The Ideal Holding Period

by Jim Stevens ([email protected])

Burlington, VT. (Oct. 12, 1998) -- If you wanted to hold stocks for the ideal holding period, what would it be? A day? A month? A year? Forever? All of the Foolish Workshop paper portfolios these days are tracked using a once a year update. This wasn't always the case -- in 1997 we tracked many of the growth models on monthly, quarterly and semiannual update schedules. The results were certainly different, and for the Unemotional Growth model it seemed brutally obvious that the higher hypothetical returns came from doing the trades every month. I don't recall how the year's returns ended up, but I do remember the monthly traded Relative Strength-IBD model was really cooking with gas early in 1997 also.

So why only annual now? If we only had to wish these real-time tracked portfolios into existence, we'd of course have every conceivable model for every time frame backtested, posted, tracked and analyzed nine ways to Sunday. All in a nifty Java applet with a soothing voice interface and advertising so subliminal you wouldn't even know it was there!

However, this is the state-of-the-art media of the future, folks, not television, so the rub comes from the fact that all this calculating takes incrementally more time and effort. Some choices have to be made as to what might be the most interesting and potentially useful models to track. It's also true that The Workshop was made to be interactive. Using the current Workshop Rankings and any of the free portfolio trackers available throughout cyberspace, Fools can pretty much track just about anything they want on their own.

The market-crushing returns of the monthly Unemotional Growth screen, combined with the popularity it gained with the publishing of Robert Sheard's successful book The Unemotional Investor, will probably ensure this model continues to be tracked. Other monthly models that don't have widely available backtests haven't been getting a lot of attention.

Aside from the extra work for Fools tracking them and publishing them online, what are the pros and cons of monthly vs. annual? I'll run through some of the obvious and not-so-obvious points. A meaningless disclosure: I was doing monthly updates in my personal investing for a year and a half but gave them up about a year ago mostly because I didn't care for the frequent trading.

Commissions: This one's simple and straightforward. If you are trading monthly you are obviously going to rack up many more commissions than if you are trading once a year. Some positions won't turn over every month, but in general, your broker will probably be into you for 7-10 times the price trading monthly instead of annually. If that means a difference of more than 1-2% of your investment, it's cause for concern.

Bid/Ask Spreads: A less obvious transaction-related cost, this is what the specialists or market makers cull off the top of every trade. It's not accounted for in the backtests, where trades hypothetically get filled at the closing price every time. Unlike commissions, this cost goes up just as fast as your portfolio's value. A number of Workshop readers have taken stabs at how to estimate this cost. I don't have any hard data, but it seems likely you'd get beat out of an eighth to a quarter for each round turn -- buying, and subsequently selling, a security position. Figuring stocks average $40-$50 a share, the costs would probably be less than a half of a percent for annual models... but here, again, 7-10 times that for monthly updates.

Taxes: Yuck! It's less of a concern in tax-deferred accounts, not a concern at all in tax-free accounts. This one really stacks the deck in favor of a 12-month holding period. No matter what your tax bracket, holding for a year will shave plenty from your tax bill. Just considering federal taxes, someone in the 28% bracket would have to see pre-tax monthly held returns about 11.1% higher then one-year held returns, just to break even. (e.g., 25% annual is equivalent to 27.78% monthly). The monthly methods start out even further behind the eight ball if you are in a higher bracket and/or get an additional state income tax break for long-term holding.

Backtesting: This is a less measurable advantage, and in my mind it favors the monthly models. Since a lot of the screens have backtests of twelve years or less, the effects of one or two stocks that happen to be wildly profitable or have desperately damaging returns in a single year can skew the average results more than any one or two stock picks in the backtest of a monthly traded model. When the one ranked stock is "skipped" and a particular model backtest improves, it makes me wonder whether the improvement isn't just evidence of the sample size and length of the test being inadequate.

Obviously, a case can be made for both holding strategies, depending on a lot of factors. There are undoubtedly even more differences and nuances, but we'll leave those for another column.

Stay Foolish!

Check out the latest file updates for the Workshop:
New Rankings | 1998 Returns | New Database