Tearsheet

A tearsheet condenses a strategy's entire simulated life into one report. A good one answers, at a glance: did it make money, how bumpy was the ride, and would I have survived holding it?

What's on it

The equity curve with its underwater (drawdown) plot. Total and annualized returns, versus buy-and-hold. The risk numbers: Sharpe ratio, max drawdown, volatility. Trade statistics: count, win rate, average win and loss, expectancy, largest winner and loser, average holding time. And the costs actually charged, fees and slippage, so gross-vs-net is visible rather than hidden.

How to read one critically

Start with net-of-costs numbers, not gross. Check the trade count: 30 trades prove little, 500 give the statistics some weight. Look at when the money was made; one lucky month carrying the whole result is a warning. And treat a beautiful in-sample tearsheet as a hypothesis to be tested with forward-walk analysis, not as a result. On AlphaProve, every completed backtest renders its tearsheet automatically, with trades plotted on the chart they came from.

On AlphaProve

Beyond the core numbers, the report carries the tail-risk metrics that decide whether a smooth-looking strategy is quietly short volatility: the Omega ratio, tail ratio, 95% VaR and CVaR, and the Ulcer index for drawdown pain over time. Once a run completes, the AI reads the whole report and writes a plain-language analysis of it — what worked, where risk concentrated, what to check next — so the page is not a wall of statistics you have to interpret alone.