forward-testing","walk-forward-test"],"relatedGuide":null},{"slug":"walk-forward-test","term":"Walk-Forward Test","category":"ai-ml","definition":"A walk-forward test (also called walk-forward analysis or rolling-window optimisation) divides the historical price record into a sequence of in-sample / out-of-sample windows. The strategy parameters are optimised on the first in-sample window, then evaluated on the immediately following out-of-sample window without further tuning. The windows then advance by one step — the previous out-of-sample becomes part of the next in-sample — and the process repeats to the end of data. Stitching together the out-of-sample windows produces a simulated equity curve that represents how the strategy would have performed if it had been periodically re-optimised in production. The technique is the canonical defence against curve-fitting because every reported return derives from data the optimiser never saw. A strategy that performs comparably across in-sample and out-of-sample segments is meaningfully robust; large divergence indicates overfitting.","relatedTerms":["walk-forward-optimization","walk-forward-analysis","out-of-sample-test"],"relatedGuide":null},{"slug":"walk-forward-analysis","term":"Walk-Forward Analysis","category":"ai-ml","definition":"Walk-forward analysis extends the walk-forward test by adding formal evaluation metrics on top of the rolling-window simulation. The key output is the walk-forward efficiency ratio (out-of-sample net profit divided by in-sample net profit, expressed as a percentage): ratios above 50% indicate that out-of-sample performance preserves a usable fraction of the in-sample edge, while ratios below 20% strongly suggest overfitting. Secondary outputs include parameter stability heatmaps (whether the optimal parameters drift wildly across windows), out-of-sample drawdown profile, and a comparison of in-sample vs out-of-sample Sharpe. Walk-forward analysis is the gold-standard quantitative validation for retail algorithmic strategies because it forces honest accounting of how much performance survives the move from fitted to unseen data. MetaTrader does not natively implement it; specialised tools (Forex Tester, Strategy Quant, Python frameworks like Backtrader) are typically used.","relatedTerms":["walk-forward-test","walk-forward-optimization","overfitting-curve-fitting"],"relatedGuide":null},{"slug":"look-ahead-bias","term":"Look-Ahead Bias","category":"metrics","definition":"Look-ahead bias is a category of backtest error in which the strategy's decision-making references data that would not have been available at the simulated decision time. Common forms include: using the current bar's close to decide on an entry that supposedly happens at the open; querying a revised economic data series instead of the originally-released figure; using restated earnings data that was not public until months later; indexing into a future bar by off-by-one error in the code; or using point-in-time-invalid technical indicators. The bias is insidious because the backtest report shows attractive returns that cannot be replicated live, since the look-ahead information genuinely is unavailable. Disciplined development uses point-in-time data, runs strict bar-by-bar simulation, and cross-validates backtest results against a careful manual replay of a small sample of trades. Look-ahead bias is the single most common cause of strategies that 'work in backtest but lose live'.","relatedTerms":["overfitting-curve-fitting","backtest","modelling-quality"],"relatedGuide":null},{"slug":"in-sample-period","term":"In-Sample Period","category":"ai-ml","definition":"The in-sample period is the segment of historical price data that the strategy developer uses for hypothesis testing, parameter optimisation, and selection between candidate rule sets. Because the strategy's specification — entries, exits, filters, parameter values — is chosen to perform well on the in-sample data, statistical performance on that same data is by construction biased upward and is not a reliable estimate of future performance. A standard split is 60-80% of available history for in-sample and the remainder for held-out out-of-sample evaluation; walk-forward analysis generalises this with rolling windows. Reporting in-sample backtest results as if they were a performance forecast is one of the most common mis-uses of backtesting; the in-sample figure is more honestly described as 'best case under the assumption that the past repeats exactly'.","relatedTerms":["out-of-sample-period","backtest","overfitting-curve-fitting"],"relatedGuide":null},{"slug":"out-of-sample-period","term":"Out-of-Sample Period","category":"ai-ml","definition":"The out-of-sample period is the segment of historical data that is deliberately withheld from strategy development and reserved for post-hoc validation. Because the strategy's specification was not tuned to fit this data, performance measured here is an honest estimate of how the strategy generalises — and therefore a much closer analogue of expected live performance than in-sample figures. The standard workflow freezes all strategy parameters at the end of development, then runs the unchanged code through the out-of-sample window in a single non-iterative pass. Any tweak made in response to disappointing out-of-sample results re-contaminates the data: the held-out segment becomes part of the development set and a new, smaller out-of-sample window is needed. Walk-forward analysis is the rolling-window generalisation that uses many short out-of-sample segments rather than one large one.","relatedTerms":["in-sample-period","out-of-sample-test","walk-forward-analysis"],"relatedGuide":null},{"slug":"out-of-sample-test","term":"Out-of-Sample Test","category":"ai-ml","definition":"An out-of-sample test is the validation step where a strategy whose parameters have been frozen on an in-sample period is evaluated on a separate, previously unused historical window. The execution is non-iterative: the code runs through the out-of-sample data once, and the results are accepted as the validation outcome regardless of whether they are good or bad. The relationship between in-sample and out-of-sample performance is diagnostic: similar performance across both segments suggests a genuine edge; sharply lower out-of-sample performance suggests curve-fitting; out-of-sample performance comparable to a coin flip suggests no real edge at all. The out-of-sample test is the cheapest credible defence against the most common retail-strategy failure mode — over-optimising on history. It is necessary but not sufficient evidence: a small out-of-sample window can still mislead, and live forward testing remains the final check.","relatedTerms":["out-of-sample-period","in-sample-period","walk-forward-test"],"relatedGuide":null},{"slug":"optimisation","term":"Optimisation","category":"ai-ml","definition":"Optimisation in algorithmic trading is the systematic search across parameter combinations to find values that maximise a chosen objective function — net profit, profit factor, Sharpe ratio, Calmar ratio, or a composite — when applied to historical data. In MetaTrader the Strategy Tester provides brute-force grid search and a genetic-algorithm mode; specialised tools like Strategy Quant offer multi-objective and walk-forward optimisation. The technique is necessary because no developer can guess optimal parameters by hand, but it is also the principal source of overfitting risk: with sufficient parameters and search density, almost any rule set can be made to look good on any history. Defences include using few parameters, optimising on coarse grids, validating on out-of-sample data, comparing optimal parameter stability across windows, and selecting parameter regions rather than single points (preferring robust 'plateau' settings over a single 'peak').","relatedTerms":["genetic-algorithm","walk-forward-test","overfitting-curve-fitting"],"relatedGuide":null},{"slug":"genetic-algorithm","term":"Genetic Algorithm","category":"ai-ml","definition":"A genetic algorithm (GA) is a heuristic optimisation method that searches a parameter space by maintaining a population of candidate solutions, evaluating each on a fitness function, and producing the next generation through selection, crossover, and mutation operators. Higher-fitness candidates are more likely to contribute to the next generation, driving the population toward optimal regions. In MetaTrader 5's Strategy Tester the GA mode is the default for non-trivial parameter spaces because brute-force grid search becomes computationally expensive — a GA can locate near-optimal parameters in a small fraction of the total search space. The trade-off is that GA results are non-deterministic (different random seeds produce different optima) and may converge on local rather than global maxima. For robust strategy development, GA-found parameter regions should still be validated on out-of-sample data, walk-forward tested, and stress-tested across multiple market regimes before being accepted for live trading.","relatedTerms":["optimisation","walk-forward-test","overfitting-curve-fitting"],"relatedGuide":null},{"slug":"tick-data-quality","term":"Tick Data Quality","category":"metrics","definition":"Tick data quality is the fidelity with which a backtest data set captures the actual sequence of bid/ask updates that occurred in the live market. The progression from low to high quality runs: M1 OHLC (open/high/low/close per minute, ignoring intra-bar movement), M1 OHLC + spread, tick data with spread snapshots, and finally full true-tick data with both bid and ask updates at every observed market event. MetaTrader 5's Strategy Tester supports 'Every tick based on real ticks' mode when tick data is available from the broker; the alternative 'Every tick' mode interpolates synthetic ticks from M1 bars, which produces optimistic fill assumptions and misleadingly attractive results for any strategy that touches spread (scalpers, news traders, anything below H1). For credible backtests of execution-sensitive strategies, sourcing real tick data — from the broker, from third-party providers like TickData or Dukascopy, or via paid feeds — is essential.","relatedTerms":["modelling-quality","backtest","slippage"],"relatedGuide":null},{"slug":"modelling-quality","term":"Modelling Quality","category":"metrics","definition":"Modelling quality is MetaTrader's headline metric — displayed as a percentage in the Strategy Tester report — that summarises how faithfully the backtest engine reproduced the live price stream during simulation. Values below 25% indicate that the test ran on M1 OHLC data with little intra-bar resolution; 90% indicates 'Every tick' interpolation mode, which manufactures synthetic ticks from M1 bars; 99% indicates 'Every tick based on real ticks' mode using broker-supplied tick data, which is the only mode credible for execution-sensitive strategy validation. Strategies on H1+ timeframes that exit on bar close are relatively insensitive to modelling quality (the OHLC bar contains the decision-time information). Strategies that touch intra-bar conditions — scalpers, stop-loss-management EAs, anything triggering on spread or bid-ask events — produce wildly different results at low and high modelling quality, and the high-quality result is the only one to trust.","relatedTerms":["tick-data-quality","backtest","strategy-tester"],"relatedGuide":null},{"slug":"curve-fitting","term":"Curve-Fitting","category":"ai-ml","definition":"Curve-fitting is the retail-trading vernacular for overfitting — adjusting a strategy's parameters and structure until the equity curve on a specific historical window looks attractive, where the adjustments are fitting to historical noise rather than to any persistent market behaviour. The signature pattern: a strategy with many parameters, tuned through extensive optimisation, produces a smooth backtest equity curve with high profit factor and low drawdown, then collapses when deployed live or on out-of-sample data. The mechanism is statistical: any sufficiently flexible model can be made to describe any specific data set, but only the systematic component of that data is informative about the future; the noise component is unique to the window. Defences are parameter parsimony, walk-forward validation, scepticism toward 'too good to be true' backtest results, and a willingness to discard otherwise attractive-looking strategies whose out-of-sample performance is materially worse than in-sample.","relatedTerms":["overfitting-curve-fitting","walk-forward-test","look-ahead-bias"],"relatedGuide":null},{"slug":"sample-size","term":"Sample Size","category":"metrics","definition":"Sample size in trading-strategy evaluation is the number of independent trades in the available record. It governs the statistical reliability of every performance metric: a 60% win rate over 20 trades is consistent with anything from 35% to 80% true win rate; the same 60% over 200 trades narrows the confidence interval substantially. Standard guidance is to treat results below 30 trades as suggestive only, results in the 30-100 range as a weak signal, results above 100 as material evidence, and results above ~500 as well-conditioned. For low-frequency strategies (one trade per week) accumulating sufficient sample requires years; for high-frequency strategies it accumulates in weeks. When sample size is small, ranking strategies on point-estimate metrics like Sharpe or profit factor confuses noise with edge — confidence-interval reporting is the discipline that prevents this confusion.","relatedTerms":["expectancy","backtest","win-rate"],"relatedGuide":null},{"slug":"default-value","term":"Default Value","category":"files-config","definition":"A default value is the parameter setting included with an EA or indicator at installation — the value that applies if the trader does not explicitly override it through the inputs dialogue or a .set file. Vendors supply defaults that produce reasonable behaviour during demo and reduce support burden, but defaults are almost never optimal: they were chosen for portability across brokers and symbols rather than for maximum performance on any specific deployment. Production use of an EA on a specific broker and symbol pair requires re-optimisation of timeframe-sensitive and execution-sensitive parameters (lot size, stop and target distances, time filters, news filters). Documented defaults serve as a reference checkpoint — if optimisation drifts wildly from defaults, the result is either a discovery of broker-specific edge or evidence of overfitting; both require investigation before deployment.","relatedTerms":["set-file","magic-number","tpl-template"],"relatedGuide":null},{"slug":"demo-account-parity","term":"Demo Account Parity","category":"platform","definition":"Demo account parity is the degree to which a broker's demo (paper-trading) environment reproduces the conditions traders will experience on a real-money account. Honest brokers run demos against the same price feed and execution engine as live accounts, with identical spread, latency, fill quality, and requote behaviour; the only difference is that orders generate paper P&L rather than real settlement. Many retail brokers, especially in the marketing-driven mass market, run demos with materially better conditions — tighter spreads, instant fills, no requotes, no news-event slippage — to give prospects an attractive first impression. The trader who optimises and validates an EA on such a demo discovers, on switching to a funded account, that the strategy's economics are different: spread is wider, slippage is real, fills during news are degraded. The practical defence is to forward-test on a small live account at the target broker before scaling capital based on demo results.","relatedTerms":["demo-account","forward-test","live-track-record"],"relatedGuide":null},{"slug":"signal","term":"Signal (EA layer)","category":"strategy","definition":"Signal is