Algorithmic Trading Glossary — Every Term Defined (India)

Adjusted price Data

A historical price series modified to remove the mechanical jumps caused by splits, bonuses and dividends, so returns are continuous. Using unadjusted prices creates false gaps that distort signals and performance. also: adjusted close, back-adjusted price

Algorithmic trading Core

Trading in which orders are generated and often submitted by a computer program following a predefined set of rules, rather than by manual discretion. The rules cover entry, exit, sizing and risk, and can run at any speed from daily to sub-second. also: algo trading, algorithmic execution

Alpha Strategy

The component of return attributable to skill or a genuine predictive edge, above what market exposure (beta) alone would deliver. In practice it is the risk-adjusted excess return a strategy adds. also: excess return

API error API

A structured failure response from an API, such as an invalid order, insufficient margin, rate-limit or server error. Distinguishing retryable from terminal errors is essential to avoid duplicate or lost orders.

Authentication API

The process by which an API confirms a client's identity, typically via API keys, tokens or a login-and-session flow. Handling credentials and session expiry correctly is a common source of live outages.

Automated trading Core

Trading in which order placement is executed by software with no manual step, typically via a broker API. Automation is about execution mechanics; the strategy behind it may be systematic or not. also: auto trading

Automation workflow API

The end-to-end orchestration that ties data fetch, signal generation, order placement, monitoring and reporting into an unattended pipeline. Its reliability, not the strategy alone, determines whether a live system runs cleanly.

Backtesting Backtesting

Simulating a strategy on historical data to estimate how it would have behaved, including returns, drawdowns and turnover. A backtest is a hypothesis test about a rule set, not a promise of future results.

Beta Metric

A measure of how much an instrument or strategy moves with a benchmark market. Return explained by beta is ordinary market exposure, distinct from alpha, which is the skill-based excess.

Bid-ask spread Execution

The gap between the highest price buyers will pay (bid) and the lowest sellers will accept (ask). It is an immediate cost of crossing the market and a basic measure of liquidity. also: spread, bid-offer spread

Breakout Strategy

An entry taken when price moves beyond a defined range boundary, such as a prior high, on the premise that the break signals a new move. Breakout systems must handle false breaks that reverse immediately.

Broker API API

The programmatic interface a broker exposes so software can fetch data and place, modify and cancel orders. In India, retail algo access is typically through such APIs under the broker's and exchange's terms. also: trading API

CAGR Metric

The compound annual growth rate — the constant yearly rate that would take the starting equity to the ending equity over the period. It smooths a lumpy equity curve into a single annualised figure. also: compound annual growth rate

Calmar ratio Metric

A risk-adjusted measure equal to annualised return (CAGR) divided by the maximum drawdown over the same period. It captures return relative to the worst peak-to-trough loss an investor would have endured.

Capital allocation Risk

How total capital is distributed across strategies, instruments or sub-portfolios. Good allocation balances expected return against correlation and concentration risk.

CI/CD Programming

Continuous integration and continuous delivery — automated pipelines that test and deploy code changes reliably. They reduce the chance a broken change reaches a live trading system.

Circuit breaker Risk

An automatic halt that stops trading when a preset limit is breached — at exchange level (index bands) or within a trading system (loss or error limits). It buys time to prevent a runaway situation from compounding.

Cointegration Strategy

A statistical property where two or more non-stationary price series move together such that a linear combination of them is stationary (mean-reverting). It is the rigorous basis for pairs and spread trading, and is stronger than simple correlation.

Configuration file Programming

An external file (such as YAML, JSON or .env) that holds parameters, credentials and settings separately from code. It lets the same code run different strategies or environments without edits and keeps secrets out of source.

Corporate action Data

An event such as a dividend, split, bonus issue or merger that changes an instrument's shares or price mechanically. Backtests must adjust historical prices for these or signals and returns will be wrong around the event.

Curve fitting Backtesting

The act of adjusting parameters or adding rules until the equity curve looks good on past data. It is the mechanism by which overfitting happens and is the single most common way backtests mislead.

Data cleaning Data

Detecting and correcting errors in raw market data — bad ticks, wrong decimals, duplicated or out-of-order rows — before it feeds a model. Silent data errors are a common source of illusory backtest edges.

