Glossary

89 terms, defined plainly. Each links to the lesson that explains it properly.

Adverse selection: The tendency for a liquidity provider to be filled most often by traders who know where price is going next, so fills cluster just before losses. Lesson
Alpha: Return in excess of what a strategy’s exposure to known risk factors would explain. The part you can credit to skill. Lesson
Arbitrage: Profiting from a price difference between equivalent instruments. True arbitrage is riskless; most real "arb" carries execution, venue or basis risk. Lesson
ATRAverage true range: The average of each bar’s true range (high–low, extended to include gaps from the prior close). A standard measure of volatility used for stops and position sizing. Lesson
Autocorrelation: Correlation of a series with its own past values. Positive at lag one suggests momentum; negative suggests mean reversion. Lesson
Avellaneda–Stoikov: A market-making model that skews quotes away from accumulated inventory and widens spreads as volatility and risk aversion rise. Lesson
Backtest: A simulation of how a trading rule would have performed on historical data. Only as honest as its assumptions about data, timing and costs. Lesson
Basis: The difference between a futures price and the spot price of its underlying. Converges to zero at expiry. Lesson
Basis pointbp, bps: One hundredth of a percent (0.01%). Also the unit of edge you earn on this site.
Bayes’ rule: P(A|B) = P(B|A)P(A)/P(B). How to update a prior belief after seeing evidence. Lesson
Beta: Sensitivity of an asset or strategy to a benchmark. A beta of 1.5 moves about 1.5% for each 1% benchmark move. Lesson
Black–Scholes: The closed-form price of a European option under lognormal prices, constant volatility and continuous hedging. Lesson
Bollinger Bands: Bands set a number of standard deviations above and below a moving average. Used as a z-score for mean-reversion and breakout rules. Lesson
Brownian motionWiener process: A continuous random walk with independent, normally distributed increments whose variance grows linearly with time. Lesson
Calmar ratio: Annualised return divided by maximum drawdown. Intuitive, but noisy and dependent on sample length. Lesson
Carry: The return earned from holding a position if prices do not change, such as interest differentials or funding payments. Lesson
Central limit theorem: The sum of many independent, finite-variance random variables is approximately normally distributed. Lesson
Cointegration: A property of two or more non-stationary series whose linear combination is stationary. The statistical basis of pairs trading. Lesson
Contango: When futures trade above spot, or longer-dated futures above shorter-dated ones. The opposite is backwardation. Lesson
Covariance matrix: The matrix of variances (diagonal) and covariances (off-diagonal) of a set of assets. Portfolio variance is wᵀΣw. Lesson
Cross-validation: Estimating out-of-sample performance by repeatedly fitting on part of the data and testing on the rest. In finance it must respect time order. Lesson
DCADollar-cost averaging: Investing a fixed amount on a fixed schedule regardless of price. Lowers average cost to the harmonic mean of prices. Lesson
Deflated Sharpe ratio: The probability that a strategy’s true Sharpe exceeds zero after correcting for the number of trials, skewness and kurtosis. Lesson
Delta: Sensitivity of an option’s price to the underlying price. Also the hedge ratio for a delta-neutral position. Lesson
Drawdown: The percentage decline from a running equity peak. Maximum drawdown is the worst such decline in a period. Lesson
Efficient frontier: The set of portfolios with the highest expected return for each level of risk. Lesson
Embargo: A buffer of observations removed after each test fold in cross-validation to stop information leaking forward into training. Lesson
Expected shortfallCVaR: The average loss in the worst (1 − α) fraction of outcomes. A coherent alternative to VaR. Lesson
Expected value: The probability-weighted average outcome of a random variable. Lesson
Fat tails: More extreme outcomes than a normal distribution predicts. Measured by kurtosis above 3. Lesson
Funding rate: Periodic payment between longs and shorts on a perpetual swap that keeps its price near spot. Positive means longs pay shorts. Lesson
Gamma: Rate of change of an option’s delta. Long options are long gamma: they gain from large moves in either direction. Lesson
GBMGeometric Brownian motion: The standard continuous-time model of a price: proportional drift plus proportional Brownian noise. Implies lognormal prices. Lesson
Grid trading: Placing ladders of buy orders below and sell orders above the price to harvest oscillation. Effectively short volatility. Lesson
Half-life: For a mean-reverting process, the expected time for a deviation from the mean to shrink by half: ln 2 / θ. Lesson
Implementation shortfall: The total cost of executing an order, measured against the price at the moment the decision was made. Lesson
Implied volatility: The volatility that makes a pricing model match an option’s market price. The market’s priced forecast of future volatility. Lesson
Information ratio: Active return over a benchmark divided by tracking error. Sharpe’s analogue for benchmarked managers. Lesson
Itô’s lemma: The chain rule of stochastic calculus. Adds a second-order term because (dW)² behaves like dt. Lesson
Kelly criterion: The bet size that maximises long-run logarithmic growth: edge divided by odds, or μ/σ² for continuous returns. Lesson
Kurtosis: The fourth standardised moment: a measure of tail weight. A normal distribution has kurtosis 3. Lesson
Limit order: An order to trade at a specified price or better. It rests in the book and provides liquidity. Lesson