the EA layer responsible for the trade-decision logic — the rules that read incoming price data and decide whether to enter the market and on which side. A typical signal layer evaluates technical indicators (moving averages, RSI, MACD), price patterns (breakouts, reversals, order blocks), or external inputs (economic-calendar events, sentiment scores) on each new bar or tick, and emits a recommendation: long, short, or stay flat. The signal layer is the most visible part of an EA and the part most discussed in marketing, but it is also the component that contributes least to live performance once strategy class is fixed; quality of risk sizing, filter discipline, and execution typically dominates the long-run equity curve. Strong signal layers are simple, transparent, and parameter-light; complex signal logic with many tunable parameters is the dominant source of curve-fitted EAs.","relatedTerms":["regime-filter","news-filter","signal-parameter"],"relatedGuide":null},{"slug":"signal-parameter","term":"Signal Parameter","category":"strategy","definition":"Signal parameters are the configurable inputs that govern the decision-making logic in an EA's signal layer: the period of a moving average, the overbought/oversold thresholds of an RSI, the lookback window for a range breakout, the number of bars required for a confirmation pattern. Because the signal layer is typically the largest source of strategy expressiveness, signal parameters are also the largest source of overfitting risk during optimisation — the parameter space is wide enough that almost any historical equity curve can be replicated by some combination. Disciplined development keeps signal-parameter count low (under 5-7 tunable inputs is a useful informal target), prefers integer or coarse-grained parameter values, validates on out-of-sample data, and tests parameter robustness by perturbing optimal values 10-20% and confirming that performance degrades gracefully rather than catastrophically.","relatedTerms":["signal","risk-parameter","execution-parameter"],"relatedGuide":null},{"slug":"risk-parameter","term":"Risk Parameter","category":"risk","definition":"Risk parameters are the EA inputs that govern position sizing and loss-limit behaviour — risk-per-trade as a percentage of equity, stop-loss distance in pips or ATR multiples, maximum open positions, maximum portfolio exposure, daily-loss cap, drawdown cutoff. These parameters have outsized impact on long-run results because they determine survival rather than entry quality: a perfect signal layer with reckless sizing blows up the account on the first adverse cluster; a mediocre signal with conservative sizing survives long enough to compound. Risk parameters are also far less subject to curve-fitting than signal parameters because optimal values are constrained by mathematics (Kelly criterion, risk-of-ruin formulae) rather than by historical pattern. Common settings: 0.5-2% risk per trade, daily loss cap at 3-5% of equity, hard EA-stop at 10-15% peak-to-trough drawdown.","relatedTerms":["position-sizing","stop-loss","daily-loss-limit"],"relatedGuide":null},{"slug":"execution-parameter","term":"Execution Parameter","category":"execution","definition":"Execution parameters are the EA inputs that govern how trade decisions are translated into broker-side orders and how those orders are managed once active. Typical execution parameters include order type (market execution vs limit pending), maximum acceptable slippage in pips before the EA rejects a fill, requote-handling retries, partial-close thresholds, trailing-stop activation and step sizes, breakeven-stop triggers, and order modification cadence. Execution parameters are often invisible in marketing material — the buyer sees signal logic and risk caps but not the fill-quality assumptions — yet they dominate live-vs-backtest divergence for scalping and short-horizon strategies. Best practice exposes execution parameters as configurable inputs so the trader can adapt to specific broker conditions; EAs that hardcode tight slippage tolerances or aggressive retry policies often misbehave on slower-execution venues.","relatedTerms":["fill-rate","slippage","market-execution"],"relatedGuide":null},{"slug":"news-filter","term":"News Filter","category":"strategy","definition":"A news filter is an EA component that suppresses trading activity around scheduled high-impact economic announcements. Around major releases — US Non-Farm Payrolls, CPI prints, central bank rate decisions, central bank press conferences — spread widens dramatically (50-200% of typical), slippage spikes on stop-loss orders, fills become unreliable, and the market often moves in spike patterns that defeat technical signal logic. Most retail trend-following and mean-reversion strategies are net-loss during news windows because the execution degradation outweighs any signal advantage. A news filter typically loads an upcoming-events calendar (ForexFactory, Investing.com, or broker-provided JSON), classifies events by impact level, and blocks new entries from N minutes before to M minutes after each red-flag event. News-trading strategies — designed specifically to exploit announcement volatility — are the exception; for everything else, a news filter is one of the highest-ROI improvements available.","relatedTerms":["regime-filter","news-trading","economic-calendar"],"relatedGuide":null},{"slug":"regime-filter","term":"Regime Filter","category":"strategy","definition":"A regime filter is an EA component that classifies the current market state (trending vs ranging, high vs low volatility, risk-on vs risk-off) and enables or disables trading accordingly. The rationale is that most strategies have a natural regime — trend-following strategies make money in trends and lose in chop, mean-reversion strategies make money in ranges and lose during strong directional moves, scalpers need stable spreads and lose during volatility spikes. A regime filter that correctly classifies the current state and disables the strategy when the regime is wrong improves long-run Sharpe by removing the negative-expectancy tail of the equity curve. Common implementations use ADX thresholds for trend strength, ATR percentiles for volatility regime, or Hidden Markov Models for more sophisticated state inference. Like all filters, a regime filter adds parameters and therefore degrees of freedom for overfitting; validation discipline matters.","relatedTerms":["market-regime","regime-detection","news-filter"],"relatedGuide":null},{"slug":"market-regime","term":"Market Regime","category":"ai-ml","definition":"A market regime is a persistent state of the market characterised by distinctive statistical properties of returns, volatility, and correlation. The canonical regime dichotomies are trending vs ranging (directional vs mean-reverting), high vs low volatility, and risk-on vs risk-off (positive correlation among riskier assets vs flight-to-quality). Regimes persist on horizons of weeks to months, then transition with varying degrees of warning. Strategies have natural regimes where their core thesis holds — trend-following needs trending markets, mean-reversion needs ranging, scalpers need stable low-volatility conditions — and regimes where the thesis fails systematically. Regime detection is the discipline of classifying current market state for downstream filtering. Approaches range from simple rules (ADX above 25 = trending) to statistical models (Hidden Markov Models, change-point detection) to machine-learning classifiers. Awareness of regime is a primary differentiator between robust and fragile strategies.","relatedTerms":["regime-detection","regime-filter","trend-following"],"relatedGuide":null},{"slug":"ea-retirement-criteria","term":"EA Retirement Criteria","category":"risk","definition":"EA retirement criteria are the pre-defined, written rules that govern when a live-trading EA is decommissioned — either temporarily suspended or permanently shut down. Typical criteria include: cumulative drawdown crossing a hard threshold (e.g. 2x the worst backtested drawdown); rolling N-month profit-factor falling below a chosen level; consecutive monthly losses exceeding a count; statistically significant divergence between live and backtested equity curve; regime detector flagging persistent unsuitable conditions. The discipline matters because, in the absence of pre-defined criteria, an underperforming EA tends to be left running on hope — 'one more month' becomes another quarter, a 10% drawdown becomes a 25% drawdown. Pre-commitment to retirement criteria — written down before deployment, ideally with automatic enforcement — converts the decision from emotional to mechanical and protects capital during the inevitable degradation phase of every strategy.","relatedTerms":["drawdown-trigger","pre-commitment","review-cadence"],"relatedGuide":null},{"slug":"position-size-cap","term":"Position Size Cap","category":"risk","definition":"A position size cap is a hard maximum on lot size per individual trade, independent of the risk-sizing calculation that would otherwise drive the figure. Its purposes are defensive: protecting against scaling errors as account equity grows (a small bug producing 100x normal size on a million-dollar account is catastrophic in a way the same bug on a thousand-dollar account is not), against single-trade risk that exceeds broker margin or liquidity tolerance, and against execution problems with very large orders (partial fills, last-look rejection, increased slippage). Standard practice is to set the cap at 5-10x the typical normal-conditions position size, providing headroom for legitimate volatility-adjusted sizing while preventing absurd outliers. The cap is enforced at the EA's order-submission point, after the risk calculation but before the order is sent to the broker.","relatedTerms":["position-sizing","risk-parameter","daily-loss-limit"],"relatedGuide":null},{"slug":"daily-loss-limit","term":"Daily Loss Limit","category":"risk","definition":"A daily loss limit is a hard, pre-defined ceiling on the amount of equity an EA is permitted to lose in a single trading day before it disables itself until the following session. The mechanism is enforced inside the EA: at the start of each day the equity is recorded; throughout the day, if cumulative realised + unrealised loss exceeds the threshold, all positions close and no new entries are placed until the next day's session opens. Typical values are 2-5% of starting equity for the day. The defence applies in three failure modes: regime-shift days where the strategy thesis collapses and continued trading magnifies losses; system or data errors where the EA misreads conditions and submits a sequence of mistaken trades; and statistical bad-luck clusters where many trades happen to lose in sequence. Daily limits are required by most proprietary-trading-firm challenges and are best-practice for self-funded operation.","relatedTerms":["daily-loss-cap","maximum-drawdown-cutoff","position-size-cap"],"relatedGuide":null},{"slug":"daily-loss-cap","term":"Daily Loss Cap","category":"risk","definition":"Daily loss cap is functionally identical to daily loss limit — the maximum cumulative loss an EA accepts in a single day before suspending until the next session. The terminology distinction is purely descriptive: 'daily loss cap' is the more common phrasing in proprietary trading firm challenge documentation (FTMO, MyForexFunds, The Funded Trader), where exceeding the cap is grounds for immediate challenge failure. For self-funded retail operation, the mechanism and the recommended value range (2-5% of starting daily equity) are the same. Some implementations include unrealised P&L in the calculation, suspending the EA on open positions when paper losses cross the cap; others count only realised losses and let positions run to their original stops. The unrealised version is more conservative and more common in prop-firm contexts.","relatedTerms":["daily-loss-limit","maximum-drawdown-cutoff","risk-of-ruin"],"relatedGuide":null},{"slug":"maximum-drawdown-cutoff","term":"Maximum Drawdown Cutoff","category":"risk","definition":"Maximum drawdown cutoff is a hard equity threshold — measured as peak-to-trough percentage from the EA's high-water mark — at which the EA permanently shuts itself down, requiring human intervention to restart. Typical values are set at 1.5x to 2x the worst drawdown observed in robust backtest and walk-forward analysis: a strategy whose live drawdown reaches twice the historical worst case has produced statistically meaningful evidence that its thesis is broken in current conditions, and continued operation is more likely to compound losses than to recover. The cutoff is the long-horizon analogue of the daily loss limit: where the daily limit protects against bad days, the maximum drawdown cutoff protects against bad regimes. It is an essential complement to EA retirement criteria — even when the trader fails to act on rolling-underperformance signals, the cutoff stops the bleeding mechanically.","relatedTerms":["drawdown-trigger","ea-retirement-criteria","drawdown"],"relatedGuide":null},{"slug":"drawdown-trigger","term":"Drawdown Trigger","category":"risk","definition":"A drawdown trigger is any rule that initiates a pre-defined defensive action when the EA's drawdown (peak-to-trough equity decline) crosses a specified threshold. Single-stage triggers map to one action — typically full shutdown at the maximum drawdown cutoff. Layered triggers handle escalating severity with graduated responses: a 5% drawdown might halve position size, a 10% drawdown might disable aggressive entries while permitting conservative ones, a 15% drawdown might pause all new trades pending review, and a 20% drawdown might trigger full shutdown. The layered design balances two concerns: avoiding overreaction to ordinary drawdown variance (every strategy has 5% drawdowns) while still responding before drawdowns become unrecoverable. Implementation discipline matters: triggers must be evaluated against current equity and acted on automatically, since the human tendency in drawdown is to override defensive rules out of hope or panic.","relatedTerms":["maximum-drawdown-cutoff","ea-retirement-criteria","rolling-profit-factor"],"relatedGuide":null},{"slug":"mt5-metatrader-5","term":"MT5 (MetaTrader 5)","category":"platform","definition":"MT5 (MetaTrader 5) is the fifth-generation retail trading platform from MetaQuotes Software, succeeding the dominant MT4 platform launched in 2005. MT5 introduces a richer order model (including stop-limit and netting accounts), multi-asset support beyond forex (equities, futures, options where the broker provides them), a faster strategy tester with multi-currency and tick-data modes, an upgraded MQL5 programming language, and a built-in Market and Signals marketplace. The platform is provided to retail traders by brokers under licence and is the de facto standard for retail algorithmic forex trading globally. Most commercial EAs and the educational material on this site target MT5 specifically; MT4 remains supported but is now in long-tail maintenance. The MT5 desktop client, the MetaEditor IDE for MQL5 development, and the Strategy Tester for backtesting form the trader's core toolset.","relatedTerms":["metatrader-5","metaeditor","mql"],"relatedGuide":null},{"slug":"metaeditor","term":"MetaEditor","category":"platform","definition":"MetaEditor is the integrated development environment bundled with the MetaTrader 4 and MetaTrader 5 desktop clients, used for writing, compiling, and debugging code in the MQL4 and MQL5 languages. It supports Expert Advisor projects, custom indicators, scripts, services, and library files; provides a tree-style code navigator, a property-editor for input parameters, syntax-highlighting and code-completion, an integrated debugger that attaches to the Strategy Tester or to live charts, a profiler for performance analysis, and direct upload to the MQL5.com marketplace. MetaEditor is