Glossary topic
Risk & Quant
48 terms
Why this topic matters
Risk and quantitative methods turn a trading idea into numbers you can test: how much it can lose, how its returns are distributed, how positions move together, and whether a backtest result holds up. This topic covers risk metrics, portfolio theory, position sizing, and the statistics behind them.
Each entry works the formula with real numbers and states its assumptions. Use these tools to size trades, compare strategies on a risk-adjusted basis, and catch the backtesting errors that make results look better than they are.
Topic questions
Risk & Quant FAQ
A
1 termB
2 termsBacktesting
Backtesting is the practice of applying fully specified trading or investment rules to historical data to estimate how they would have behaved in the past.
Beta
Beta is a measure of how strongly a security’s returns move with a selected market benchmark, equal to return covariance divided by benchmark return variance.
C
5 termsCalmar Ratio
The Calmar ratio measures risk-adjusted return by dividing compound annual return by the maximum drawdown over the same window, conventionally the trailing 36 months.
Capital Asset Pricing Model (CAPM)
The Capital Asset Pricing Model (CAPM) is a model that sets an asset’s required return at the risk-free rate plus beta times the expected market risk premium.
Conditional VaR (Expected Shortfall)
Conditional value at risk (CVaR), or expected shortfall, is the average loss across the worst selected fraction of outcomes, such as the worst 5% at 95% confidence.
Correlation
Correlation is a unitless measure, from −1 to +1, of the strength and direction of the linear relationship between two series, such as two assets’ returns.
Covariance
Covariance is a measure of how two return series move together, calculated as the average product of their deviations from their own means.
D
1 termE
3 termsEfficient Frontier
The efficient frontier is the set of portfolios that offer the highest expected return for each level of volatility, given the available assets and constraints.
Efficient Market Hypothesis
The efficient market hypothesis (EMH) is the theory that market prices reflect available information, described in weak, semi-strong and strong forms.
Exposure
Exposure is the size of a position or its sensitivity to a named market factor, measured as market value, notional value, delta or scenario P&L.
F
1 termG
1 termH
1 termI
1 termK
2 termsKelly Criterion
The Kelly criterion is a sizing rule that chooses the fraction of capital to stake by maximizing expected logarithmic wealth under a specified payoff model.
Kurtosis and Excess Kurtosis
Kurtosis is the fourth standardized moment of a distribution, measuring tail weight, and excess kurtosis subtracts the normal distribution’s value of 3.
L
2 termsLog-Normal Distribution
A log-normal distribution is the distribution of a variable whose natural logarithm is normal, which keeps the variable above zero and makes it right-skewed.
Look-Ahead Bias
Look-ahead bias is the error of building or testing a trading rule with information that was not yet available when the simulated decision would have been made.
M
4 termsMarginal VaR
Marginal VaR is the change in portfolio Value at Risk from a small change in one position, the partial derivative of portfolio VaR with respect to that position size.
Minimum Variance Portfolio
The global minimum variance portfolio is the lowest-volatility mix available within a defined asset universe and set of weight constraints.
Modern Portfolio Theory
Modern Portfolio Theory is Harry Markowitz’s framework for combining assets by expected return, variance and correlation to maximize expected return at each risk level.
Monte Carlo Simulation
Monte Carlo simulation is a method that estimates prices, probabilities and risk measures by evaluating a model on many random samples and aggregating the results.
N
1 termO
2 termsOut-of-Sample Testing
Out-of-sample testing is the evaluation of a frozen trading strategy on reserved data that played no part in its design, to estimate how its rules generalize.
Overfitting
Overfitting is the tailoring of a model or trading strategy to sample-specific variation, so it fits its historical data well but performs worse on new observations.
P
1 termR
5 termsR-Squared
R-squared is the share of an outcome’s variation that a fitted regression explains, such as the portion of a fund’s return variation explained by its benchmark.
Random Walk
A random walk is a series built by adding independent random increments to a starting value, so the past path does not change the distribution of the next step.
Regression
Regression is a statistical method that fits an equation describing how a dependent variable changes with one or more independent variables.
Risk Parity
Risk parity is an allocation method that sets portfolio weights so each asset contributes an equal share of a chosen risk measure, such as portfolio volatility.
Risk-Reward Ratio
The risk-reward ratio compares a trade’s planned loss at invalidation with its planned gain at a target price.
S
6 termsSharpe Ratio
The Sharpe ratio is average return above the risk-free rate divided by the standard deviation of those excess returns, a measure of excess return per unit of volatility.
Skewness
Skewness is the standardized third moment of a return distribution, a measure of asymmetry whose sign shows whether cubed deviations below or above the mean dominate.
Sortino Ratio
The Sortino ratio is a risk-adjusted return measure that divides mean return above a target by downside deviation, so only returns below the target count as risk.
Standard Deviation
Standard deviation measures how widely observations spread around their mean in the same units as the data, and in finance it is the usual measure of return volatility.
Stress Testing (Scenario Analysis)
Stress testing is the revaluation of a portfolio under selected severe price, volatility, rate and liquidity scenarios to measure the gain or loss each one produces.
Survivorship Bias
Survivorship bias is the distortion created when a dataset keeps only the stocks, funds or strategies that still exist and omits those that closed, merged or delisted.
T
3 termsTracking Error
Tracking error is the standard deviation of a portfolio’s returns minus its benchmark’s returns, a measure of how much active return varies from period to period.
Trading Expectancy
Trading expectancy is the average profit or loss expected per trade from a strategy based on its win probability and average winning and losing outcomes.
Treynor Ratio
The Treynor ratio is a portfolio’s mean return above the risk-free rate divided by its beta, a measure of excess return per unit of market sensitivity.
V
2 termsValue at Risk (VaR)
VaR, or value at risk, is the portfolio loss that a chosen share of outcomes, such as 95%, stays at or below over a set horizon, a quantile of the loss distribution.
Variance
Variance is the average of the squared deviations of an asset’s returns from their mean, a statistical measure of dispersion whose square root is the standard deviation.
W
2 termsWalk-Forward Analysis
Walk-forward analysis is a backtesting method that repeatedly optimizes a strategy on past data and tests it on the next unseen segment, stepping forward through time.
Win Rate
Win rate is the percentage of completed trades that meet a consistently defined profitable outcome after applicable trading costs.
Z
2 termsZ-Score
A z-score is a standardized value that measures how many standard deviations an observation lies above or below the mean of its distribution.
Z-Test
A z-test is a hypothesis test that compares a sample mean or proportion with a hypothesized value using the standard normal distribution or a normal approximation.