introduction.html

<p>Algorithms usually perform differently between backtesting and live trading over the same time period. Backtests are simulations where we model reality as close as possible, but the modeling isn't always perfect. To measure the performance differences, we run an out-of-sample (OSS) backtest in parallel to all of your live trading deployments. The <a href="/docs/v2/cloud-platform/live-trading/results">live results page</a> displays the live equity curve and the OOS backtest equity curve of your algorithms.</p>

<img class="img-responsive" src="https://cdn.quantconnect.com/i/tu/live-strategy-equity.png" alt="The live and OSS backtest equity curves of an Alpha">

<p>If your algorithm is perfectly reconciled, it has an exact overlap between its live and OOS backtest equity curves. Deviations mean that the performance of your algorithm has differed between the two execution modes. Several factors can contribute to the deviations.</p>

<p>To generate the OOS reconciliation curve for a live deployment and overlay live vs. backtest equity and order fills from the Research Environment, see <a href="/docs/v2/research-environment/meta-analysis/live-reconciliation">Live Reconciliation</a>.</p>