functionally adequate but spartan compared to general-purpose IDEs like VS Code or JetBrains products; some developers use external IDEs with custom toolchains for editing and reserve MetaEditor for compilation and debugging only. Compiled artefacts are .ex4 / .ex5 binary files that the MetaTrader client loads as EAs, indicators, or scripts.","relatedTerms":["mql5-compiler","mql","expert-journal"],"relatedGuide":null},{"slug":"mql5-compiler","term":"MQL5 Compiler","category":"platform","definition":"The MQL5 compiler is the toolchain integrated with MetaEditor that translates MQL5 source code (.mq5 files) into MetaTrader-executable bytecode (.ex5 files). It performs lexical, syntactic, and semantic analysis, then produces an optimised bytecode artefact that runs inside the MetaTrader 5 client's MQL5 virtual machine. Important quirks for production work: the compiler is strict about implicit type conversions and emits warnings that often correspond to subtle runtime bugs (silent precision loss on doubles, sign-mismatched comparisons, dead code paths); it does not enforce strong static typing as strictly as C++ despite syntactic similarity; and it allows certain unsafe constructs (uninitialised pointer dereferencing, out-of-bounds array access) that fail silently at runtime rather than at compile time. Production-grade EA development treats every compiler warning as an error to be addressed before shipping.","relatedTerms":["metaeditor","mql","ex5-file"],"relatedGuide":null},{"slug":"mql5-marketplace","term":"MQL5 Marketplace","category":"platform","definition":"The MQL5 Marketplace is MetaQuotes' official catalogue for distributing commercial Expert Advisors, custom indicators, signals subscriptions, and library code targeting MT4 and MT5. Vendors register on mql5.com, submit binaries that pass an automated review, and list products at fixed or rental pricing; buyers receive the binary directly through MetaEditor's built-in Market panel after payment. The marketplace's structural advantages over off-platform distribution: identity-verified vendor accounts, persistent buyer reviews that cannot be deleted by the vendor, optional live-account performance feeds via the Signals subsystem, and standardised refund mechanisms for non-functional products. Quality varies enormously — the bulk of marketplace EAs are still curve-fitted or grid-based — but the structural transparency is higher than for products sold via Telegram or affiliate channels. Caveat emptor still applies.","relatedTerms":["mt5-metatrader-5","vendor-transparency","verifiable-performance"],"relatedGuide":null},{"slug":"ex5-file","term":".ex5 File","category":"files-config","definition":"A .ex5 file is the compiled binary output of the MQL5 compiler — the executable form of an Expert Advisor, custom indicator, script, or service that the MetaTrader 5 client loads and runs inside its MQL5 virtual machine. The corresponding source file is the .mq5 text file; the .ex5 binary is produced by MetaEditor's compile step. Vendors distribute .ex5 files (not .mq5) to protect strategy intellectual property; buyers receive a binary they can run but cannot read or modify. The .ex5 format is proprietary to MetaQuotes and is partially obfuscated, though full decompilation is technically possible with specialised tools and is a recurring concern for commercial vendors. Buyers should be aware that running a closed-source .ex5 means trusting the vendor with order-submission privileges on their broker account; reputable vendors are accountable through marketplace identity and reviews, anonymous .ex5 distributions are higher risk.","relatedTerms":["mql5-compiler","metaeditor","mql5-marketplace"],"relatedGuide":null},{"slug":"expert-journal","term":"Expert Journal","category":"platform","definition":"The Expert Journal is the log surface in MetaTrader's Toolbox that records every event raised by running Expert Advisors — initialisation messages, order submission requests and responses, deinitialisation reasons, runtime errors, broker rejections, custom Print() output written by the EA developer, and timer events. It is the primary diagnostic surface during both development and live operation: every unexpected EA behaviour leaves a trace there. The log is persisted to disk in the MetaTrader 'Logs' directory and rotates daily by default; entries older than a few weeks may be archived or purged depending on settings. For production EA operation, a discipline of reading the Expert Journal daily — or programmatically tailing it via a watchdog script — catches anomalies (broker rejections, malformed orders, indicator initialisation failures) before they compound into losses.","relatedTerms":["journal-error-count","market-watch","metaeditor"],"relatedGuide":null},{"slug":"market-watch","term":"Market Watch","category":"platform","definition":"The Market Watch window is MetaTrader's symbol-and-quote panel that lists the instruments the trader has selected for active monitoring, with current bid/ask, spread, last update time, and (optionally) high/low/depth-of-market details. Its operational significance for EA users: an Expert Advisor will only receive price-update events for symbols visible in Market Watch. An EA attached to a chart for a symbol that has been removed from Market Watch silently stops receiving ticks and ceases to trade, often without writing any error to the Expert Journal. Routine checks should confirm that all symbols traded by active EAs are present in Market Watch. The 'Show All' option (right-click → Show All) makes every broker-offered symbol visible at once; some traders prefer this configuration to eliminate the silent-failure mode.","relatedTerms":["expert-journal","mt5-metatrader-5","ohlc"],"relatedGuide":null},{"slug":"journal-error-count","term":"Journal Error Count","category":"metrics","definition":"Journal error count is the number of error-severity entries written to the MetaTrader Expert Journal by an EA over a defined measurement window — per session, per day, per week. It is a useful operational health metric because almost every failure mode in live EA operation surfaces as a journal error: broker rejections of orders (price too far from market, invalid stops, account margin issues), indicator initialisation failures after a broker connection drops and reconnects, MQL runtime exceptions, and OS-level resource errors. A stable EA on a stable broker connection produces near-zero errors during normal operation; rising counts indicate a problem requiring investigation — often a degraded broker connection, a configuration drift, or a regime where the EA's expected market conditions no longer hold. Monitoring this count manually or via an external script is part of production-grade EA operation.","relatedTerms":["expert-journal","fill-rate","ea-retirement-criteria"],"relatedGuide":null},{"slug":"arbitrage-ea","term":"Arbitrage EA","category":"strategy","definition":"An arbitrage EA seeks to exploit transient price differences between two or more related markets. Variants include: latency arbitrage (trading the same instrument across two brokers when one lags), statistical arbitrage (trading the spread between correlated pairs that have temporarily diverged), triangular arbitrage (trading round-trips through three currency pairs when cross rates are inconsistent), and cash-vs-CFD arbitrage (trading between equity CFDs and underlying equities). True latency arbitrage requires sub-millisecond execution, co-located servers, and tier-1 broker access, and has been infeasible at retail scale since the mid-2010s; brokers actively detect and ban it. Statistical and triangular arbitrage remain viable at retail scale but are capacity-limited — opportunities are small in pip terms and disappear under heavy volume. Most commercial 'arbitrage EAs' sold to retail traders are not actually arbitrage in the technical sense; they are mean-reversion strategies marketed with the term to imply risk-free returns.","relatedTerms":["latency-arbitrage","co-location","fill-rate"],"relatedGuide":null},{"slug":"grid-ea","term":"Grid EA","category":"strategy","definition":"A grid EA places a sequence of buy or sell orders at fixed price intervals (the 'grid') as price moves against the initial entry, on the principle that adding to a position at lower (or higher) prices reduces the average entry and a subsequent reversal will close the entire grid at profit. The structure produces attractive-looking equity curves in backtests over ranging markets because most grids close profitably; it produces catastrophic losses in sustained trends because each grid step adds exposure to a position that continues moving against the trader. The failure mode is asymmetric: many small wins followed by a single account-ending loss when a trend extends beyond the grid's depth or the account's margin. Grid EAs are heavily marketed to retail buyers because the win rate and equity-curve smoothness look excellent; the buyer typically discovers the tail risk only after deploying real capital. Genuine, robust grid strategies exist but require sophisticated risk management; the mass-market commercial offering rarely does.","relatedTerms":["martingale","grid-trading","risk-of-ruin"],"relatedGuide":null},{"slug":"scalper-ea","term":"Scalper EA","category":"strategy","definition":"A scalper EA targets very small per-trade profits — typically 3-15 pips — by entering and exiting positions over short horizons ranging from seconds to a few hours. The economics rely on high trade frequency and high win rate to compound modest per-trade edges into meaningful aggregate returns. Scalpers are acutely sensitive to execution-quality variables that other strategies tolerate: spread directly consumes a large fraction of target profit, slippage on stop and limit orders changes per-trade P&L sign, fill rate below 95% destroys statistical expectancy, latency above 10ms erodes the edge from time-sensitive entries. Successful scalping requires a tier-1 ECN broker with raw spread plus low commission, co-location or proximity-hosted VPS, and disciplined news filtering. Most retail scalper EAs marketed at low prices fail not because the signal logic is wrong but because the assumed execution conditions do not exist at the buyer's broker.","relatedTerms":["scalping","fill-rate","co-location"],"relatedGuide":null},{"slug":"trend-following-ea","term":"Trend-Following EA","category":"strategy","definition":"A trend-following EA identifies an established directional move (using moving-average crossovers, breakout systems, momentum oscillators, or regime classifiers) and enters in the direction of the trend, holding the position until reversal signals fire. The strategy class has the longest documented track record in algorithmic trading — managed-futures CTAs, hedge funds, and retail EAs alike have run trend-following profitably for decades. The trade-off is psychological difficulty: trend-followers have low win rates (often 30-40%), endure long flat or losing periods between trending markets, and depend on a small minority of trades to produce most of their P&L. In ranging or chop-dominated environments — which constitute most of the calendar most years — trend-following EAs lose small amounts steadily, testing trader discipline. Realistic expectations include drawdowns of 15-30%, recovery periods measured in many months, and total reliance on capturing the few major trends per year.","relatedTerms":["trend-following","trend","regime-filter"],"relatedGuide":null},{"slug":"correlation-ea-portfolio","term":"Correlation (EA portfolio)","category":"strategy","definition":"Correlation in an EA portfolio is the statistical measure of how synchronously the equity curves of multiple Expert Advisors move over time. Two trend-following EAs trading similar timeframes on related currency pairs will have high positive correlation (0.7+) — when one drawdowns, the other typically does too, because both are exposed to the same regime risk. Two strategies with different theses (e.g. a trend-follower and a mean-reverter) on different instruments will have lower or even negative correlation, providing genuine diversification. The portfolio construction discipline is to deliberately combine low-correlation strategies so that aggregate drawdown is lower than any individual EA's drawdown — the only free lunch in finance. Common mistakes in retail EA portfolios include running multiple copies of the same EA on different pairs (high correlation, no diversification) or running visually different EAs that are all variations on the same regime thesis. Measuring portfolio correlation requires several months of live equity data per EA.","relatedTerms":["correlation","portfolio-drawdown","regime-filter"],"relatedGuide":null},{"slug":"risk-of-ruin","term":"Risk of Ruin","category":"risk","definition":"Risk of ruin is the mathematical probability that a trading strategy will at some point reduce the account to zero (or below a defined liquidation threshold) given its empirical win rate, risk-reward ratio, and per-trade risk as a fraction of equity. For a fixed-fractional sizing strategy the formula is approximately RoR ≈ ((1-A)/(1+A))^U, where A is the trader's edge per unit risk and U is the number of units the account can lose before ruin. The metric quantifies the survival question that win-rate and profit-factor headlines hide: a 60%-win-rate, 1:1-RR strategy risking 5% per trade has approximately a 30% probability of eventually busting; the same strategy risking 1% has a sub-1% probability. Conservative position sizing — typically targeting risk of ruin below 1-2% — is the price paid for the right to compound over decades rather than dramatic short-term performance.","relatedTerms":["kelly-criterion","position-sizing","drawdown"],"relatedGuide":null},{"slug":"capital-scaling","term":"Capital Scaling","category":"risk","definition":"Capital scaling is the disciplined process of increasing the trading capital allocated to a strategy in defined increments as live performance accumulates evidence that the strategy works as expected. A typical schedule starts the strategy on the minimum viable capital (often 5-10% of intended target), holds at that level for 2-3 months to confirm live performance matches forward-test expectations, then increases by 50-100% per validation milestone with mandatory pauses between increments. The discipline matters for two reasons: it limits exposure to strategies that fail in live conditions despite passing backtest and forward-test (a non-trivial fraction), and it prevents the psychological scaling-on-winning-streak pattern that converts a fortunate run into a catastrophic regression to the mean. Pre-defined scaling criteria written before deployment work better than ad-hoc judgement during operation.","relatedTerms":["scaling-discipline","pre-commitment","review-cadence"],"relatedGuide":null},{"slug":"scaling-discipline","term":"Scaling Discipline","category":"risk","definition":"Scaling discipline is the practice of mechanically applying pre-defined capital-scaling rules to a live trading strategy, resisting the emotional pulls to deviate. The most common failure modes are upward: a strategy on a winning streak feels increasingly safe, the trader scales capital faster than the schedule permits, and a subsequent regime shift produces drawdown on the inflated capital base. Downward failures also occur: a strategy in normal-variance drawdown feels broken, the trader reduces capital below the planned floor, and the strategy recovers to higher equity than the trader is now exposed to. Empirically, retail traders with positive-expectancy strategies regularly destroy their accounts through scaling-discipline failure rather than through strategy failure. The defence is to commit the scaling schedule to writing before deployment, expose it to a trusted observer, and treat deviations as the major event they actually are.","relatedTerms":["capital-scaling","pre-commitment","review-cadence"],"relatedGuide":null},{"slug":"pre-commitment","term":"Pre-Commitment","category":"risk","definition":"Pre-commitment