Log return: ln(Pₜ/Pₜ₋₁). Adds across time, which makes it convenient for compounding and statistics. Lesson
Look-ahead bias: Using information in a backtest that was not available at the time of the simulated decision. Lesson
MACD: Moving average convergence–divergence: the difference between a fast and slow EMA, with a signal line. A trend and momentum indicator. Lesson
Maker / taker: Makers add liquidity with resting limit orders; takers remove it with marketable orders. Venues usually charge takers more. Lesson
Market impact: The adverse price move caused by your own trading. Roughly proportional to the square root of size relative to volume. Lesson
Market making: Continuously quoting both a bid and an ask to earn the spread while managing inventory risk. Lesson
Martingale (betting): Increasing position size after losses to recover them in one win. Guarantees ruin given a long enough losing streak. Lesson
Mean reversion: The tendency of a quantity to return towards an average after deviating from it. Lesson
Meta-labelling: Training a secondary model to predict whether a primary signal will be correct, and using that probability to size or filter trades. Lesson
Micro-price: A fair-value estimate that weights the best bid and ask by the opposite side’s size. Better than the mid for short-horizon prediction. Lesson
Momentum: The tendency of recent winners to keep outperforming and recent losers to keep underperforming over medium horizons. Lesson
Order-book imbalance: The difference between bid and ask size at the top of the book, scaled by their sum. Predicts the next small mid-price move. Lesson
Ornstein–Uhlenbeck process: A continuous-time mean-reverting process: dX = θ(μ − X)dt + σdW. Lesson
Overfitting: Fitting a model to noise in the sample, so that in-sample performance overstates future performance. Lesson
Paper trading: Running a strategy against live market data with simulated orders and no real capital. Lesson
PCAPrincipal component analysis: Decomposing a covariance matrix into uncorrelated factors ordered by how much variance each explains. Lesson
Perpetual swapPerp: A futures contract with no expiry, kept near spot by periodic funding payments. Lesson
Pine Script: TradingView’s scripting language for indicators and strategies. Most specs in the vault are written in it.
Purging: Removing training observations whose label windows overlap the test set’s, to prevent leakage. Lesson
Put–call parity: C − P = S − Ke^(−rT) for European options on the same strike and expiry. Ties call and put prices together. Lesson
R-squared: The fraction of variance in a dependent variable explained by a regression. Lesson
Risk parity: A portfolio in which each asset contributes equally to total risk, rather than receiving equal capital. Lesson
Risk-neutral pricing: Pricing derivatives as if all assets earned the risk-free rate, justified by the ability to hedge. Lesson
RSIRelative strength index: An oscillator from 0 to 100 comparing recent gains to losses. Readings above 70 or below 30 are commonly treated as overbought or oversold. Lesson
Sharpe ratio: Mean excess return divided by its standard deviation. The standard measure of risk-adjusted performance. Lesson
Skew (volatility): The pattern of implied volatility across strikes. Equity indices show higher IV for low strikes. Lesson
Skewness: The third standardised moment: asymmetry of a distribution. Negative skew means occasional large losses. Lesson
Slippage: The difference between the expected and actual execution price of a trade. Lesson
Sortino ratio: Like Sharpe, but divides by downside deviation only, so upside volatility is not penalised. Lesson
Spread: The gap between the best ask and the best bid. The cost of a round trip with market orders. Lesson
Stat-arbStatistical arbitrage: Trading mean-reverting residuals across many related assets, typically after removing common factors. Lesson
Stationarity: A process whose mean, variance and autocovariance structure do not change over time. Lesson
Supertrend: A trend indicator that trails price by a multiple of ATR and flips when price closes through it. Lesson
Survivorship bias: Testing only on assets that still exist, which excludes the failures and flatters results. Lesson
t-statistic: An estimate divided by its standard error. For a strategy, approximately the annual Sharpe times the square root of years. Lesson
Theta: Sensitivity of an option’s value to the passage of time. Usually negative for long options. Lesson
Time-series momentumTrend following: Going long assets with positive past returns and short those with negative past returns, each judged on its own history. Lesson
Triple-barrier method: Labelling each trade by which comes first: a profit target, a stop loss, or a time limit, all scaled by volatility. Lesson
TWAP: Time-weighted average price: executing an order in equal slices over equal intervals. Lesson
Value at RiskVaR: The loss threshold not exceeded with a given probability over a given horizon. Lesson
Vega: Sensitivity of an option’s value to implied volatility. Lesson
Volatility: The standard deviation of returns, usually annualised. Lesson
Volatility drag: The gap between arithmetic and geometric returns, about σ²/2. Why volatile assets compound more slowly than their average return. Lesson
Volatility targeting: Scaling position size inversely to recent volatility to keep portfolio risk roughly constant. Lesson
VWAP: Volume-weighted average price. Both a benchmark and an execution algorithm that trades in line with market volume. Lesson
Walk-forward analysis: Repeatedly fitting on a past window and testing on the next one, then stitching the out-of-sample pieces together. Lesson
Z-score: How many standard deviations a value sits from its mean: (x − μ)/σ. Lesson