Data layer Architecture

The part of a system responsible for ingesting, storing, cleaning and serving market data to the rest of the pipeline. Its correctness underpins every downstream signal and backtest.

Data snooping Backtesting

Reusing the same dataset to test many hypotheses until one appears significant by chance. Without correcting for the number of trials, the winner is often a false positive rather than a real edge. also: data dredging, p-hacking

Data structures Programming

The way data is organised in code — arrays, hash maps, queues and time-indexed frames — chosen for the access patterns a strategy needs. Good choices keep backtests fast and live loops within their latency budget.

Data validation Data

Automated checks that confirm incoming data meets expected rules — plausible ranges, monotonic timestamps, no duplicates — before it is trusted. It stops corrupt data from silently poisoning live signals.

Discretionary trading Core

Trading where a human makes each decision using judgement, context and experience rather than a fixed rule set. It is the conceptual opposite of systematic trading and is hard to backtest objectively.

Drawdown Metric

The decline in equity from a prior high-water mark at any point in time, expressed as a percentage or amount. A strategy spends much of its life in some level of drawdown between new equity highs.

Edge Strategy

A statistical advantage that makes a strategy profitable on average over many trades, net of costs. An edge is expressed through positive expectancy, not through any single winning trade.

Error handling Architecture

The code paths that anticipate and recover from failures such as rejected orders, dropped connections and bad data. Robust handling decides whether a glitch is a non-event or an uncontrolled loss.

Event streaming API

Delivering a continuous flow of events — ticks, order updates, fills — from source to consumers as they happen. It underpins real-time strategy reaction and is usually carried over WebSockets or message queues.

Event-driven architecture Architecture

A system design where components communicate by emitting and reacting to events (bars, ticks, fills, signals) rather than calling each other directly. It mirrors how a live trading system actually receives information and decouples the stages.

Event-driven backtest Backtesting

A backtest that steps through data one event at a time, passing each bar or tick to the strategy as it would arrive live. It is slower but mirrors production logic and naturally prevents look-ahead bias.

Execution engine Architecture

The subsystem that turns a target position or signal into actual orders, choosing order type, size and timing and handling responses. It sits between strategy logic and the broker API.

Execution quality Execution

How well orders are filled relative to a benchmark such as the arrival price, measured through slippage, fill rate and market impact. Poor execution quietly erodes an otherwise sound strategy.

Expectancy Metric

The average profit or loss expected per trade, computed as (win rate x average win) minus (loss rate x average loss). Positive expectancy is the mathematical definition of an edge.

Failover Architecture

The ability of a system to switch to a backup process, connection or server when the primary one fails, ideally without losing position state. It protects a live strategy from single points of failure.

Forward testing Backtesting

Evaluating a fixed strategy on new data as it arrives going forward, whether on paper or with small live capital. Unlike a backtest, it cannot be re-run or tuned, so it is a cleaner test of genuine edge. also: forward test

High-frequency trading Strategy

Automated trading at very short horizons, often sub-second, that depends on low latency and high order throughput. It is largely an institutional, infrastructure-heavy domain rather than a retail one. also: HFT

Historical data Data

Past market data used to research and backtest strategies. Its quality, adjustment and point-in-time accuracy determine how trustworthy any backtest built on it can be.

Idempotency API

A property where sending the same request more than once has the same effect as sending it once. For order placement it prevents a retry after a timeout from accidentally creating duplicate positions.

In-sample Backtesting

The portion of historical data used to develop and fit a strategy's parameters. Performance measured here is optimistic because the rules were tuned to it. also: training data

Kelly criterion Risk

A formula that computes the bet fraction maximising the long-run growth rate of capital given an edge and its odds. Full Kelly is highly volatile, so practitioners commonly use a fraction of it to reduce drawdowns.

Kill switch Risk

A mechanism that instantly halts all trading and, ideally, flattens open positions when triggered manually or by a safety rule. It is the last line of defence against a malfunctioning strategy or feed.

Latency Execution

The time delay between an event occurring and the system reacting to it, spanning data receipt, computation and order round-trip. Lower latency matters most for market making and short-horizon strategies.

Limit order Execution

An instruction to trade only at a specified price or better. It guarantees the price but not a fill, and may rest unexecuted if the market never reaches the limit.

