The MarketDelta Blog

(This is a continuation in our blog series on creating a trading system from start to finish. Want to see more? Click here for our main RTL support page, which links to all the articles in this series and many more tutorials. Questions? Click here for the RTL Community Forum where you can get help on your programming.)

Realization is a simulation of real trading. It is not just backtesting, it is the actual trading process re-enacted for the most accurate picture of a strategy’s past results.

We now have a trading system based on the Market Profile’s Initial Balance range that has backtested well in many different scenarios. Our current incarnation allows for two exits to our trade, so we can “scale out” and move our stop-loss levels to break-even (the entry price) after the first profit is taken.

In our last article, we performed the backtest for the same 250-day period as in all other tests and, after $5 round-turn commissions, found a net profit of $3,970 for 2 contracts when using a 2-point stop, 5-point total profit, and a 2-point scale-out.

Using optimization, we can see that we would have increased profit substantially if we instead went for 15 point profits, a 9-point stop and a first scale of 4 points.

Here’s the problem

It is fun to know today what would have worked in the past. But last year, if we had started trading this system, we would not have had the luxury of optimizing it from price action that would occur in the future. We only know what has already happened, not what we would have done with knowledge we didn’t yet have. Just because we know now that the prior year worked well with 15 point profit targets, does not mean that we would have actually used this in our strategy at that time — because we didn’t know that this was going to be optimal in the months that followed.

It is for this reason that optimization cannot show you potential profits for the future, even if it can help you structure a strategy and give you ideas.

Realization

However, MarketDelta does offer a featured called “Realization” that accurately captures the real results you would have enjoyed in the past, if you optimized, traded off that optimization, then optimized again later, and continued to adjust your strategy based on recent market conditions. In this way, you can see the effects of optimization on real-time performance.

Realization works as follows… You set an “Optimization period” — this is the amount of time to include in each optimization. Then you set a “Realization period” — this is how often you optimize based on prior data and alter your strategy based on the results of that optimization.

MarketDelta runs an optimization, then performs a simple backtest using these optimization results for the period following the optimization. Then at the end of that period (the realization period), it runs another optimization and uses those results for the next backtest. This process continues until all the data is tested.

It effectively captures how you might have traded a system and only uses optimization results for past data when calculating the actual backtests on future data.

To set it up, just press the Realize button in the Optimization window.

Make your settings, then press Realize. For these settings, we get these results:

This is a real result that could have occurred without knowing the prices of the future. Still profitable, but the reality is obvious: not as profitable as a straight-up optimization for the entire period. If your system can perform well in the Realization phase, you have many reasons to be more confident.

If we alter our realization for different trading behavior, we would alter our profits over the 250-day period:

These settings include more data in each optimization along the way, and then re-optimize every 30 days. If you traded this system, you would not change your profit targets and stops as often, and these levels would be based on more price history. Here are the results for this realization:

Somewhat comparable.

Now let’s try a test for more aggressive changes over time to the strategy. Suppose we optimize every 10 days based on the most recent 10 days, and then use those results in trading for the following 10 days, before repeating this procedure:

Keep in mind that your strategy is likely to change much more dramatically every 10 days since it is using optimal results from a short testing period for each change.

Here are the results:

Here are our best Realization results yet; so, at least for this strategy, frequent optimization and strategy changes proved profitable over the long term.

This would not necessarily be the case with other strategies.