---
title: Backtest Panel
description: "Set benchmark, capital, historical dates, IS/OOS divisions and the dedicated Super out-of-sample reservation before running a backtest."
url: "https://portfolioboss.com/documentation/backtest-panel-guide/"
updated: "2026-11-07"
category: "Chapter 4 - Strategy Builder"
---

# Backtest Panel

The Backtest panel controls the historical period, comparison benchmark and sample divisions used to evaluate a strategy. Find it in the configuration area of **Strategy Builder**. Review these settings before interpreting any result.

**On this page**

- [Choose a benchmark and starting amount](#comparebenchmark)
- [Set the historical date range](#datefromparam)
- [Divide development data into IS and OOS](#insampleratio)
- [Reserve Super out-of-sample data](#super-out-of-sample)
- [Run and interpret a manual backtest](#run)

## Choose a benchmark and starting amount

**Compare to Benchmark** selects the reference instrument shown with the strategy’s results. Type its name or symbol and choose the intended result. Use a benchmark appropriate to the instruments and period you want to compare; a different benchmark changes the context of the comparison.

The benchmark’s chart button opens its price history in the Instrument report. Historical availability, including the benchmark’s available data, can affect the usable backtest range. A delisted instrument is not a valid benchmark selection.

**Initial Amount** is the starting capital assumption used by the backtest. It helps express the simulation in monetary terms; it does not fund a broker account or allocate money to a dashboard strategy. Account funding is managed in the [Trading Dashboard](https://portfolioboss.com/documentation/manage-strategy-funding-and-rebalancing/).

## Set the historical date range

| Control | Meaning |
| --- | --- |
| Date From → Earliest date possible | Starts as early as the strategy’s instruments, benchmark, cash equivalent and indicator lookback requirements allow. |
| Date From → Earliest date, ignoring rule instruments | Can extend the range by ignoring rules that depend on instruments until those instruments have sufficient data. The earlier segment therefore does not use exactly the same available rules as the later segment. |
| Date From / Date To → Set manually | Lets you specify a date using the field or calendar. Check the effective trading date shown by the application. |
| Date To → Latest date available | Uses the latest available data within the selected backtest mode, not necessarily today’s calendar date. |
| Date To → Latest date available (before SOOS) | Keeps the run before the reserved Super out-of-sample period. This is the mode shown in the current example. |

A shorter-than-expected date range can be caused by an instrument’s history or an indicator’s warm-up requirement. Read the effective dates displayed by Portfolio Boss instead of assuming the requested calendar range was used.

## Divide development data into IS and OOS

**In sample (IS)** periods are used to develop or tune a strategy. **Out of sample (OOS)** periods let you examine its behavior on other portions of historical data. Repeatedly changing a strategy in response to OOS results reduces their independence.

| Control | Effect |
| --- | --- |
| Number of In Sample Periods | Sets how many IS segments divide the development range. |
| Total In Sample Ratio | Sets the share assigned to IS; the remainder is assigned to OOS. |
| First Date Period is In Sample | Controls whether the earliest segment is IS. |
| Last Date Period is In Sample | Controls whether the latest segment is IS. |
| Periods bar | Shows the resulting date divisions. Hover over a segment to inspect its date range. |

Compare IS and OOS results in the reports, paying attention to the length and market conditions of each period. Similar historical behavior is useful evidence for review, but does not guarantee similar future returns or eliminate drawdowns. See the [Metrics panel guide](https://portfolioboss.com/documentation/metrics-panel-guide/) for metric definitions.

## Reserve Super out-of-sample data

[![Current Backtest panel highlighting Date To, In Sample Periods and the dedicated Super out-of-sample section and period bar.](https://portfolioboss.com/wp-content/uploads/2022/08/pb72-soos.png)](https://portfolioboss.com/wp-content/uploads/2022/08/pb72-soos.png)
The dedicated Super out-of-sample controls reserve a final period. The example settings are not a recommendation or a factory-default claim. [Click to enlarge.](https://portfolioboss.com/wp-content/uploads/2022/08/pb72-soos.png)

**Super out-of-sample (SOOS)** is a separately reserved period for a later review after development and selection decisions. The current panel has a dedicated **Super out of sample period** control. Set the reserved duration and verify the dates in the period bar; keep **Date To** in the before-SOOS mode while working with the development period.

Older guide instructions used only a manually shortened Date To as a workaround for holding data back. Manual dates remain useful for a historical cutoff, but the dedicated SOOS controls make the reserved period explicit in the current interface.

Decide when and why to examine the reserved period before looking at it. Once you use its results to tune or select again, that period is no longer an untouched final check.

## Run and interpret a manual backtest

1. Confirm the strategy, date range, data status and sample divisions.
2. Choose the manual **Start** action in Strategy Builder. If the confirmation dialog appears, add a memo and review it before continuing.
3. Wait for the run to finish, then review the reports and [Manual Results](https://portfolioboss.com/documentation/manual-results-tab/).

If data is still downloading, inspect [Sync Details](https://portfolioboss.com/documentation/sync-details-search-results-panel/). A run made with incomplete data can give incomplete or misleading results. The confirmation preference can be changed in [Settings → Miscellaneous](https://portfolioboss.com/documentation/miscellaneous-tab-guide/).

For historical signal investigation, a manual Date To acts as the calculation cutoff. Set the relevant historical trading date, rerun the test and inspect Trade Signals and the Instrument chart. Historical model signals are not evidence of fills in a live broker account.