Interpreting Walk Forward Results

Walk Forward Pro is a feature-rich application and it can take a short while for new users to understand its full capabilities. This guide provides a quick-start to interpreting the walk forward results screens.

Overview

As you already know, a WFO (Walk Forward Optimization) consists of multiple stages. This section explains how the information from each stage relates to the data shown in the application’s user interface.

Step through the Walk Forward Analysis Process

Let’s take it step-by-step. During Stage 1, all of the parameter combinations for the trading system that were requested by the user on the ‘WFA Setup’ screen are tested. The parameter combination that is considered to be the most robust from the optimization phase, is then used in the first out-of-sample walk forward test. The parameters that were chosen can clearly be seen in the first row of the table on the ‘Out-of-Sample Walk Forward Tests’ screen as shown below. Note that the rules used to select the parameters to use in the walk forward phase will depend on the settings selected for the ‘Selection Performance Metric’ and the ‘Selection Basis’ on the ‘WFA Setup’ tab.

Most robust optimization parameters used in walk forward test
Figure 1 - Most robust optimization parameters used in walk forward test

The results from this first out-of-sample walk forward test (using the most robust parameters) can be seen in the first row of the table on the ‘Out-of-Sample Walk Forward Tests’ screen, and the equity curve can be seen as the first section of the chart, as shown below.

First stage walk forward phase results
Figure 2 - First stage walk forward phase results

Progressing to stage 2 of the Walk Forward Analysis Process

Following completion of the walk forward test, Walk Forward Pro automatically moves onto Stage 2. In line with best practice, the optimization phase of Stage 2 moves forward so that the end of the optimization period is the same as the end of the previous walk forward test period.

All combinations of the parameters requested by the user are once again tested during this second optimization phase.

Optimization window moved forward
Figure 3 - Optimization window moved forward

Once the optimization has completed, the most robust parameter combination from optimization 2 is then used as the parameters for the next out-of-sample walk forward test.

This illustrates one of the major benefits of using the WFO process:

Each walk forward test is undertaken using the parameters that performed best against ‘recent’ market price data. This means that as market dynamics and market personality change over time, WFO is able to keep your trading system in tune with this by selecting the parameters that work best in that type of market.

Key take-away fact
Most robust stage 2 optimization parameters used in next walk forward test
Figure 4 - Most robust stage 2 optimization parameters used in next walk forward test

The parameters that were chosen from Optimization 2 can be seen highlighted above. In this case they were the same as those chosen in Stage 1, but this will not always be the case.

The performance of the chosen parameters against the out-of-sample data can be seen in the second row of the table and in the second section of the equity chart as shown below.

Stage 2 walk forward phase results
Figure 5 - Stage 2 walk forward phase results

Completion if the entire Walk Forward Analysis Process

The WFO process continues until all of the stages that were requested have completed. The chart now represents the full cumulative out-of-sample results.

All walk forward stages in WFA (Walk Forward Analysis)
Figure 6 - All walk forward stages in WFA (Walk Forward Analysis)

Pre-live Optimization

If the user requested a ‘Pre-Live Optimization’ to be run from the ‘WFA Setup’ tab, this will now automatically start. Pre-Live Optimizations are used to identify what the best performing parameters are in the period of time immediately prior to the end of the overall test period. This makes these parameters ideal for live trading since they are based on the most recent dynamics in the market. These parameters can be seen on the ‘Pre-Live Optimization’ screen as shown below.

Pre-live results of WFA (Walk Forward Analysis)
Figure 7 - Pre-live results of WFA (Walk Forward Analysis)

The Pre-Live Optimization screen shows the parameters that performed ‘best’ during the pre-live optimization. This does NOT necessarily mean they are ‘good enough’ to trade in a live, real money account, and nor do they give any indication of whether the user’s trading system being tested is fit for purpose or not. The user will need to form their own judgement from the other results screens to determine if the their system is ready to trade on a live account or not.

Key take-away fact

Like what you've read today? Then please consider sharing

Subscribe to our newsletter

By subscribing you are giving your consent to send you emails in accordance with our privacy policy

About Us

We are passionate about algorithmic trading, and about helping other algorithmic traders reach their full potential.

We help traders to develop robust trading systems that deliver results in live accounts.

Contact

Built in Yorkshire, UK
Proudly serving Europe and the World