How Alto evaluates a strategy.

Every Alto strategy begins with a predefined ruleset. Those rules determine which markets qualify, which markets do not qualify, the period being evaluated, and the limits applied to individual positions.

The rules should be defined before performance is evaluated.

Where sufficient historical data exists, Alto applies the strategy’s rules to previous periods. A historical test may evaluate multiple seasons, tournaments, election cycles, economic releases, or other comparable periods depending on the strategy.

The available testing period varies by strategy and data availability.

Only markets meeting the strategy’s stated criteria are counted as qualifying observations. Alto displays the number of qualifying observations so users can distinguish between strategies supported by large samples and those with more limited historical data.

Strategies may include allocation limits intended to reduce the impact of any single market on the strategy. The applicable allocation framework is shown on the individual strategy page.

Alto does not publish every internal research or optimization parameter, but users should be able to understand the material rules governing the strategy they are viewing.

Whenever possible, historical performance is shown by independent periods rather than only as one cumulative number. For example, a baseball strategy tested across several seasons may display the return for each season separately.

This makes it easier to see whether historical performance was consistent or driven primarily by one period.

When historical periods are displayed independently, the simulated bankroll may be reset to the same starting value for each period. This prevents compounding across unrelated historical periods from making the results appear larger than they were.

When this methodology is used, Alto states it directly on the chart.

Settled positions are accounted for using their actual outcome. A losing position is not removed from the historical record simply because it performed poorly. A settled winning position is reflected at its applicable payout rather than being artificially represented as an open position.

Historical results may not perfectly reflect the prices, liquidity, fees, spreads, execution quality, or availability a user could experience in a live market. Where assumptions are used, Alto aims to identify them clearly.

Backtests are simulations of what a defined set of rules would have produced using the available historical data and assumptions. They are not forecasts.

A strategy that performed positively in previous periods can perform negatively in the future.