Find and manage backtests
Open Strategies → Backtests to list your backtests. The page opens on your results, newest first. Select a backtest to open its results.
Above the list:
- The usage pill shows this week's SimTime and your storage. Select it to open the Usage page.
- The Describe a strategy idea box sends an idea to the AI Editor. See Create and run a backtest.
If you have no backtests yet, the page offers AI Editor, Simple and Templated to start one. It also lists featured runs made from MesoSim's built-in templates. Featured runs belong to MesoSim and can't be deleted.
Follow running backtests
Your queued and running backtests are shown in one line above the results, with the number of running and queued runs and the progress of each. Expand the line to list them as the first rows.
Select Cancel… in this line to cancel all active backtests, only the queued ones, or the ones that match your search. To cancel one backtest, use Cancel job in its row menu or on its page.
See Run states and recovery for what each state means.
Search and filter
Combine these filters to find the runs worth a closer look:
- State: All, Active, Finished, Failed or Cancelled, each with its count.
- Search: type part of a name or strategy. The suggestions list your strategies; pick one or more to show only their runs. Press Enter to search names and strategies for the text. Press
/to jump to the search box. - Symbol: show the runs of one symbol.
- Date: filter by the date a backtest was created or finished, or by its backtest period. Use a preset, such as Last 7 days or Older than 6 months, select bars in the chart, or enter a range.
- + Condition: filter by a metric, for example
Sharpe ≥ 1.5orCAGR < 0. A backtest must match all conditions.
Conditions can use CAGR, Total Return, Sharpe, Sortino, Calmar, Profit Factor, Max Drawdown from Peak, Max Drawdown on Initial Capital, Annualized Volatility, Trade Count, Win Count, Loss Count and Avg Days In Trade. The quick picks find losing runs (CAGR < 0), runs with a negative Sharpe, runs with a maximum drawdown from peak below −50%, and runs with fewer than 10 trades. See Quantitative Metrics for the definitions.
Each active filter is shown as a chip, with the number of matching backtests and the storage they take. Clear all removes the filters.
Choose columns
By default, the list shows the state, name, created date, strategy, symbol, backtest start and end, CAGR, Max Drawdown on Initial Capital, Sharpe, Alpha and Beta. Narrow screens show fewer columns.
Use the column picker to add other columns, such as Started, Finished, SimTime, Storage or Max Drawdown from Peak. Reset layout restores the defaults.
Select backtests
Select the check box of a row to select it. You can select only your own backtests.
The check box in the header selects the visible page. To include every backtest that matches your filters, select Select all N matching. The selection bar then shows the number of selected backtests and their size. It offers Delete, and Cancel when the selection includes queued or running backtests.
Cancelling 50 or more backtests, or deleting 50 or more permanently, asks you to type CANCEL or DELETE to confirm.
Delete backtests
Delete from the selection bar, from a row's menu, or with More → Delete on the backtest page. Queued and running backtests must be cancelled first.
The dialog offers two choices and shows how much storage each one frees:
| Choice | Keeps | Deletes |
|---|---|---|
| Keep summaries, remove results (recommended) | Name, strategy and metrics. The backtest stays in the list as Results removed. | Charts, tearsheet, trades, event log and other reports. Portfolios use these results too. |
| Delete permanently | Nothing | The backtest and all its data. This can't be undone. |
After results are removed, a shared link to the backtest shows only its summary. After a permanent delete, the link stops working.
Deletion continues in the background, and you can leave the page. Deletion history lists earlier deletions.
Run a backtest again
A backtest with removed results keeps its page, but its result tabs are hidden. Select Run again to start a new backtest with the same strategy. It calculates new results with the current MesoSim version, so they can differ from the removed ones. The new run uses SimTime. If you need the original data, export it before you remove results. On a backtest someone shared with you, Run a copy runs it in your account.
Free up space
Free up space on the Usage page suggests backtests to clean up:
- failed and cancelled backtests
- repeated runs: the same strategy run again with identical settings, symbol and period. The newest run of each is kept.
- backtests finished more than 6 months ago
- the largest finished backtests
Each suggestion opens the matching backtests on this page, so you can review them before deleting.