We have normally associated the word back-testing with some scientific process or dealing with mountains of data. While both these statements are partially correct, their interpretation is not. Firstly, back-testing is extremely useful when it comes to investment and trading. Secondly, while it is true that the process of back-testing entails mountains of data working in the background and complex calculations, websites offer you the facility to back test your strategy with real historical data. How to do backtesting is something that has now become available even to small investors. Let us understand what is backtesting in trading and look at some key back tested trading strategies.
What do we understand by back-testing investment strategy?
Let us assume that you plan to invest in a stock at the current market price for a period of 5 years. Obviously, the next 5 years could include a lot of events over which you may have little or no control. What you are more interested in is the comparison of three alternate strategies. In the first approach you just buy the stock in its entire quantity today and hold for five years. In the second approach, you adopt a SIP kind of approach wherein you allocate your entire investment in the stock in a phased manner over the next 1 year and then hold on to it for the same period. In the third approach, you are a little more dynamic and you time your purchase quantities based on P/E range of the stock. The big challenge is to compare these three alternatives. Evaluating these 3 strategies over the next 5 years will be an exercise in uncertainty and so the best that you can do is to back test these 3 strategies. Here is how to go about it.
How to go about back testing the 3 investment approaches?
In back testing what you do is to run the strategy on live data in a retrospective way rather than a prospective way. Let us understand this further. You are exploring these 3 alternatives in 2018 and hope to see the results in 2023. Instead, you assume that you had adopted these 3 approaches in 2013 and then use real market data to evaluate the outcome of these 3 alternatives as of 2018. The back testing of the 3 approaches on live data will give you a clear picture of which strategy would have worked better in the last 5 years. Of course, then there is the subjective judgement that you need to apply before adopting the approach but you, at least, have the data in front of you to make the decision.
How can investors benefit by adopting the back-testing approach?
One can argue that back testing entails past data and therefore it is not compatible with the future. The last five years may have been a bull market and the next five years could be a bear market; then how do you treat the back testing as reliable. The answer is that there are 3 distinct advantages that emanate from the back testing of your strategy:
Firstly, back testing helps you to measure your thinking and your strategy against the market reality. For all you know, the best approach may be an approach that is entirely different from the 3 that you have shortlisted. These kinds of insights are available from back testing on live data.
Secondly, back testing puts your strategies in perspective. You may run the live data on these 3 approaches and eventually realize that all the 3 strategies have actually underperformed the index. That means; you would have been better off just staying invested in an index fund rather than going through all this trouble.
Thirdly, back testing results can be made a lot more useful with further refinements. For example, you can stress test you back data for higher volatility, lower volatility, lower valuations, higher valuations etc. These refinements will give you a much clearer picture of how your strategies could perform under different market conditions.
Finally, remember that this is back data after all!
Back testing is just a data based approach to decision making. Today it is possible to do a back test of your strategy through websites but as an investor you need to be aware of the shortcomings of the same. If the underlying market structure changes drastically in 5 years then back testing may not work (for e.g. pre-liberalization versus post-liberalization). Secondly, you are not sure of the actual software process that goes on behind the black box like back-testing model. The assumptions made and the data considered matters a lot. Lastly, back testing may lull investors into a false sense of complacency. Of course, that can be overcome by consciously treating the back testing outcome as just another set of inputs for your decision making.
In a nutshell, the facility to back test is a big addition to your investment approach. Use it intensely but be wary of the limitations of back testing and apply accordingly.