Liquidity Execution

The ease of trading an instrument in size without moving its price much, reflected in volume, depth and spread. Illiquid instruments produce wider spreads, larger slippage and unreliable backtest fills.

Logging Architecture

Recording a durable, timestamped trail of decisions, orders, fills and errors as a system runs. Good logs are essential for reconciliation, debugging live issues and auditing behaviour after the fact.

Look-ahead bias Backtesting

Using information in a backtest that would not have been known at the moment of the simulated decision, such as filling on a close price the strategy used to trigger. It manufactures returns that are impossible to capture live. also: lookahead bias

Lookback window Core

The number of past bars or the time span an indicator or signal uses to compute its current value. Too short makes signals noisy; too long makes them slow, and the choice is a frequent target of overfitting. also: lookback period

Market making Strategy

Continuously quoting both a bid and an ask to earn the spread by providing liquidity, while managing inventory risk. Profitability depends on tight spreads, low latency and disciplined inventory control.

Market order Execution

An instruction to buy or sell immediately at the best available price. It guarantees a fill but not a price, so in thin or fast markets it can suffer meaningful slippage.

Maximum drawdown Metric

The largest peak-to-trough decline in equity before a new peak is reached, usually stated as a percentage. It is a core measure of pain and of how much capital a strategy can put at risk. also: max drawdown, MDD

Mean reversion Strategy

A strategy family that bets a price stretched away from a reference level will return toward it. It often has a high win rate but risks large losses when a move keeps trending instead of reverting.

Minute data Data

Intraday bars aggregated to one-minute (or similar) intervals, a common middle ground between ticks and daily data. It supports most intraday backtests without the storage burden of full tick data.

Missing data Data

Gaps in a series caused by holidays, halts, feed outages or unlisted periods. How gaps are filled or excluded changes indicator values and returns, so the policy must be explicit and consistent.

Momentum Strategy

The tendency of assets that have performed well (or poorly) recently to continue in the same direction over a horizon. Momentum systems rank or time instruments by recent relative strength.

Monitoring Architecture

Continuously observing a running system's health, positions, latency and profit-and-loss, with alerts when values leave expected bounds. It is how operators catch a stuck feed or runaway strategy before losses grow.

Monte Carlo simulation Backtesting

A technique that generates many randomised scenarios — for example by resampling the trade sequence — to study the range of possible outcomes rather than a single historical path. It helps estimate drawdown risk and the sensitivity of results to luck.

OHLC Data

The open, high, low and close prices summarising trading over a fixed interval such as a day or a minute. It is the most common compact format for charting and bar-based backtests. also: OHLCV, candle, bar

OMS Architecture

An order management system — the component that creates, submits, tracks and reconciles orders and their state through the broker. It is the authoritative record of what has been sent, filled or cancelled. also: order management system

Order book Execution

The live list of resting buy and sell limit orders at each price level for an instrument. Its depth and shape reveal available liquidity and help estimate the price impact of a trade. also: limit order book, depth of market

Order lifecycle Core

The states an order moves through — created, submitted, acknowledged, partially or fully filled, or rejected/cancelled. Tracking these states accurately is essential to keep position and risk in sync with reality.

Out-of-sample Backtesting

Data withheld during development and used only afterwards to test a fixed strategy. It gives a more honest estimate of real behaviour because the rules never saw it during tuning. also: OOS, holdout

Overfitting Backtesting

Tuning a strategy so closely to historical data that it captures noise rather than a repeatable pattern, producing a great backtest that fails live. More parameters and more tuning iterations increase the risk.

Pairs trading Strategy

A statistical-arbitrage technique that goes long one instrument and short a related one when their price spread diverges from its historical norm, expecting convergence. The pair should be economically linked, not merely correlated by chance.

Paper trading Backtesting

Running a strategy against live market data with simulated orders and no real money. It surfaces data-feed, timing and operational issues that a backtest cannot, but fills are still assumed rather than real.

Partial fill Execution

When only part of an order's quantity executes and the remainder stays open or is cancelled. Systems must track filled versus outstanding quantity so position and risk state stay correct.

