Portfolio Backtesting – How to Test Trading Strategies
Growth Stock Portfolio Backtesting Strategy
Strategy Testing on a Basket of Stocks
Portfolio backtesting allows you to test basket trading strategies to see whether your idea of making money in the stock market has any historical merit or not. How do you come up with a portfolio backtesting strategy? What are some methods to test it? What are some free online resources?
What is Portfolio Backtesting?
Portfolio backtesting is a method test a strategy, or many strategies at once, on a large universe of stocks. As an example, you may want to buy all the companies that have high revenue and earnings growth but a low price-to-earnings ratio. Is this a sound strategy or is will is result in disaster? You can use portfolio backtesting software to find out. This quantitative strategy is run across broad markets as you hold a basket of stocks that meet the buying criteria.
Fundamental and Technical Analysis
What goes into your strategy? You must consider what factors make up a winning company. Most investors will rely heavily on fundamental analysis - which involves such things as cash flow, net asset value, dividends, revenue, and earnings. Traders, or even investors looking to gain a favorable entry and exit, will use technical analysis which involves analyzing share price and the volume traded. A good portfolio backtesting service will allow you to build trading strategies using both fundamental and technical analysis filters.
Example of Strategy Testing
Let’s run through a quick example of how we can test a strategy for historical gain. Imagine that I believe the following to be true:
- Growth stocks generally outperform the broad market
- Sales growth is the important backbone to earnings growth
- Growth should be evident when comparing quarterly results, trailing 12 months, and 5 year averages
From here I can quickly build filters that will simulate buying all the companies that fit the above criteria. I need to decide on how long I will hold the companies before re-balancing my portfolio. In this instance I will choose 3 months. After running these stock selelction criteria through my online portfolio backtesting software over the past 3 years, I get the following results:
- Growth strategy delivers 26.26% compound annual growth rate
- S&P 500 delivers 13.81% compound annual growth rate
Checking for Robustness
It would initially appear that my strategy is wildly successful. However, this problem is well-known to portfolio backtesting software – it can cherry pick the results. This can occur when you ‘fortunately’ pick the right entry date and stumble across winning stocks when re-balancing.
But what if you purchased a week sooner? A week later? Two months later? Thus we have a need to check our strategy for robustness. Robustness checking can be performed by testing the strategy with variety of entry points. You buy and hold for a specified period using your strategy at random entry points. The average gain or loss can be compared to see whether this is a strong strategy or a fluke.
Using over 150 random entry points over the past 3 years, while holding for 3 months at a time, the growth strategy beats the market by 7% per year on average. This is far below the 12-13% annual excess gain we tested earlier and the robustness check provides us with a more realistic statistic.
Free Online Portfolio Backtesting Software
Most portfolio backtesting sites cost an arm and a leg. Expect to pay up to a thousand or more per month if you are getting institutional grade software or hundreds per month if getting all the features with Zacks Research Wizard.
One service called Portfolio123, has a free membership level that lets you backtest one strategy at a time with up to one year of trailing data. If you use the above link you can also get any membership level for an extra long trial period of 45 days, after which you can downgrade to the 100% free membership if you so choose.
Why do I like Portfolio123 so much?
- No survivorship bias. Stocks that got delisted are still in the database to keep it honest.
- No look ahead bias. You only get access to information that was available at the time of screening.
- Over 550 programmable fundamental and technical filters.
- Stock screens that mimic investing masters like Warren Buffett, Peter Lynch, Benjamin Graham, and Joel Greenblatt.
- Institutional grade software of the highest quality but a fraction of the cost of the competition (membership levels start at free but I use the $50/month)
- Other tools such as market timing, ETF portfolio backtesting, using ranking systems, simulated trading environments, and much more.
Of course, many individual investors are simply looking for 'hedge fund' strategies that others have already worked tirelessly on and would prefer to replicate such a fund in their own account if a simple method was available to do so. This would by-pass any net-worth restriction that traditional hedge funds have, eliminate the ridiculous fees, give total control to the investor and be 100% transparent. If that interests you, I recommend you head over to Portfolio Cafe to see why I decided to hire on as an equity model builder with this new type of 'investment fund' - if you will.
If you are serious about testing out trading strategies in the stock market, or if you are simply curious how good a popular trading system is – you owe it to yourself to try out a portfolio backtesting service.