PEG5: How long to hold
By Moe Chernick (moebruin)
EL SEGUNDO, CA (August 3, 1999) -- The number one question I get about the PEG screen is "How long do I hold the stocks?" This is a tough question to answer because the PEG works well for all tested periods. In fact, what length to use depends on each investor's personal needs. So let's look at the data, analyze it, and discuss what period will work best for you.
Below I've listed the average returns for various portfolios starting from January, 1987. I've also listed all listed enough disclaimers, I hope, to protect me from the lawyers out there.
Notes on the data listed below:
1) The 12-month October portfolio includes only the first nine months of the final year, i.e. through June 1999.
2) This backtest was done using old Value Line data sets and, while I feel it accurately portrays how the screen has done during the test period, I cannot guarantee its accuracy.
3) This data does not include trading costs.
Twelve-Month Hold Start month 5 Stocks 4 Stocks January 45.3% 46.5% April 19.4% 24.8% July 40.3% 44.8% October 26.0% 28.3% Six-Month Hold Start months 5 Stocks 4 Stocks Jan/July 48.7% 53.3% Apr/Oct 33.4% 39.1% Three-Month Hold Start months 5 Stocks 4 Stocks Jan/Apr/July/Oct 42.0% 49.3%From this data you can see some good news and bad news. The bad news is the annual returns for April and October, while good, are not nearly as good as from January and July. Why that is, I'm not sure, however, since the returns for April and October are still higher than Standard & Poor's 500 Index I do not believe this should be a major concern in using the screen. In fact, this year, portfolios started in April and October are doing fantastic. A 5-stock PEG started in October of 1998 is up 82.5% for the first nine months, while a 5-stock PEG started in April of 1999 is up an amazing 45.7% for the first 3 months.
Just about everything else is good news. The semiannual and quarterly returns look so good I get lots of questions about doing PEG as a monthly screen. Unfortunately, I don't have access to monthly data, so it is currently untested. But for the curious, I am experimenting with a monthly PEG within my own portfolio and, at least for the first eight months, it has done well -- but not any better then a semiannual port.
So what length is best for you? In a taxable account, this is really a tough question. I think it comes down to how you choose to interpret the data. If you believe that April and October suffer from a seasonality affect and that the returns will continue on par with the backtested results, then stick with an annual screen.
On the other hand, if you dismiss seasonality, then you might want to go with the shorter period. There is no known right or wrong way to go, although if you are not in a retirement account, the tax benefits of sticking with a one-year hold are substantial.
For IRAs and other tax-free accounts, you probably want to pick a shorter holding period, but the question is which one. The most popular version of the PEG these days is what is called the dual-semi. This version takes advantage of the strong returns of the 4-stock semiannual version.
With dual-semis, a person would buy the top 4 PEG stocks in the beginning of January (or on any other start date). Three months after their start date, they would buy the top 4 PEG stocks again but would continue to hold their initial PEG stocks. Six months after starting, the initial stocks would be sold and replaced by the current list. This process is then repeated every 3 months, selling the stocks you bought six months ago and replacing them with the stocks from the current list. The advantage of this plan is that it gives you eight stocks instead of four, and the historic returns show outstanding results with a very low standard deviation.
If you can't afford eight stocks, or you don't need the added diversity in your portfolio, the four-stock quarterly version will give you similar results with a little higher volatility.
Finally, the data shows that for all start dates and all holding periods, four-stock portfolios have better returns then five. If you need some diversity, go with the five, but if your portfolio all ready holds 12 or more stocks you probably want to stick with the top four stocks.
So while deciding yearly, semiannually, or quarterly may get your head spinning, if the past proves to be a reliable indicator of the future, you should do fine no matter which period you choose.