Getting Started with Backtesting

Getting Started with Backtesting

You can’t determine whether mechanical entry and exit signals are viable for profitable trading unless you test it over a broad range of trading vehicles through different market environments. Recent price action can create bias and isolated examples are not enough to make decisions from. You have to start with your own watch list and signals and let the results of the data tell you what works. The goal of back testing is to see what signals will create bigger profits than losses. They should keep you in trends and get you out early enough to lock in profits. They should also get you out of losing trades quick enough to keep you losses small.

The first step in your process to create your own back tested moving average systems is your watch list. You need to create a diversified list of the markets, stocks, and ETFs that you want to back test. These should be things that you trade regularly along with other markets that you want to add. Your list should have enough variety to give you the opportunity for signals during different market cycles. This list includes index ETFs along with the sector ETFs of the S&P 500. This is a great starting point with different market caps and different sectors. While these are all equity holdings it is a diversified list for the stock market and a great place to start. You can add other markets that you plan on trading.

The next step is to choose the moving average signals that you want to back test on your watchlist. It is good to have a diversified list of signals to back test inside your own timeframe or even signals that diversify across timeframes. This list contains the 5day /20 day ema that can capture market momentum. The 10 day / 50 day ema can profit from both market swings from lows to highs as well as trend trades higher.

The single longer term moving averages can act as lines in the sand between bull markets and bear markets. Price crossing under a longer term moving average can market the end of a bull market and the beginning of a correction. A longer term moving average can also give investors a signals to get back into a market as price crosses back over signaling the end of a down trend.

The shorter term moving averages crossing over the longer term moving averages help to filter out the noise that happens around the longer term moving averages as price can give many false signals at turning points.

It is important to have diversity in the signals that you will be backtesting. You will learn a lot about what does and does not work in the market. You will also see how the patterns of price action can be different across indexes and sectors.

You will need to back test your signals on your watchlist across a meaningful time period to see how will your signals did during up trends, down trends, panics, crashes, and bubbles.

The January 2000 through 2018 time period is a back test through the entire 21st century and captures the end of the internet bubble, the bull market from 2003-2007, the financial panic of 2008, and after the 2009 bottom the bull market from 2009 to 2018. This will give you a very diverse look at how your signals did through multiple extreme market environments.

Be aware that some stocks and ETFs did not exist through all of this time period so consider if the time period is diverse enough to make judgments on the system performance.

For a fast path to learning about backtesting my eCourse is available here: Getting Started with Backtesting