is the behavioural-economics practice of binding oneself in advance to a course of action that future-self would otherwise be tempted to deviate from. In trading it manifests across multiple horizons: tactical pre-commitment is placing stop-loss orders before any trade is taken (the position must have a stop before submission, not after observing how it moves); operational pre-commitment is writing the capital-scaling schedule and EA retirement criteria before deployment so subsequent decisions are mechanical reference rather than emotional judgement; strategic pre-commitment is choosing the broker, the strategy class, and the maximum acceptable lifetime drawdown before opening the account. The pattern matters because human decision-making degrades sharply under monetary stress — in drawdown, on winning streaks, after losing trades — and the only reliable defence is to make the key decisions before stress is present.","relatedTerms":["scaling-discipline","ea-retirement-criteria","stop-loss"],"relatedGuide":null},{"slug":"review-cadence","term":"Review Cadence","category":"risk","definition":"Review cadence is the scheduled frequency at which a trader formally inspects strategy performance, broker statements, EA journal entries, and operational metrics, with the purpose of catching problems before they compound. A standard cadence includes a weekly tactical review (every trade closed in the past week, journal errors, fill quality, any unexpected EA behaviour) and a monthly strategic review (rolling Sharpe and drawdown vs expectations, regime fit, capital-scaling milestones, retirement-criteria status). The discipline matters because operational problems and strategy degradation are typically gradual: each individual unusual event looks tolerable in isolation, and only the pattern across weeks reveals that intervention is needed. Without a fixed cadence the trader's attention is dragged by salient single events (a big winner, a stop-out, a news spike) and the slow degradation that destroys most accounts goes unnoticed until it is too late to correct cheaply.","relatedTerms":["ea-retirement-criteria","trading-plan","trade-journal"],"relatedGuide":null},{"slug":"rolling-profit-factor","term":"Rolling Profit Factor","category":"metrics","definition":"Rolling profit factor is the profit-factor metric (gross profit divided by gross loss) computed over a moving time window rather than over the strategy's full history. Typical window lengths are 30, 60, and 90 days. The metric is diagnostic in a way that aggregate profit factor is not: a strategy with lifetime PF of 1.5 but rolling 60-day PF that has trended down from 1.8 to 1.1 is showing measurable degradation that warrants investigation, even if the lifetime number still looks healthy. The aggregate metric is dominated by past performance and lags real changes; the rolling metric responds quickly. Used in EA retirement criteria, rolling profit factor below a chosen threshold (e.g. 1.0 for two consecutive windows) is a clear mechanical signal to suspend or shut down. Plotted as a time series, it provides visual evidence of how the strategy is interacting with current regime.","relatedTerms":["profit-factor","ea-retirement-criteria","drawdown-trigger"],"relatedGuide":null},{"slug":"peak-drawdown","term":"Peak Drawdown","category":"metrics","definition":"Peak drawdown — also called maximum drawdown or max DD — is the largest percentage decline from any equity high-water mark to the subsequent low within a measurement window. It is conventionally the single most important risk metric for evaluating a trading strategy because it represents the worst experience the trader had using the strategy, which is the experience that determines whether the trader stays with the strategy. A backtest peak drawdown of 15% over five years means the trader, in real time, would have watched 15% of capital evaporate and waited for recovery; tolerance for that experience is a personal risk-appetite question, not a strategy-quality question. Peak drawdown is asymmetric in recovery: a 50% drawdown requires 100% gain to recover; a 30% drawdown requires 43%. Strategies with attractive returns and high peak drawdowns are often less compoundable than strategies with modest returns and low drawdowns.","relatedTerms":["drawdown","maximum-drawdown","recovery-factor"],"relatedGuide":null},{"slug":"portfolio-drawdown","term":"Portfolio Drawdown","category":"metrics","definition":"Portfolio drawdown is the drawdown measured on the aggregate equity curve of multiple trading strategies running concurrently on the same account, rather than on any individual strategy in isolation. The relationship between portfolio drawdown and individual-strategy drawdowns depends on correlation: highly correlated strategies produce portfolio drawdown close to the average of individuals (no diversification benefit), uncorrelated strategies produce portfolio drawdown materially smaller than the worst individual drawdown (genuine diversification), and anti-correlated strategies can produce portfolio drawdown smaller than the smallest individual drawdown (rare in practice). Building portfolios for low aggregated drawdown is the central craft of multi-EA operation — and the reason serious automated traders run portfolios rather than single 'best EA' deployments. Measurement requires running each strategy on a sub-account or with magic-number isolation so individual contributions can be decomposed from the aggregate.","relatedTerms":["correlation-ea-portfolio","peak-drawdown","drawdown"],"relatedGuide":null},{"slug":"pnl-volatility","term":"PnL Volatility","category":"metrics","definition":"PnL volatility is the standard deviation of periodic returns on the strategy's equity curve, where the period is conventionally daily, weekly, or monthly depending on trade frequency. It quantifies the smoothness of performance: a strategy producing 20% annual return with 8% PnL volatility is dramatically more attractive than the same return with 25% volatility, because the former compounds more reliably and is psychologically much easier to hold through drawdown periods. PnL volatility is the denominator of the Sharpe ratio (return divided by volatility), the dominant industry metric for risk-adjusted performance. Strategy choices that reduce volatility for similar return — better regime filtering, smaller position sizing, diversification across uncorrelated strategies — almost always increase realised long-run compounding even when they reduce headline return. Volatility minimisation is a more durable goal than return maximisation for retail algorithmic trading.","relatedTerms":["sharpe-ratio","drawdown","rolling-profit-factor"],"relatedGuide":null},{"slug":"realised-expectancy","term":"Realised Expectancy","category":"metrics","definition":"Realised expectancy is the empirical per-trade expectancy figure — (win rate × average win) - ((1 - win rate) × average loss) — computed from the live trades a strategy has actually produced, as distinct from the theoretical expectancy estimated by backtest or walk-forward analysis. The gap between backtest expectancy and realised expectancy is the diagnostic measure of how much edge has been lost in transit from simulation to live deployment, with causes typically including spread widening, slippage on stops, execution latency, broker B-book interference, and regime shift since the backtest window closed. A strategy with backtest expectancy of $30/trade and realised expectancy of $5/trade is technically still profitable but has clearly lost most of its assumed edge — usually a sign that conditions have moved or that the backtest contained optimistic assumptions. Realised expectancy needs a sample of at least 100 live trades to be meaningfully estimated.","relatedTerms":["expectancy","sample-size","modelling-quality"],"relatedGuide":null},{"slug":"trade-journal","term":"Trade Journal","category":"risk","definition":"A trade journal is a written record kept by the trader that captures the reasoning behind each trading decision, the resulting outcome, and the lessons learned. For manual traders it covers individual trades — entry rationale, stop placement, execution observations, exit decisions, psychological notes. For algorithmic traders the relevant granularity is one level up: strategy deployments, parameter changes, capital-scaling actions, EA retirement decisions, observations about live-vs-backtest divergence. In both cases the discipline is the same: writing down the reasoning before the outcome is known forces clarity and creates an honest record for later review. Without a journal, the trader's memory of past decisions is reconstructed in hindsight to flatter the present, and the same mistakes are repeated indefinitely. The journal is the substrate that converts experience into learning; without it experience is just elapsed time.","relatedTerms":["trading-plan","review-cadence","pre-commitment"],"relatedGuide":null},{"slug":"trading-plan","term":"Trading Plan","category":"risk","definition":"A trading plan is the written operating manual for the trader's algorithmic-trading practice. A minimum-viable plan specifies: which strategies (EAs) are deployed, on which symbols and brokers, with what capital allocation each; the risk caps applied at strategy level (risk per trade, daily loss limit, max drawdown cutoff) and at portfolio level (total open exposure, correlation limits); the operational cadence (when reviews happen, how often statements are reconciled, what triggers an unscheduled review); the criteria for adding new strategies, scaling existing ones, and retiring underperformers; and the explicit exclusions (what the trader will not do under any circumstances). The plan exists to make decisions mechanical rather than discretionary, which is the only way to operate algorithmic strategies without succumbing to the emotional pulls that destroy retail accounts. Plans must be written, not held mentally, because written plans resist drift.","relatedTerms":["pre-commitment","trade-journal","review-cadence"],"relatedGuide":null},{"slug":"trading-session","term":"Trading Session","category":"basics","definition":"A trading session is a regional market window during which trading activity and liquidity concentrate for instruments associated with that region. Forex operates 24/5 across the major sessions: Asian session (Tokyo, roughly 23:00-08:00 UTC) is associated with JPY, AUD, and NZD pair activity; European session (London, 07:00-16:00 UTC) is the highest-volume session overall and drives most EUR and GBP pair movement; US session (New York, 12:00-21:00 UTC) overlaps with London for several hours, producing the highest-liquidity window of the day, and is followed by a thinner US-only afternoon. Session timing affects spread (typically tightest during overlaps, widest during regional handoffs), volatility (highest during overlaps and news), and execution quality. EAs frequently include session filters that disable trading outside the session their strategy was developed for; running a London-session strategy through Asian hours typically destroys its edge.","relatedTerms":["news-filter","volatility","liquidity"],"relatedGuide":null},{"slug":"ask","term":"Ask","category":"basics","definition":"The ask, also called the ask price or offer, is the lowest price at which a seller in the market is willing to part with a currency pair. From the retail trader's perspective it is the price at which a buy order will be executed for the base currency. The ask is always higher than the bid; the difference between them is the spread, which represents the broker's revenue plus market-maker compensation. Quoted as the right-hand of the two numbers in a typical platform display (e.g. EUR/USD 1.0850 / 1.0852, where 1.0852 is the ask). For market orders, the ask is the relevant price for buy entries and for sell exits. See the canonical entry: ask-price.","relatedTerms":["ask-price","bid","spread"],"relatedGuide":null},{"slug":"bid","term":"Bid","category":"basics","definition":"The bid, also called the bid price, is the highest price at which a buyer in the market is willing to acquire a currency pair. From the retail trader's perspective it is the price at which a sell order will be executed for the base currency. The bid is always lower than the ask; the difference between them is the spread. Quoted as the left-hand of the two numbers in a typical platform display (e.g. EUR/USD 1.0850 / 1.0852, where 1.0850 is the bid). For market orders, the bid is the relevant price for sell entries and for buy exits. Most charting engines plot the bid as the canonical price line; some platforms offer an option to overlay the ask. See the canonical entry: bid-price.","relatedTerms":["bid-price","ask","spread"],"relatedGuide":null},{"slug":"lot","term":"Lot","category":"basics","definition":"A lot, in retail forex, is the standardised unit of transaction size. The four conventional sizes are: standard lot (100,000 units of the base currency), mini lot (10,000 units, 0.1 standard), micro lot (1,000 units, 0.01 standard), and nano lot (100 units, 0.001 standard, supported by fewer brokers). MetaTrader's lot input accepts fractional values and is the principal sizing knob for retail EAs. Lot size directly determines pip value: one standard lot of EUR/USD produces roughly $10 per pip of price movement, one mini lot $1, one micro lot $0.10. Risk-per-trade sizing is calculated by dividing the desired dollar risk by stop-loss distance times pip value per lot. See the canonical entry: lot-size.","relatedTerms":["lot-size","pip","position-sizing"],"relatedGuide":null},{"slug":"expert-advisor-ea","term":"Expert Advisor (EA)","category":"platform","definition":"An Expert Advisor (EA) is an algorithmic-trading program written in MQL4 or MQL5 that runs inside the MetaTrader 4 or MetaTrader 5 client and is authorised to open, manage, and close trades on the attached account without human intervention. The EA encodes a trading strategy as code — signal layer (when to enter), filter layer (when not to enter), risk layer (how much to size), execution layer (how to submit and manage orders). MetaTrader provides the runtime, the live price feed, and the broker connection; the EA provides the decision logic. EAs are distributed as compiled .ex5 / .ex4 binaries, traded through the MQL5 Marketplace or sold by independent vendors, and customised through input parameters and configuration .set files. See the canonical entry: expert-advisor.","relatedTerms":["expert-advisor","mt5-metatrader-5","mql"],"relatedGuide":null},{"slug":"pip-value","term":"Pip Value","category":"basics","definition":"Pip value is the cash amount, expressed in the account-deposit currency, corresponding to a one-pip price movement on a position of a given size. For accounts denominated in USD trading dollar-quote pairs like EUR/USD, the calculation is direct: one standard lot of EUR/USD has a pip value of approximately $10 (because 100,000 units × 0.0001 USD per unit = $10). For non-dollar pairs (e.g. GBP/JPY on a USD account) the value depends on the current exchange rate of the quote currency to USD and must be recalculated dynamically. Pip value is the central conversion factor for risk-per-trade sizing: desired dollar risk divided by stop-loss distance in pips, divided again by pip value per lot, equals the lot size to trade. MetaTrader and most broker platforms display pip value in the symbol's specifications.","relatedTerms":["pip","lot-size","position-sizing"],"relatedGuide":null},{"slug":"vps-virtual-private-server","term":"VPS (Virtual Private Server)","category":"tools","definition":"A Virtual Private Server (VPS) is a remote computing instance — typically running Windows Server with MetaTrader installed — rented from a hosting provider on a monthly basis. For algorithmic forex traders the VPS solves the always-on requirement: an EA running on a home computer is interrupted