The Art Of Back Testing (Part 1)


Back testing is the simple process of creating a sampling based on certain criteria and seeing what the outcome would have been. We are looking to see what may have the highest probability of happening.

For example some individuals would like to see whether entering a trade at any time after the first hour of trading would yield any different results. Others would like to back test different stop loss results based on other types of criteria.

Back testing makes sense since we are shaving off years of trading with unreliable and inconsistent results preventing us from losing monies.

Back testing is designed to create a measure of confidence so that we can approach trading with measurable results.

How do we Begin a Back Test?

The very first think we need to identify is the time frame that we are seeking to be in a trade.

For example a person who is looking to be in a trade for several months does not need to look at a 30-minute bar and vice versa. The person who is looking to be in a trade from one day to the next will not need to look at a monthly chart.

So it all begins with a time frame.

A huge part of this process is the ability to analyze data so you need to keep good records.

Create a back testing journal.

What we are looking for are patterns in something happening more often than not and these patterns are not necessarily technical in nature.

For example I used to enter my trades between 9:30AM EST to 10:00AM EST since it was convenient for me to enter the trades at that time since I usually took a coffee break from work at that time only to find at the end of the day I usually paid the highest price that option went for that day.

What I did was realize was that there are times in the US when certain key economic data (that relates to housing starts, Consumer Confidence, etc.) can take place at 10:00AM EST and these events can impact our position.

So these days I prefer to place trades after 10:30AM EST.

A good idea is to locate all of the FOMC, Non-farm payroll and Earnings dates so that we can see that if a trade did not work out it may have been impacted by anyone of these key events.

In that case you would have either stayed out or created a hedged position prior to the event and have been negatively impacted.

The Back Testing Tools

We need to have access to reliable price quotes and price charts. The key word here is “reliable.”

We need to realize that when we use a free tool the data may not be totally accurate so it would be something that may be useful to use if direction is what we are seeking.

However when we are seeking precise price quotes then we will need to invest in a tool with accurate quotes.

When it comes to option trading it is virtually impossible to recreate the exact option quote since there are many.

There are option quotes for the every month and there are option quotes for every strike price.

Then there is the Earnings environment which can cause the volatility to go hay wire however it does not make itself manifest in a historical chart.

So we will need to make concessions here and go with the “average implied volatility.”

At least we will have something we can measure.

We will continue with the art of back testing in the next post… stay tuned!