Point-in-time data Data

Historical data reconstructed to show exactly what was known and observable at each past moment, including later-revised figures as originally reported. Using it prevents look-ahead bias from restated fundamentals or index membership. also: PIT data

Portfolio diversification Risk

Spreading capital across instruments or strategies whose returns are not perfectly correlated, so their combined equity curve is smoother than any one alone. Diversification benefit falls sharply when correlations spike in a crisis.

Portfolio engine Architecture

The component that aggregates signals and positions across strategies and instruments into a single target portfolio, applying allocation and netting rules. It resolves conflicts when multiple strategies want the same instrument.

Portfolio heat Risk

The total open risk across all live positions, measured as the sum of what would be lost if every open trade hit its stop at once. Capping heat prevents many correlated positions from combining into an outsized loss.

Position sizing Risk

Deciding how many units, lots or contracts to trade on a given signal, based on account capital, risk per trade and stop distance. It usually matters more to long-run results than the entry rule itself.

Profit factor Metric

Gross profit divided by gross loss across all trades. A value above 1 means the strategy made money in the sample, but a single large winner can inflate it, so it needs a decent sample size.

Python for trading Programming

Using Python and its ecosystem (pandas, NumPy, backtesting and broker-API libraries) to research, backtest and automate strategies. It dominates quant research for its libraries despite not being the fastest language.

Quantitative trading Core

Systematic trading whose rules are derived from statistical, mathematical or data-driven models rather than qualitative views. It emphasises measurement, hypothesis testing and empirical edge. also: quant trading

R-multiple Metric

A trade's profit or loss expressed in units of the initial risk (R) taken on it, so a trade risking 1R that gains twice its risk is +2R. It normalises outcomes across trades of different sizes and instruments. also: R multiple

Rate limit API

A cap the API sets on how many requests a client may make per interval. Exceeding it returns errors or temporary blocks, so clients must throttle, batch and back off to stay within the ceiling. also: rate limiting

Rebalancing Strategy

Periodically adjusting positions back to their target weights or signals as prices and allocations drift. More frequent rebalancing tracks targets closely but incurs more transaction cost.

Regime Core

A persistent market environment — such as trending versus range-bound, or high versus low volatility — in which a strategy's behaviour differs. Many strategies work in one regime and lose in another, so regime shifts are a key failure mode.

REST API API

A request-response web interface where the client asks for data or actions over HTTP and receives a reply. It suits order placement and snapshot queries but is not ideal for continuous real-time streaming. also: REST

Retry/backoff API

A pattern that re-attempts a failed request after waiting, with the wait growing (often exponentially) between tries. It recovers from transient failures without hammering an API that is rate-limited or briefly down. also: exponential backoff, retry logic

Risk engine Architecture

A component that checks every intended order against limits — position size, exposure, loss and portfolio heat — and can veto or resize it before it reaches the market. It enforces risk rules independently of strategy logic.

Risk of ruin Risk

The probability that a sequence of losses reduces capital below a threshold at which the account can no longer trade. It rises sharply with larger position sizes and with negative or thin expectancy.

Risk per trade Risk

The amount of capital, often a fixed small percentage, a trader is willing to lose if a single trade hits its stop. Keeping it small and consistent is what caps the depth of losing streaks.

Scheduling API

Running trading tasks at defined times or intervals, such as squaring off before market close or fetching data after the session. Robust scheduling must account for market holidays, half-days and time zones.

Sharpe ratio Metric

A risk-adjusted return measure equal to excess return divided by the standard deviation of returns. It rewards consistent returns and penalises volatility, but treats upside and downside volatility alike.

Signal Strategy

A discrete instruction produced by strategy logic — such as go long, go short, or flat — derived from data at a point in time. Signals are separate from execution, which decides how the resulting order is actually filled.

Signal generation Architecture

The stage that consumes cleaned data and produces trading signals according to strategy rules. Keeping it separate from execution makes both easier to test and reason about.

Slippage Execution

The difference between the price a strategy expected and the price at which the order actually filled. It grows with order size, thin liquidity and fast markets, and is a leading reason live results trail backtests.

Smart order routing Execution

Logic that decides how and where to send an order — splitting size, choosing order types or timing — to improve fills. Even on a single venue it governs slicing and pacing to reduce market impact. also: SOR

Sortino ratio Metric

A variant of the Sharpe ratio that divides excess return by downside deviation only, ignoring upside volatility. It better reflects an investor's real concern with losses rather than gains.

