Quantitative investing – increasing the odds of success

Time for part 3 of my Quantitative Investing posts. Please refer to  the introduction post and the getting started post. In this post I want to discuss portfolio monitoring rules and how to increase the odds of success with quant strategies.

First, lets talk about portfolio monitoring rules. The quant strategies I discuss involve investing in a portfolio of usually 25 (max 50) individual stocks on an equal weight basis and holding for a year. A common question is, ‘do I hold for a year no matter what?’. No there are some basic rules. These rules are for risk reduction, thereby working to increase your odds of success with these strategies. After all significant events can take place to individual names during your one year holding period. Here are the rules from What Works on Wall Street (my go to quant reference). Page 56 if you have the book.

1. If the company fails to verify its numbers as required by Sarbanes-Oxley, we replace the stock in the portfolio. (basically, your’re looking for an accounting scandal, failure to file, etc..)
2. If a company is charged with fraud by the federal government, we replace the stock in the portfolio.
3. If a company restates its numbers, such that it would not have qualified at the time of purchase, we replace the stock. (easier than it sounds – you just need to keep an excel spreadsheet of your results)
4. If a company received a takeover offer from a third party, we replace the stock if the price of the stock moves within 95% of the offer price. (this happened to me this year. OMX received a buy out offer from Staples)
5. If  a company drops by 50% from the point of purchase and is in the bottom 10% of all stocks for the previous 12 months of price performance, replace the stock. (this acts like a maximum stop loss. For example, with a 25 stock portfolio this would limit the loss to 1% of the total portfolio)
6. If a stock from a dividend strategy cuts its dividend by 50% or more it is replaced in the portfolio.

That’s it. You don’t make trades otherwise during the 12 month holding period. Note that these rules are meant to reduce the risk from investing in individual names. They do nothing to protect from general market declines.

The best way to increase your odds of success is to invest in strategies with the highest base-rates. One of the hardest parts of quantitative investing is that even the best strategies can under perform for long periods of time. Base-rates are a way to keep focused on the long term. Humans suffer from something called base rate fallacy. Simply, we let recent history influence our views too much when we should stay focused on long term. For example, we know that over the long term value outperforms but there are periods when it doesn’t, when growth trumps value. We can measure this by using rolling returns and calculating the percentage of periods when a strategy beats the benchmark. Below is a comparison of some selected strategies and the S&P500.

The table shows the percentage of time rolling 1,3,5,7 and 10 year returns beat the benchmark for various strategies. For example, using 3 year rolling returns the S&P500 beat its benchmark (the total stock market) in 48% of all the 3 year time periods in the sample. 1965 to 1967 is one 3 year period, the next one is 1966 to 1968, the one after that is 1967 to 1969, etc… In other words over a 3 year period you are likely to under perform over half of the time. Maybe this is why traditional buy and hold is so hard? But if you chose a low P/E stock strategy you beat the benchmark 85% of the time. And using Trending Value you would beat the bench mark 99% of the time with a 3 year holding period. High base rates increase your chance of success and more importantly increase the chance you will stick with the strategy. But base rates aren’t everything either. Take dividend stocks or the combined consumer staple/utilities portfolio. Their base rates aren’t the best (still quite good) but they have other characteristics that make them worthy of consideration. For dividend stocks its the chance to live off the dividends and for consumer staples/utilities are the fact that they are the two lowest risk sectors of the market.

In summary, there are some simple portfolio monitoring rules that increase your odds of success with quant strategies. Also, investing in strategies with high base-rates calculated over the long term increases your odds of success and more importantly the chances you will end up sticking with a given strategy. In the next post I’ll start diving into four strategies I think are worthy of consideration for more conservative investors and retirees in particular.