whenever the computer is shut down, sleeps, loses internet, or crashes; an EA running on a VPS continues operating uninterrupted as long as the hosting provider's data centre is online. Forex-specialised VPS providers (BeeksFX, NewYorkCity Servers, ForexVPS, broker-bundled options like MQL5 VPS) offer servers proximity-located near major broker data centres for low latency. For active EA operation a VPS is essentially required; for swing or position trading it is helpful but not strictly necessary. See the canonical entry: vps.","relatedTerms":["vps","persistent-vps","co-location"],"relatedGuide":null},{"slug":"swap-rollover","term":"Swap (rollover)","category":"basics","definition":"Swap, also called the rollover or overnight rollover, is the daily interest-rate adjustment applied to forex positions held past the broker's rollover cutoff — conventionally 22:00 GMT, with a triple-swap charge on Wednesdays to settle weekend financing. The swap reflects the interest-rate differential between the two currencies in the pair: holding a long position in a higher-yielding currency against a lower-yielding one produces a positive swap (credit), and vice versa. Broker markups on the underlying interbank rate make most retail swaps net-negative even on theoretically positive-carry positions. Swap matters most for position trading and carry-trade strategies that hold trades for days or weeks; for day-traders and scalpers who close intra-day it is irrelevant. The broker's specification panel for each symbol displays current long-swap and short-swap rates. See the canonical entry: swap.","relatedTerms":["swap","carry-trade","trading-session"],"relatedGuide":null},{"slug":"stop-loss-sl","term":"Stop-Loss (SL)","category":"order","definition":"A stop-loss, abbreviated SL, is a pending exit order attached to an open position that closes the position if the market moves adversely by a defined distance. For a long position the stop-loss is placed below the entry price; for a short position, above. The stop-loss serves two essential functions: it caps the maximum loss on the trade to a known figure, enabling deterministic risk-per-trade sizing; and it forces the loss to be realised mechanically rather than left open to compound through emotional indecision. Every position should have a stop-loss in place before the position is opened, not after; trades without stops have unbounded downside and convert from risk-management into hope. See the canonical entry: stop-loss.","relatedTerms":["stop-loss","take-profit-tp","risk-parameter"],"relatedGuide":null},{"slug":"take-profit-tp","term":"Take-Profit (TP)","category":"order","definition":"A take-profit, abbreviated TP, is a pending exit order attached to an open position that closes the position when the market reaches a chosen profit target. For a long position the take-profit is placed above the entry price; for a short position, below. Paired with a stop-loss on the opposite side, the two orders form the position's risk-reward bracket: the loss is bounded at the stop-loss distance, the profit is bounded at the take-profit distance, and the ratio of these distances is the trade's risk-reward ratio. The take-profit converts profitable runs into realised gains mechanically rather than relying on the trader's discretion. Some strategies prefer to let profits run with only a trailing-stop exit; both approaches are valid, with the choice depending on the strategy's edge profile. See the canonical entry: take-profit.","relatedTerms":["take-profit","stop-loss-sl","risk-reward-ratio"],"relatedGuide":null},{"slug":"overfitting","term":"Overfitting","category":"ai-ml","definition":"Overfitting is the canonical failure mode of model fitting in any domain: the model is tuned so finely to specific training data that it captures the data's random noise as well as its systematic structure, producing excellent in-sample fit and poor out-of-sample generalisation. In algorithmic trading this is called curve-fitting and is the dominant reason retail strategies that look attractive in backtest fail in live deployment. Defences are parameter parsimony (fewer tunable inputs reduce fitting capacity), regularisation (parameter values pulled toward zero or toward priors), out-of-sample validation (held-out windows), walk-forward analysis (rolling-window validation), robustness checks (perturbing optimal parameters by 10-20% and confirming graceful degradation), and disciplined scepticism toward any backtest result that looks 'too clean'. See the canonical entry: overfitting-curve-fitting.","relatedTerms":["overfitting-curve-fitting","curve-fitting","walk-forward-analysis"],"relatedGuide":null},{"slug":"hedging-account","term":"Hedging Account","category":"platform","definition":"A hedging account is a broker-side account type that permits the trader to maintain multiple positions on the same symbol simultaneously, including positions on opposing sides. A hedging account allows the trader to be long one lot of EUR/USD and short one lot of EUR/USD at the same time, each tracked as a distinct position with its own entry price, stop-loss, take-profit, and P&L. This contrasts with a netting account, where simultaneous orders on the same symbol collapse into a single net exposure (long one and short one would close to zero open exposure). Hedging accounts are the default for most MT4 brokers and many MT5 brokers serving retail. They are essential for multi-EA portfolios where different strategies may produce opposing signals on the same instrument. US-regulated brokers are prohibited from offering hedging accounts under NFA rules; non-US jurisdictions typically permit them.","relatedTerms":["hedging","netting-account","metatrader-5"],"relatedGuide":null},{"slug":"support","term":"Support","category":"analysis","definition":"A support level is a price area where a previous downward move was arrested by buying interest, producing a halt or reversal. The level becomes psychologically and operationally significant because traders observe the prior bounce and place buy orders or stop-loss orders near it, reinforcing the response in subsequent visits. Support is identified visually as a horizontal level where price has bounced multiple times, often coinciding with round numbers, prior swing lows, or technical indicators (moving averages, Fibonacci levels). The defining characteristic is asymmetric reaction: price approaches support and decelerates or reverses. When support breaks, it frequently becomes resistance — the level where prior buyers now sell into rallies to recover their losses. Support-and-resistance is the foundational structure of technical analysis. See the canonical combined entry: support-resistance.","relatedTerms":["support-resistance","resistance","trend-line"],"relatedGuide":null},{"slug":"resistance","term":"Resistance","category":"analysis","definition":"A resistance level is a price area where a previous upward move was arrested by selling interest, producing a halt or reversal. The level becomes operationally significant because traders observe the prior rejection and place sell orders or stop-loss orders near it, reinforcing the response on subsequent tests. Resistance is identified visually as a horizontal level where price has been rejected multiple times, often coinciding with prior swing highs, round numbers, or technical indicators (Bollinger upper bands, moving averages, Fibonacci extensions). The defining characteristic is asymmetric reaction: price approaches resistance and decelerates or reverses. When resistance breaks decisively, it frequently becomes support — the level where former sellers now buy on retests. Resistance and support together provide the geometric structure within which directional bias, breakout strategies, and mean-reversion strategies all operate. See the canonical combined entry: support-resistance.","relatedTerms":["support-resistance","support","trend-line"],"relatedGuide":null},{"slug":"maximum-drawdown","term":"Maximum Drawdown","category":"metrics","definition":"Maximum drawdown, also referred to as peak drawdown or simply max DD, is the largest percentage decline from any equity high-water mark to the subsequent low within a measurement window — backtest, forward-test, or live operation. It is conventionally treated as the single most important risk metric because it represents the worst experience a trader using the strategy actually endured, which is the experience that determines whether the trader stays with the strategy through to its eventual recovery. Recovery from a drawdown is asymmetric: a 50% drawdown requires a 100% gain to recover, a 30% requires 43%, a 20% requires 25%. Strategies with attractive returns and high maximum drawdowns are often less compoundable in practice than strategies with moderate returns and modest drawdowns, because the psychological tolerance test is harder to pass. See the canonical entry: drawdown.","relatedTerms":["drawdown","peak-drawdown","recovery-factor"],"relatedGuide":null},{"slug":"regulator","term":"Regulator","category":"compliance","definition":"A regulator is the government or quasi-government authority that licences, supervises, and disciplines financial-services firms operating within its jurisdiction. The dominant regulators for retail forex are tier-1 (UK FCA, US SEC/CFTC/NFA, Australia ASIC, Japan FSA, Singapore MAS, Germany BaFin) and tier-2 (Cyprus CySEC, Malta MFSA, South Africa FSCA, Mauritius FSC). Each enforces capital requirements, client-money-segregation rules, conduct standards, marketing restrictions, and complaint-handling obligations. Regulatory tier is the single largest determinant of broker trustworthiness — a tier-1 regulated broker faces meaningful consequences for misconduct, an unregulated offshore broker faces none. Traders should confirm regulatory status on the regulator's official register (not just the broker's website claim), check for warning notices, and prefer brokers under regimes with consumer-protection guarantees like the FCA's FSCS compensation scheme.","relatedTerms":["regulator-complaint","appointed-representative-ar","custody-in-trading"],"relatedGuide":null},{"slug":"economic-calendar","term":"Economic Calendar","category":"analysis","definition":"An economic calendar is a structured published schedule of upcoming macroeconomic releases (CPI, GDP, employment), central bank meetings (Fed FOMC, ECB, BoE), speeches, surveys, and other scheduled events that produce predictable spikes in market volatility. Each event is tagged with impact level (red/orange/yellow or 1-3 stars), expected vs prior value, and time of release in the trader's timezone. The canonical retail sources are ForexFactory, Investing.com, and FXStreet; institutional traders use Bloomberg, Reuters, and broker-supplied API feeds. Two main uses: news-trading strategies that deliberately trade announcement spikes, and news-filter logic that disables ordinary strategies around high-impact events to avoid spread blow-out and slippage on stops. EAs typically consume an economic calendar via a paid API or a parsed weekly download to drive their news-filter component.","relatedTerms":["news-filter","news-trading","trading-session"],"relatedGuide":null},{"slug":"netting-account","term":"Netting Account","category":"platform","definition":"A netting account is a broker-side account configuration where positions on the same symbol are aggregated into a single net exposure rather than tracked as independent buy and sell tickets. Submitting a sell order while long produces a partial or complete close of the long position, not a separate short position. Netting is the convention in futures markets, is mandated for US-regulated retail forex by the NFA, and is offered as a non-default option by MT5 brokers in other jurisdictions. The alternative — hedging — permits simultaneous long and short positions on the same symbol with independent tracking. Netting accounts are simpler conceptually but constrain multi-EA strategies where different EAs may signal in opposite directions on the same instrument; for such portfolios, hedging accounts are typically preferred. The account type is set at broker registration and is rarely changeable afterward.","relatedTerms":["hedging-account","metatrader-5","broker"],"relatedGuide":null},{"slug":"ohlc","term":"OHLC","category":"analysis","definition":"OHLC stands for Open, High, Low, Close — the four price values that summarise the price action of a single bar at a chosen timeframe. The open is the price at bar start, the close is the price at bar end, and high and low are the maximum and minimum prices observed within the bar window. The four values together collapse an arbitrary number of intra-bar ticks into a compact, comparable representation, and form the data type that drives most chart visualisations (candlestick, bar charts) and most backtest engines. Strategies that depend on intra-bar dynamics (scalpers, exact stop fills, spread-spike behaviour) require richer data than OHLC alone provides — typically true-tick data with full bid/ask sequencing. For H1+ timeframe strategies that act on bar closes, OHLC is generally adequate and is the universal interchange format for historical forex data.","relatedTerms":["candlestick","timeframe","tick-data-quality"],"relatedGuide":null},{"slug":"tick","term":"Tick","category":"basics","definition":"A tick is a single update to the price quote of a traded instrument — a change in the bid, the ask, or both, with no specification of trade size. Tick frequency varies dramatically by instrument and time of day: exotic forex pairs may tick a few times per second during quiet sessions, while EUR/USD during the London-New York overlap may tick hundreds of times per second. Tick-level data — capturing every bid and ask update — is the highest fidelity data form available and is essential for credible backtests of execution-sensitive strategies. The MetaTrader 5 Strategy Tester supports 'every tick based on real ticks' mode when tick data is provided by the broker; this mode reports 99% modelling quality and is the standard for honest scalper or news-trader validation. Sub-tick data (order book depth, time-and-sales) is available institutionally but rare at retail scale.","relatedTerms":["tick-data-quality","modelling-quality","ohlc"],"relatedGuide":null},{"slug":"pamm","term":"PAMM","category":"strategy","definition":"PAMM stands for Percentage Allocation Money Management, a broker-side account structure that allows a master trader to execute trades on a pooled balance from multiple investor accounts, with profits and losses allocated proportionally to each investor's contribution. The legitimate version is regulated as portfolio management or collective investment, requires identified licensed brokers, KYC on both master and investors, periodic audited reporting, and disclosed fee structures (typically performance fees of 20-30% on gains above a high-water mark). The unregulated version — marketed through Telegram, signal services, and 'mentor' networks — uses the PAMM label to dress up what is structurally a managed-trading scam: master takes custody, fabricates performance, exits with deposits. The discriminator is regulatory status of the broker offering the PAMM structure and identifiability of the master trader. Forsage ($300M, SEC 2022) is a documented PAMM-styled retail fraud case.","relatedTerms":["managed-trading-offer","prop-firm-funded-account","ponzi-scheme"],"relatedGuide":null},{"slug":"prop-firm-funded-account","term":"Prop Firm (funded account)","category":"strategy","definition":"A proprietary trading firm — 'prop firm' in retail vernacular — offers funded trading accounts to retail traders who pay a one-time fee to enter and pass a challenge demonstrating disciplined trading on a demo account. Successful candidates are given access to a 'funded' account (in legitimate firms, real capital; in some marginal firms, a simulated account where the firm pays out claimed profits from new challenge fees) and split