Definition
A tearsheet is a one-page report that summarises a strategy's risk and return profile in a fixed, comparable format. The convention originated in equity research — the term refers to summary pages torn from a binder — and now denotes any compact dashboard of headline metrics produced from a returns or P&L series.
The standard surface includes cumulative and annualised return, the Sharpe and Sortino ratios, maximum drawdown and recovery time, monthly return heatmap, exposure decomposition by sector or factor, turnover, and a sample equity curve. The point is comparability: a reader who has seen a hundred tearsheets can absorb the essentials of a new one in under a minute.
Why it matters
How it works
Producing a tearsheet starts from a clean daily returns series — or a P&L series and an associated capital base — and applies a fixed library of summary statistics. Python libraries such as pyfolio and quantstats package the conventional layout: top-line numbers in a header table, the equity curve and benchmark beneath, drawdown periods plotted as shaded regions, a monthly returns heatmap, factor exposures from a Fama-French style regression, and rolling Sharpe and beta to detect regime shifts.
The most useful tearsheets follow three discipline rules. They are generated from the same code path used in production, so a model in research and the model in live trading share metric definitions. They include the benchmark on every chart, so an investor can read excess return at a glance. And they distinguish in-sample from out-of-sample windows visually — typically by a vertical line at the cutover date — so a reader cannot accidentally take overfit numbers for tested ones. A tearsheet that pools the two without separation is selling rather than reporting.