Statistical arbitrage Strategy

A market-neutral approach that trades many statistically related instruments to profit from short-term mispricings while hedging broad market risk. It relies on mean-reverting spreads and diversification across many small bets. also: stat arb

Stop loss Risk

A predefined exit that closes a losing position once price reaches a chosen level, capping the loss on that trade. In live trading it is typically enforced with a stop order routed to the exchange. also: stop

Stop order Execution

An order that stays dormant until price reaches a trigger level, at which point it becomes a market or limit order. It is the standard mechanism for enforcing a stop loss or a breakout entry. also: stop-loss order, SL order

Strategy Strategy

A complete specification of when to enter, how much to trade, and when to exit, together with its risk controls. In a system it is the logic that turns market data into intended positions.

Strategy class Programming

A code structure that encapsulates one strategy's state and its logic for handling incoming data and producing orders. A clean interface lets the same class run identically in backtest and live modes.

Survivorship bias Backtesting

A distortion caused by testing only instruments that still exist today, silently excluding those that were delisted, merged or went bankrupt. It flatters results because the failures are missing from the sample. also: survivorship

System architecture Architecture

The overall design of a trading system and how its parts — data, signals, sizing, risk, execution and monitoring — fit together. A clean architecture makes a system testable, observable and safe to run live.

Systematic trading Core

An approach in which every trading decision follows an explicit, repeatable rule set, so the same inputs always produce the same action. It removes in-the-moment judgement even if a human still presses the button. also: rule-based trading

Tick data Data

The most granular market data, recording individual trades or quote updates as they occur with timestamps. It is essential for execution and microstructure research but is large and demanding to store and process.

Time zone handling Data

Correctly aligning timestamps across data sources, the exchange and the server to a consistent reference, typically the market's local time. Mismatches cause bars to be misaligned and signals to fire at the wrong moment. also: time zones

Trading lifecycle Core

The full sequence a trade passes through: data to signal, signal to sized order, order to fill, and open position to exit and settlement. Mapping it clarifies where each system component fits.

Trailing stop Risk

A stop level that ratchets in the direction of profit as price moves favourably but never loosens. It locks in gains while giving a trend room to continue.

Transaction cost Execution

The full cost of trading, including brokerage, exchange fees, taxes such as STT, the bid-ask spread and slippage. Realistic cost modelling in a backtest often turns an apparent edge into a loss. also: trading cost

Trend following Strategy

A strategy family that buys strength and sells weakness on the assumption that established price moves tend to persist. It typically has a low win rate offset by large winners and cuts losers quickly. also: momentum-of-price

Unit testing Programming

Writing automated checks that verify small pieces of logic behave as intended. In trading it guards against subtle bugs in indicators, sizing and order handling that could otherwise cost real money. also: testing

Vectorised backtest Backtesting

A backtest that computes signals and returns across the whole price series using array operations at once. It is fast for research but can hide look-ahead bias and struggles to model order-level details like partial fills. also: vectorized backtest

Version control Programming

Tracking changes to code and configuration over time, usually with Git, so every strategy version is reproducible. It is essential for knowing exactly which code produced a given backtest or live trade. also: Git

Volatility Metric

The degree of variation in an instrument's returns, usually measured by the standard deviation of returns over a window. It scales position sizing, risk estimates and many strategy signals.

Volatility targeting Risk

Sizing positions so the portfolio's expected volatility stays near a chosen level, scaling exposure down when markets are turbulent and up when calm. It stabilises risk but adds turnover and can lag sudden shocks.

Walk-forward analysis Backtesting

A validation method that repeatedly optimises on one window and tests on the next unseen window, rolling forward through history. It approximates how periodic re-optimisation would perform out-of-sample. also: walk-forward testing, walk-forward optimisation

WebSocket API

A persistent two-way connection that lets a server push live updates — quotes, ticks, order fills — to a client without repeated polling. It is the usual channel for real-time market data feeds. also: websockets

Win rate Metric

The fraction of trades that close profitable. A high win rate is neither necessary nor sufficient for profitability; it must be read together with the size of wins versus losses.