profits with the firm at ratios commonly 70/30 to 90/10 in the trader's favour. The largest legitimate operators (FTMO, Topstep, MyForexFunds, The Funded Trader, FundedNext) provide a real capital-access path for skilled traders without personal capital. Sustainability of the model depends on the funded trader pass rate — most challenges fail, providing the fee revenue that funds the few who succeed. Some copycat operators run challenges that are statistically impossible to pass; due diligence on success-rate transparency is essential before paying entry fees.","relatedTerms":["pamm","managed-trading-offer","trading-plan"],"relatedGuide":null},{"slug":"persistent-vps","term":"Persistent VPS","category":"tools","definition":"A persistent VPS is the always-on hosting requirement specifically suited to live EA operation — a virtual private server that operates continuously across long periods without unscheduled restart, sleep, network interruption, or resource throttling. The distinction matters because EA trading is intolerant of even short interruptions: a power-cycle during an open position can leave the EA out of sync with broker-side state, a missed tick during a fast move can miss an entry or stop, a maintenance reboot during news can produce orphan positions. Forex-specialised VPS providers (BeeksFX, NewYorkCity Servers, ForexVPS, broker-bundled VPS through MQL5 or specific brokers) guarantee persistent operation through redundant infrastructure and SLAs measured in uptime fractions (99.9% or higher). Cheap general-purpose VPS deals from major cloud providers technically work but lack the operational guarantees serious EA traders need; the few dollars saved per month are quickly outweighed by a single misbehaviour event.","relatedTerms":["vps","vps-virtual-private-server","co-location"],"relatedGuide":null},{"slug":"versioned-changelog","term":"Versioned Changelog","category":"files-config","definition":"A versioned changelog is a documented, ordered history of the changes made to an EA across its release history — version numbers (e.g. 1.0.0 → 1.0.1 → 1.1.0), release dates, and a per-version summary of bug fixes, feature additions, parameter changes, and behavioural modifications. For commercial vendors a changelog is part of basic accountability: buyers need to know what they are upgrading into, whether a new version preserves the parameter set of the previous one, and whether breaking changes require revalidation of their forward-tested configurations. For self-developed EAs the changelog is part of operational discipline: when a strategy is running multiple instances across brokers and accounts, knowing precisely which version is deployed where is essential for diagnosing performance differences. Conventions like Semantic Versioning (MAJOR.MINOR.PATCH) communicate the scale of change at a glance. EAs distributed without a changelog are a vendor-transparency red flag.","relatedTerms":["vendor-transparency","default-value","set-file"],"relatedGuide":null},{"slug":"withdrawal-test","term":"Withdrawal Test","category":"compliance","definition":"A withdrawal test is the practice of requesting a small withdrawal from a forex broker shortly after the initial deposit and before scaling trading capital. The test confirms two things: that the broker's withdrawal infrastructure functions as advertised, and that the broker's verification, KYC, and back-office processes operate correctly. A withdrawal that succeeds — funds arrive within the broker's promised window, no manufactured friction emerges — provides material evidence that the broker is not running a withdrawal trap. A withdrawal that fails, stalls, or attracts new unexpected requirements is a clear signal to withdraw the remainder and avoid further deposits. The test costs the buyer minor inconvenience and a small transfer fee; it can save the entire deposit. It should be performed at every new broker before any significant capital is deposited, regardless of the broker's marketing or regulatory claims.","relatedTerms":["withdrawal-trap","chargeback","vendor-transparency"],"relatedGuide":null},{"slug":"myfxbook","term":"MyFXBook","category":"execution","definition":"MyFXBook is a third-party verification service that traders use to publish their broker account performance with cryptographic auditability. The account is linked via an MT4/MT5 investor password — a read-only credential — so MyFXBook can fetch every closed trade, equity-curve point, and drawdown event without ability to modify positions. The result is a public dashboard with verifiable status badges. The editorial bar for EA evaluation is a MyFXBook account continuously tracked for 6-12+ months with trade-by-trade detail matching vendor marketing claims.","relatedTerms":["mql5-signals","live-track-record","verified-live-track","trade-by-trade-verification","verification-length"],"relatedGuide":null},{"slug":"mql5-signals","term":"MQL5 Signals","category":"execution","definition":"MQL5 Signals is the signal-following service operated by MetaQuotes inside the MetaTrader ecosystem. Providers publish a live broker account; subscribers can mirror trades onto their own MetaTrader installation. The economic mechanism (paid monthly subscription, automatic trade copying) is secondary for EA buyers — what matters is that every MQL5 Signal carries verified trade-by-trade history. Vendor-published MQL5 Signal links carry the same verification weight as MyFXBook accounts for EA evaluation purposes.","relatedTerms":["myfxbook","live-track-record","verified-live-track","trade-by-trade-verification"],"relatedGuide":null},{"slug":"live-track-record","term":"Live Track Record","category":"metrics","definition":"A live track record is the public, verifiable trading-account history that demonstrates an EA's behaviour under real broker conditions. The record must be third-party verifiable (MyFXBook, MQL5 Signals, FX Blue, or equivalent) with trade-by-trade detail; vendor-published screenshots, marketing testimonials, and \"audited\" claims without a verifiable account link do not qualify. The 2026 editorial floor is 6 months continuous history for inclusion in any evaluation tier; 12+ months for top-tier consideration; through at least one regime stress period for the strongest placements.","relatedTerms":["myfxbook","mql5-signals","verified-live-track","trade-by-trade-verification","verification-length"],"relatedGuide":null},{"slug":"verified-live-track","term":"Verified Live Track","category":"metrics","definition":"A verified live track is a stricter quality bar than a live track record: continuous third-party verification across the entirety of the published period, with no off-platform gaps. The distinction matters because some vendors maintain a verified account, disable verification briefly, hand-trade or override the EA during that window, then re-enable verification — producing a record that contains both verified and unverified periods. A verified live track excludes those mixed periods, providing higher trust at the cost of stricter inclusion requirements.","relatedTerms":["live-track-record","myfxbook","mql5-signals","trade-by-trade-verification","verification-length"],"relatedGuide":null},{"slug":"trade-by-trade-verification","term":"Trade-by-Trade Verification","category":"metrics","definition":"Trade-by-trade verification is the level of detail at which a live track record exposes individual trades rather than aggregate statistics. MyFXBook, MQL5 Signals, and FX Blue all provide this granularity by default — every closed trade appears as a separate row with entry time, exit time, instrument, position size, entry price, exit price, and realised profit. The detail enables buyer-side auditing: comparing trade timestamps against the EA's claimed signal-generation logic, instrument coverage against vendor marketing, and entry/exit patterns against the strategy class the EA claims to implement.","relatedTerms":["live-track-record","verified-live-track","myfxbook","mql5-signals"],"relatedGuide":null},{"slug":"kill-switch","term":"Kill Switch","category":"risk","definition":"A kill switch is a hard architectural risk control in an EA's logic: when intraday account loss exceeds a configured threshold (typically 2-5% of equity), the EA stops opening new positions for the remainder of the trading day. Existing positions are managed normally per their stop-loss and take-profit logic. The mechanism bounds tail risk at the calendar boundary — regardless of what the strategy logic would suggest doing, the EA cannot continue compounding losses past the daily threshold. This is the architectural feature that distinguishes safety-tier EAs from configurable systems where the trader is expected to enforce daily-loss discipline manually.","relatedTerms":["risk-per-trade","drawdown","ea-retirement-criteria"],"relatedGuide":null},{"slug":"risk-per-trade","term":"Risk Per Trade","category":"risk","definition":"Risk per trade is the percentage of account equity at risk on any single trade, computed as (stop-loss distance in pips × pip value × lots) ÷ account equity. Vendor-default settings on serious EAs are 0.5-1.5% for safety-tier, 1-2% for most-profitable-tier, 2-3% for aggressive systems. The parameter directly determines the relationship between win-rate and drawdown — at any given win-rate, drawdown scales linearly with risk-per-trade while returns scale at a similar rate, so altering this parameter changes the strategy's risk profile materially.","relatedTerms":["kill-switch","drawdown","calmar-ratio","lot-size"],"relatedGuide":null},{"slug":"underwater-curve","term":"Underwater Curve","category":"metrics","definition":"The underwater curve is a chart that plots equity as a percentage below its running maximum over time. Y-axis values are always 0 or negative — at new equity highs the curve touches 0, between highs it stays negative at the depth of the current drawdown. The chart visualises three things the headline \"max drawdown 14%\" number hides: how often the strategy is in drawdown vs at new highs, how long underwater periods last, and how frequently the strategy revisits comparable depths. A flat 0 curve with one deep spike represents different risk than a curve oscillating around -8% for months at a time.","relatedTerms":["drawdown","calmar-ratio","recovery-factor"],"relatedGuide":null},{"slug":"survivorship-bias","term":"Survivorship Bias","category":"metrics","definition":"Survivorship bias in EA evaluation occurs because the buyer's information set is skewed toward survivors. The 2026 MQL5 marketplace lists ~16,000 EAs; this is a fraction of the EAs that were listed in 2022-2024 and have since been withdrawn after vendor abandonment, performance collapse, or marketplace removal. The historical average return of the EAs currently visible overstates the historical average return of the universe that existed at any past point — because the failures were quietly removed. Buyers evaluating current EAs against past performance must account for this bias.","relatedTerms":["vendor-abandonment","live-track-record","verification-length"],"relatedGuide":null},{"slug":"correlation-cap","term":"Correlation Cap","category":"risk","definition":"A correlation cap is a position-sizing constraint applied across multiple instruments traded by the same EA. The cap limits the total directional exposure when correlated pairs (e.g. all the USD-quote majors) align — instead of opening full position sizes on 7 USD-quote pairs simultaneously during a dollar-strength regime (which would mean 7× single-pair exposure), the cap reduces per-pair sizing or rejects later signals so the aggregate exposure stays bounded. The mechanism prevents the most common multi-pair failure mode where apparent diversification dissolves into concentrated single-factor exposure during regime stress.","relatedTerms":["risk-per-trade","kill-switch","drawdown"],"relatedGuide":null},{"slug":"tier-1-ecn","term":"Tier-1 ECN","category":"execution","definition":"Tier-1 ECN is the editorial label for forex brokers offering Electronic Communication Network execution at the highest market-quality tier. The defining features: raw inter-bank pricing with sub-0.5 pip spreads on major pairs; a fixed commission per round-turn lot ($3-3.50 per lot is standard); direct routing to a panel of liquidity providers without dealing-desk intervention; sub-15ms execution latency from London (LD4) or New York (NY4) colocation. The 2026 editorial shortlist for Tier-1 ECN is short: IC Markets Raw, Pepperstone Razor, Tickmill Pro.","relatedTerms":["ecn-broker","spread","slippage","latency","ld4-colocation"],"relatedGuide":null},{"slug":"ecn-broker","term":"ECN Broker","category":"execution","definition":"An ECN (Electronic Communication Network) broker routes client orders directly to a panel of liquidity providers — typically large banks and prime-brokers — via FIX-protocol connections, rather than executing trades against its own book. Revenue model: a fixed commission per round-turn lot ($3-7 per lot is the retail range), with spreads passed through from the underlying liquidity providers. The ECN architecture eliminates the conflict-of-interest inherent in market-maker models where the broker profits when the trader loses.","relatedTerms":["tier-1-ecn","spread","slippage"],"relatedGuide":null},{"slug":"swap-rate","term":"Swap Rate","category":"execution","definition":"The swap rate (also called rollover, overnight financing, or carry) is the interest charge or credit applied to forex positions that remain open past the broker's daily rollover time (typically 17:00 New York time). The rate reflects the interest-rate differential between the two currencies in the pair: long the higher-yielding currency receives a credit, short the higher-yielding currency pays a charge. Broker mark-up sits on top of the underlying inter-bank rates. Wednesday-to-Thursday positions accrue triple swap to cover Saturday and Sunday weekend rollover.","relatedTerms":["spread","tier-1-ecn","ecn-broker"],"relatedGuide":null},{"slug":"tick-data","term":"Tick Data","category":"execution","definition":"Tick data is the most granular form of forex price data, recording every bid/ask price change with millisecond or sub-second timestamps. Each tick represents one quote update from a liquidity provider — in normal market conditions, EURUSD generates 5,000-20,000 ticks per day. Bar data (M1, M5, H1, etc.) aggregates these ticks into OHLC (open-high-low-close) summaries; tick data preserves the underlying micro-structure that bar aggregation throws away. For scalping EAs operating at sub-bar timescales, tick data is the only valid backtest input.","relatedTerms":["backtest","slippage","spread"],"relatedGuide":null},{"slug":"prop-firm","term":"Prop Firm","category":"execution","definition":"A proprietary trading firm (prop firm) funds traders' accounts with the firm's own capital and shares the resulting profits with the trader. The 2022-2025 retail prop-firm market grew explosively around the \"funded trader\" model: the trader pays $50-1,000 for an evaluation challenge (typically achieving 8-10% profit within strict drawdown rules over 30-60 days), and upon passing receives a funded $10,000-$200,000 account with a 70-90% profit split. Prop-firm structures have different rules, execution venues, and risk constraints than retail accounts, which materially affects EA deployment.","relatedTerms":["kill-switch","drawdown","risk-per-trade"],"relatedGuide":null},{"slug":"multi-strategy","term":"Multi-Strategy EA","category":"strategy","definition":"Multi-strategy EAs combine multiple independent strategy modules into a single deployment unit, allocating risk across them rather than concentrating in one strategy class. Typical architectures: a trend module for trending regimes, a mean-reversion module for range regimes, a session-breakout module for volatility-expansion regimes, and a news-event filter that disables trading during high-impact macro releases. Per-module risk allocation (e.g. 30% trend, 30% mean-reversion, 25% breakout, 15% session-specific) lets the portfolio earn in different regimes than any single module would handle. The architecture is the 2026 editorial default for professional-tier products serving high-capital allocations.","relatedTerms":["strategy-class","strategy-attribution","correlation-cap","risk-per-trade"],"relatedGuide":null},{"slug":"strategy-class","term":"Strategy Class","category":"strategy","definition":"Strategy class is the architectural category that defines how an EA generates trade signals and manages position lifecycle. Common classes: trend-following (rides directional momentum, low win-rate + high R-multiple wins), mean-reversion (trades counter-direction to range extremes, high win-rate + small wins), breakout (captures volatility expansion at session opens or consolidation breaks), scalping (M1-M5 high-frequency, very tight risk per trade), grid (places multiple staggered orders, dangerous tail risk), martingale (doubles position after losses, structurally inevitable blow-up), arbitrage (exploits price discrepancies, rare in modern retail forex), news-based (trades around macro releases, sensitive to broker news-filter rules).","relatedTerms":["multi-strategy","scalping"],"relatedGuide":null},{"slug":"strategy-attribution","term":"Strategy Attribution","category":"metrics","definition":"Strategy attribution is the verification-platform feature that decomposes a multi-strategy EA's aggregate live performance into per-module contributions. Instead of seeing only \"this account returned 6% last month\", buyers see \"trend module +3.2%, mean-reversion +1.8%, breakout +1.0%, news-filter 0%\". The attribution lets buyers understand which strategies are producing the edge, which are decaying, and whether the aggregate performance is concentrated in one module (single-strategy disguised as multi-strategy) or genuinely diversified across the architecture.","relatedTerms":["multi-strategy","strategy-class","verification-length","live-track-record"],"relatedGuide":null},{"slug":"verification-length","term":"Verification Length","category":"metrics","definition":"Verification length is how long the EA's public live track record (MyFXBook, MQL5 Signals, FX Blue) has been continuously tracked. The metric is consequential because verification quality compounds with time — a 14-month track record includes more regime variance than a 4-month one, providing materially stronger evidence of edge robustness. Verification length cannot be backfilled or accelerated; vendors with short records are simply newer in their public-verification cycle regardless of internal track record they claim privately.","relatedTerms":["live-track-record","verified-live-track","myfxbook","mql5-signals"],"relatedGuide":null},{"slug":"realised-return","term":"Realised Return","category":"metrics","definition":"Realised return is the actual percentage return an EA produced on its public live trading account during a defined measurement window — typically monthly returns shown on MyFXBook, MQL5 Signals, or equivalent verification platforms. The distinction from related concepts matters: realised return is what actually happened on real capital under real broker conditions; backtest return is the theoretical result of replaying historical data; marketed return is the vendor's published claim. Editorial evaluation prioritises realised return over the alternatives because only realised return reflects the friction of live execution.","relatedTerms":["calmar-ratio","live-track-record","verification-length","drawdown"],"relatedGuide":null},{"slug":"neural-network","term":"Neural Network","category":"ai-ml","definition":"A neural network is a parameterised machine-learning model structured as a series of layers, each containing many \"neurons\" that combine the previous layer's outputs through weighted sums plus a non-linear activation function (sigmoid, ReLU, tanh). Training optimises the weights to minimise prediction error on a labelled dataset. In forex EA applications, neural networks typically serve as either (a) signal filters that score rule-generated trade signals and reject low-quality ones, or (b) full strategy classifiers that predict probability of trade success from a feature vector of price-derived inputs.","relatedTerms":["ml-filter","retraining-cadence","concept-drift"],"relatedGuide":null},{"slug":"ml-filter","term":"ML Filter","category":"ai-ml","definition":"An ML filter is a machine-learning model trained to score rule-generated trade signals and reject those below a probability threshold. The underlying strategy remains rule-based and auditable; the ML layer sits on top, screening for higher-quality opportunities. Compared to end-to-end ML strategies, ML filters keep the rule layer transparent — buyers and vendors can still inspect why the EA generated a candidate signal — while compounding the rule layer's edge with ML-driven precision.","relatedTerms":["neural-network","retraining-cadence","concept-drift"],"relatedGuide":null},{"slug":"retraining-cadence","term":"Retraining Cadence","category":"ai-ml","definition":"Retraining cadence is the schedule on which a ML-augmented EA's underlying model is retrained on new market data. Common cadences: weekly (used by GoldStrike AI's premium tier), monthly (typical for ML filter products), quarterly (lower-engagement vendors). The mechanism: as new market data accumulates, the model is re-fit on a rolling window that drops the oldest period and adds the newest, so the model's parameters track evolving market microstructure rather than being frozen at training-time. Without retraining, ML models decay as the gap between training data and live market widens.","relatedTerms":["neural-network","ml-filter","concept-drift"],"relatedGuide":null},{"slug":"concept-drift","term":"Concept Drift","category":"ai-ml","definition":"Concept drift is the change over time in the underlying distribution of the data a machine-learning model is operating on. In supervised learning terms, both P(X) (the distribution of input features) and P(Y|X) (the relationship between features and target outcomes) can drift. In forex EA applications, concept drift manifests as decaying ML model performance over months — features that once predicted trade success become less predictive as market microstructure shifts, algorithmic competition increases, and macro regimes transition. Concept drift is the architectural reason ML-augmented EAs require periodic retraining.","relatedTerms":["neural-network","ml-filter","retraining-cadence"],"relatedGuide":null},{"slug":"latency","term":"Latency","category":"execution","definition":"Latency in forex EA context is the round-trip time between a price-data event and the corresponding broker confirmation. Components: tick delivery (broker to VPS), EA decision time (typically sub-1ms), order submission (VPS to broker), order matching (broker server), broker confirmation (broker to VPS). Total round-trip latency from premium VPS colocated in LD4 (London) or NY4 (New York) to Tier-1 ECN brokers is 10-15ms; retail home internet to standard VPS routes 60-150ms. For scalping strategies, the latency difference between LD4 colocation and retail VPS converts a profitable strategy into a losing one.","relatedTerms":["ld4-colocation","institutional-vps","tier-1-ecn","slippage"],"relatedGuide":null},{"slug":"ld4-colocation","term":"LD4 Colocation","category":"tools","definition":"LD4 is the Equinix London Docklands datacenter facility — one of the most concentrated forex-trading infrastructure venues globally. Many Tier-1 ECN brokers (IC Markets, Pepperstone, Tickmill) operate their matching engines in LD4; many liquidity providers host their pricing systems there as well. LD4 colocation refers to running a trader's VPS in the same datacenter, which minimises network latency between the VPS and the broker server to typically 1-3ms — versus 60-150ms from typical retail VPS provider locations.","relatedTerms":["latency","tier-1-ecn","institutional-vps"],"relatedGuide":null},{"slug":"institutional-vps","term":"Institutional VPS","category":"tools","definition":"Institutional VPS is the highest-quality VPS infrastructure tier available to retail and institutional EA deployment. Distinguishing features beyond standard VPS: documented uptime service-level agreements (99.95% or higher), SOC2 compliance for institutional regulatory requirements, LD4 or NY4 colocation as standard rather than premium add-on, redundant network connectivity (multiple backbone providers), institutional support tier with documented response times, and SLA financial commitments for downtime. The premium typically costs $75-200/month vs $25-50/month for retail-grade VPS.","relatedTerms":["ld4-colocation","latency"],"relatedGuide":null},{"slug":"vendor-abandonment","term":"Vendor Abandonment","category":"scam-patterns","definition":"Vendor abandonment is the operational failure where an EA's vendor ceases active maintenance — stops responding to support emails, stops publishing updates, allows the verified live account to go stale, and effectively disappears from the public infrastructure that documented the product. The buyer is left with a static product whose edge decays as markets evolve. Common pattern: vendor goes quiet for 6-18 months after initial release, support channel becomes unresponsive, then verified account disappears or stops updating. For ML-based EAs, abandonment essentially means model decay without recovery.","relatedTerms":["vendor-transparency","survivorship-bias","retraining-cadence"],"relatedGuide":null},{"slug":"demo-first-verification","term":"Demo-First Verification","category":"risk","definition":"Demo-first verification is the editorial-mandated practice of deploying every new EA purchase to a demo account before live capital deployment. Duration: 2 weeks minimum, 4 weeks for serious capital commitments, 8 weeks for institutional-tier allocations. The demo period serves three purposes: confirming the EA installs and runs correctly on the buyer's broker setup, observing how the EA's execution differs between vendor's verified broker and buyer's broker, and developing a monitoring routine before real capital is at stake. Skipping this step is the single most common avoidable EA deployment error.","relatedTerms":["live-track-record","myfxbook","mql5-signals","kill-switch"],"relatedGuide":null},{"slug":"repaint","term":"Repaint","category":"analysis","definition":"Repaint is the failure mode where an indicator's historical values change as new bars print, making the indicator appear to predict past events while being unusable in live trading. A repainting indicator might show a clear \"buy\" signal at a past bar in retrospect but at the time that bar was forming, the signal wasn't visible — it only appeared after subsequent bars revealed the trend. Repainting indicators are the largest single source of fraudulent EA marketing because backtests on repainting indicators show impossibly strong results that don't replicate in live deployment.","relatedTerms":["backtest"],"relatedGuide":null},{"slug":"mt4","term":"MT4 (MetaTrader 4)","category":"platform","definition":"MT4 (MetaTrader 4) is the original MetaTrader retail trading platform from MetaQuotes Software Corp. Launched in 2005, MT4 introduced the MQL4 programming language that powers the majority of commercial forex EAs in circulation. Despite MetaQuotes announcing MT4 end-of-life multiple times since 2018, the platform remains dominant because most retail brokers continue supporting it, the EA library is enormous, and the simpler MQL4 language has a lower learning curve than MQL5. Strategies written for MT4 require porting to run on MT5 due to the different programming model.","relatedTerms":["mt5-metatrader-5","mql","mql4","mql5","expert-advisor-ea"],"relatedGuide":null},{"slug":"mt5","term":"MT5 (MetaTrader 5)","category":"platform","definition":"MT5 (MetaTrader 5) is the modern multi-asset trading platform from MetaQuotes Software Corp., positioned as the successor to MT4. MT5 introduced a more sophisticated programming language (MQL5, object-oriented), 64-bit architecture, multi-threaded backtesting, multi-instrument Strategy Tester, and Marketplace integration for EA distribution. Despite slow retail adoption for years, MT5 has become the editorial preference for new EA development because the platform's capabilities are materially superior for multi-pair strategies, ML-augmented systems, and modern backtest discipline.","relatedTerms":["mt5-metatrader-5","mql","mql5","mql4","expert-advisor-ea"],"relatedGuide":null},{"slug":"mql4","term":"MQL4","category":"files-config","definition":"MQL4 (MetaQuotes Language 4) is the proprietary programming language MetaQuotes designed for MT4 platform development. The language is C-like with procedural syntax, supports basic data types, includes built-in functions for technical analysis and trading operations, and compiles to platform-specific binaries (.ex4 files). MQL4's simpler syntax has produced the larger retail EA developer community, but newer architectures (ML-augmented strategies, multi-strategy ensembles) are more naturally expressed in MQL5.","relatedTerms":["mql","mql5","mt4","expert-advisor-ea"],"relatedGuide":null},{"slug":"mql5","term":"MQL5","category":"files-config","definition":"MQL5 (MetaQuotes Language 5) is the proprietary programming language for MT5 platform development. Compared to MQL4, MQL5 introduces object-oriented programming with proper class hierarchies, stricter typing, a substantially larger standard library, multi-threaded execution capabilities, and native support for the more sophisticated EA architectures that 2026's serious vendors deploy. Source compiles to .ex5 binaries.","relatedTerms":["mql","mql4","mt5-metatrader-5","expert-advisor-ea"],"relatedGuide":null},{"slug":"cfd","term":"CFD (Contract for Difference)","category":"basics","definition":"A Contract for Difference (CFD) is a leveraged financial derivative that lets retail traders speculate on price movement of an underlying asset (forex pair, stock index, commodity, crypto) without owning the asset. The trader and broker agree to exchange the cash difference between opening and closing price; gains and losses are settled in cash. CFDs are how most retail forex trading actually happens — when traders \"trade EURUSD,\" they're typically trading a EURUSD CFD with their broker, not the underlying spot forex market.","relatedTerms":["leverage","margin","spread"],"relatedGuide":null},{"slug":"machine-learning","term":"Machine Learning","category":"ai-ml","definition":"Machine learning (ML) is a subset of artificial intelligence focused on systems that learn from data rather than following hand-coded rules. In forex EA applications, ML models are trained on historical trade outcomes (labelled \"successful\" or \"unsuccessful\") to predict whether new trade signals are likely to succeed. Common ML algorithms in EAs: neural networks, gradient-boosted decision trees (XGBoost, LightGBM), support vector machines, and random forests. The market is saturated with \"AI EA\" marketing that uses ML as a buzzword; substantive ML implementation is rare relative to the marketing language.","relatedTerms":["neural-network","ml-filter","concept-drift"],"relatedGuide":null},{"slug":"arbitrage","term":"Arbitrage","category":"strategy","definition":"Arbitrage is the simultaneous purchase and sale of the same or related assets in different markets to exploit price discrepancies. Classical arbitrage is risk-free — buy at one price, sell at another higher price, instantly. In modern retail forex, true arbitrage opportunities are extraordinarily rare and capacity-limited because institutional electronic trading systems exploit any meaningful discrepancy within milliseconds. Most products marketed as \"forex arbitrage EAs\" to retail audiences are either statistical arbitrage (which carries genuine risk) or outright fraudulent products that don't actually arbitrage.","relatedTerms":["scalping","trend-following"],"relatedGuide":null},{"slug":"capital-floor","term":"Capital Floor","category":"risk","definition":"Capital floor is the editorial framework for the minimum deposit at which an EA can operate as the vendor intends. Below capital floor, several frictions compound: position sizing becomes coarse (micro-lot precision issues), margin pressure becomes constant (small adverse moves trigger margin calls), per-trade commissions consume meaningful fraction of returns, and broker-side execution becomes harder. Vendors who publish honest capital floors enable buyer-side fit assessment; vendors who advertise \"any account size\" while having strategy that requires $5,000 to operate properly produce predictable disappointment.","relatedTerms":["risk-per-trade","kill-switch","drawdown"],"relatedGuide":null},{"slug":"broker-tier","term":"Broker Tier","category":"execution","definition":"Broker tier in this editorial vocabulary classifies forex brokers by execution quality and regulatory standing. The 2026 editorial framework uses three tiers: Tier-1 ECN (premium execution, strong regulation, institutional-grade infrastructure — IC Markets Raw, Pepperstone Razor, Tickmill Pro); Tier-2 Standard (major regulated brokers but with mark-up pricing or less premium execution — Standard accounts at the Tier-1 brokers, mid-tier regulated brokers); Tier-3 Offshore (offshore-regulated brokers offering higher leverage and looser conditions — elevated regulatory tail risk).","relatedTerms":["tier-1-ecn","ecn-broker"],"relatedGuide":null},{"slug":"demand-zone","term":"Demand Zone","category":"analysis","definition":"A demand zone is a price range where historical buying volume produced sharp upward price moves. Visually identified on charts as zones where price spent time in tight consolidation immediately before accelerating upward; the consolidation reflects institutional accumulation, the subsequent acceleration reflects the moment buying overwhelmed supply. Demand zones are foundational to supply-and-demand trading methodology and to many breakout EAs. The principle is that when price returns to the historical demand zone, the same buying interest may reappear and produce another upward reversal.","relatedTerms":["supply-zone"],"relatedGuide":null},{"slug":"supply-zone","term":"Supply Zone","category":"analysis","definition":"A supply zone is a price range where historical selling volume produced sharp downward price moves. Visually identified on charts as zones where price spent time in tight consolidation immediately before accelerating downward; the consolidation reflects institutional distribution, the subsequent acceleration reflects the moment selling overwhelmed buying. Supply zones are the bearish mirror of demand zones in supply-and-demand trading methodology. When price returns to a historical supply zone, the same selling interest may reappear and produce another downward reversal.","relatedTerms":["demand-zone"],"relatedGuide":null},{"slug":"maximum-adverse-excursion","term":"Maximum Adverse Excursion (MAE)","category":"metrics","definition":"Maximum adverse excursion (MAE) is the largest unrealised loss a trade reached during its lifetime, regardless of where it ultimately closed. A trade that opened at 1.1000, dropped to 1.0950, and closed at 1.1050 has MAE of 50 pips even though it closed at +50 pips profit. MAE analysis separates favourable closes from properly-managed trades: trades with large MAE that closed profitable were lucky to recover; those with small MAE reflect tighter risk management. The metric is foundational to evaluating stop-loss placement effectiveness.","relatedTerms":["drawdown","kill-switch"],"relatedGuide":null},{"slug":"individual-leg-drawdown","term":"Individual Leg Drawdown","category":"metrics","definition":"Individual leg drawdown is the drawdown experienced by one strategy module (trend leg, mean-reversion leg, breakout leg) within a multi-strategy EA, calculated independently from the aggregate portfolio. A multi-strategy EA's aggregate drawdown may be 15% while the trend leg's individual drawdown was 35% — the aggregate number is smoothed by other legs earning during the trend leg's stress period. Per-leg drawdown analysis reveals whether the diversification benefit is real or whether one leg is masking dangerous concentration risk during specific market regimes.","relatedTerms":["drawdown","strategy-attribution","multi-strategy"],"relatedGuide":null},{"slug":"martingale-cap","term":"Martingale Cap","category":"risk","definition":"A martingale cap is the parameter limiting how many position-doubling steps a martingale EA will execute before halting. Standard martingale logic doubles position size after each loss, indefinitely; martingale caps stop the doubling after a configured number of steps (typically 4-7), accepting the cumulative loss rather than continuing to double. The cap bounds the worst-case immediate exposure but doesn't fix martingale's structural problem — caps just defer the inevitable failure mode.","relatedTerms":["kill-switch","risk-per-trade"],"relatedGuide":null},{"slug":"spread-cap","term":"Spread Cap","category":"execution","definition":"A spread cap is an EA parameter that prevents trade entries when the current bid-ask spread exceeds a defined threshold (e.g. 2.5 pips on EURUSD, 30 USD/oz on XAUUSD). The mechanism protects against entering trades during news events, low-liquidity periods, or unusual broker conditions where the wider spread would erode the strategy's edge. Without a spread cap, EAs may take trades at 5-10× normal spread cost during the 10-30 seconds around scheduled high-impact events, producing immediate losses that vendor's verification on tight-spread brokers doesn't expose.","relatedTerms":["spread","kill-switch"],"relatedGuide":null},{"slug":"grid-expansion","term":"Grid Expansion","category":"strategy","definition":"Grid expansion is the rule governing how a grid EA's pre-placed buy and sell orders extend their range as price moves outside the initial grid. Aggressive grid expansion continues adding new orders at increasing intervals deeper into adverse moves; conservative expansion caps the grid's extent and stops adding orders beyond defined boundaries. The expansion parameter is the central risk-control mechanism in grid EAs — and the primary source of grid architecture's unbounded tail risk when expansion is aggressive.","relatedTerms":["kill-switch","drawdown"],"relatedGuide":null},{"slug":"multi-timeframe","term":"Multi-Timeframe (MTF) Analysis","category":"analysis","definition":"Multi-timeframe (MTF) analysis evaluates trade setups across multiple chart timeframes simultaneously to filter signals. A trader looking for long entries on the M15 timeframe might confirm the broader H4 trend is also upward before taking M15 long entries, reducing exposure to counter-trend signals. The discipline reflects the principle that higher timeframes carry more market structure information than lower ones; trading lower-timeframe signals without higher-timeframe context produces excessive false signals.","relatedTerms":["higher-time-frame","mtf-alignment"],"relatedGuide":null},{"slug":"higher-time-frame","term":"Higher Time Frame (HTF)","category":"analysis","definition":"Higher time frame (HTF) is the broader-context chart timeframe relative to the primary trading timeframe. The relative nature matters: an M15 day-trader's HTF is H4 or D1; an M1 scalper's HTF is M15 or H1. HTFs provide market structure context that lower timeframes don't expose — major trend direction, key support/resistance levels, regime classification. Most disciplined trading methodologies require HTF context before lower-timeframe entries.","relatedTerms":["multi-timeframe","mtf-alignment"],"relatedGuide":null},{"slug":"mtf-alignment","term":"MTF Alignment","category":"analysis","definition":"MTF alignment is the trading condition where multiple chart timeframes (HTF, intermediate, entry timeframe) all signal the same directional bias or market context. Long entries with MTF alignment occur when daily, H4, and H1 timeframes all show upward trend; the M15 entry signal happens during the alignment. MTF-aligned entries have higher historical success rates than unaligned entries because the multi-timeframe agreement reduces false-signal probability.","relatedTerms":["multi-timeframe","higher-time-frame"],"relatedGuide":null},{"slug":"volatility-band","term":"Volatility Band","category":"analysis","definition":"A volatility band is a dynamic price envelope around a central moving average, calculated using volatility measurements. Bollinger Bands (most common variant) use 2 standard deviations of price action over the moving-average period; Keltner Channels use ATR. The bands expand during volatile periods (wider envelope) and contract during quiet periods (narrower envelope), automatically adjusting to market conditions. Volatility bands serve as adaptive support/resistance levels and provide volatility-regime classification.","relatedTerms":["bollinger-bands","atr"],"relatedGuide":null},{"slug":"session-timing","term":"Session Timing","category":"analysis","definition":"Session timing is an EA filter that restricts trade entries to specific forex market sessions based on session-specific volatility and liquidity patterns. The 24-hour forex market is divided into Asian (Tokyo, 23:00-08:00 UTC), European (London, 08:00-16:00 UTC), and American (New York, 13:00-22:00 UTC) sessions, with the London-NY overlap (13:00-16:00 UTC) being the highest-liquidity period. Session timing matters because volatility, spreads, and edge characteristics vary materially by session.","relatedTerms":["spread"],"relatedGuide":null},{"slug":"dxy-correlation","term":"DXY Correlation","category":"analysis","definition":"DXY correlation is the statistical relationship between individual currency pairs and the US Dollar Index — a weighted basket of EUR, JPY, GBP, CAD, SEK, CHF that represents broad dollar strength. Currency pairs are correlated to DXY by their currency composition: USD-quote pairs (EURUSD, GBPUSD, AUDUSD) typically have negative DXY correlation (rising dollar = falling pair); USD-base pairs (USDJPY, USDCHF, USDCAD) typically have positive DXY correlation. DXY correlation analysis reveals when multi-pair EA positions are actually concentrated dollar-direction exposure rather than diversified.","relatedTerms":["correlation-pair","currency-strength","correlation-cap"],"relatedGuide":null},{"slug":"currency-strength","term":"Currency Strength","category":"analysis","definition":"Currency strength is the aggregate performance of a single currency measured across all its forex pairs simultaneously. Rather than analysing EURUSD as a single pair, currency-strength analysis decomposes the movement into EUR strength and USD strength components separately. When USD strengthens broadly, every USD-base pair rises and every USD-quote pair falls — currency-strength indicators visualise this directly, helping traders distinguish 'EUR weakness' from 'USD strength' which may have different driving factors and persistence.","relatedTerms":["dxy-correlation","correlation-pair"],"relatedGuide":null},{"slug":"correlation-pair","term":"Correlation Pair","category":"analysis","definition":"Correlation pairs are forex pairs whose price changes show statistical correlation — positive correlation means pairs move in the same direction, negative correlation means opposite directions. EURUSD and GBPUSD are strongly positively correlated (both have USD as quote; broad USD moves affect both). EURUSD and USDCHF are strongly negatively correlated (USD-quote vs USD-base, opposite reactions to dollar moves). For multi-pair EAs, understanding correlation pairs prevents accidental concentration: running both EURUSD and GBPUSD on long bias is essentially a single concentrated USD-short bet despite the appearance of multi-pair diversification.","relatedTerms":["correlation-cap","dxy-correlation","currency-strength"],"relatedGuide":null},{"slug":"decoupling","term":"Decoupling","category":"analysis","definition":"Decoupling is the period during which historically-correlated forex pairs (or other instruments) lose their typical correlation. EURUSD and GBPUSD normally have +0.85 correlation; during decoupling periods (often triggered by GBP-specific events like Brexit-style catalysts), this correlation can drop to +0.20 or even reverse. Decoupling matters for risk management because the diversification assumptions baked into multi-pair EA architectures depend on stable correlations; when correlations decouple, the EA's assumed risk distribution changes materially.","relatedTerms":["correlation-pair","correlation-cap"],"relatedGuide":null},{"slug":"ecb-event","term":"ECB Event","category":"analysis","definition":"ECB event refers to scheduled European Central Bank communication events — primarily the monthly Governing Council meeting interest-rate decisions (typically every 6 weeks) and the subsequent press conference. ECB events produce significant EUR volatility: typical EURUSD moves are 50-150 pips during the 60-minute window starting at the announcement; spreads widen 2-3× normal levels for 5-30 minutes; news-related execution friction makes EA trading during the window unprofitable for most strategies.","relatedTerms":["spread-cap"],"relatedGuide":null},{"slug":"eia-release","term":"EIA Release","category":"analysis","definition":"EIA release refers to the US Energy Information Administration's weekly Petroleum Status Report — published Wednesdays at 14:30 UTC (15:30 if DST), reporting US crude oil inventory levels for the prior week. The release moves WTI and Brent oil prices materially: typical immediate WTI move is 2-5%, sometimes 8-10% on large surprises. The volatility also transmits to oil-correlated currencies (USD/CAD via Canadian oil exposure, NOK pairs via Norwegian oil) producing tradeable patterns and EA execution friction.","relatedTerms":["opec-plus"],"relatedGuide":null},{"slug":"opec-plus","term":"OPEC+ Event","category":"analysis","definition":"OPEC+ events are the scheduled and unscheduled meetings of the Organization of Petroleum Exporting Countries (OPEC) plus its expanded coalition including Russia and other non-OPEC oil producers. The group meets approximately every 1-2 months to discuss production quotas; quota changes affect global oil supply and trigger major oil-price moves. Unlike scheduled news events, OPEC+ announcements sometimes occur outside trading hours or with unusual timing, complicating EA news-filter logic.","relatedTerms":["eia-release"],"relatedGuide":null},{"slug":"training-data-window","term":"Training Data Window","category":"ai-ml","definition":"Training data window is the historical period from which an ML model's training examples are drawn. A 12-month training window means the model learns from the past 12 months of price data; a 5-year window means 5 years of history. The choice trades off two concerns: short windows produce models tightly fitted to recent market microstructure (high recency relevance) but vulnerable to regime shifts (limited regime coverage); long windows produce regime-robust models but slower to adapt to recent microstructure evolution.","relatedTerms":["machine-learning","neural-network","concept-drift","retraining-cadence"],"relatedGuide":null},{"slug":"z-score","term":"Z-Score","category":"metrics","definition":"Z-score is the standardised distance of a value from a distribution's mean, expressed in standard-deviation units. Z-score = (value − mean) ÷ standard deviation. A z-score of 0 means the value equals the mean; +1 means one standard deviation above the mean; +2 means two standard deviations above; etc. In financial applications, z-scores measure how extreme a current price, indicator value, or spread is relative to its historical distribution. Z-scores are foundational to statistical arbitrage, mean-reversion strategies, and regime-state classification.","relatedTerms":["volatility-band"],"relatedGuide":null}]}