Most of the tests done on selecting the optimal holding period of a portfolio after considering the costs of trading - commission, spread, etc. - reveal that the best holding period is monthly. This may seem at first glance to be counterintuitive and nonsensical because if one selects a monthly five-stock portfolio which is rebalanced 12 times per year and if it is given that many of the newer picks will change and some or all of one's portfolio will turn over, how could that produce a CAGR higher than a quarterly, semi-annual, or annual rebalance periods. If instead of a monthly five-stock portfolio I select an annual five-stock portfolio, and if I pay $10 per transaction, both buy and sell, my commissions are significantly less for the annual rebalance (only five stocks bought one year before sale date and if any still remain on the screen, I can continue to hold them another year). But I may be required to turnover my entire portfolio twelve times during the same year(buys and sells) if I select the monthly rebalance. So how could one, over the long haul possibly do better selecting a monthly rebalance rather than a yearly?
Yet, despite these eventualities, backtesting strategies continue to reveal, test after test after test, that a monthly rebalance produces a greater after-cost return than longer holding dates.
If one considers why this happens, it may make some sense. When a particular screen's instructions happen to contain essentially a relative strength aspect, the start date for the monthly and the yearly will produce, using the same screen, the same five stocks. However, when one holds annually, one does not care what the current screen selections next month or quarter are because they just hold the portfolio for a year and then look to the screen for next year's stocks at the end of that holding period. So if a particular stock had a high RS on January 1, for some reason or another, it may go down and lose RS as time goes on. But for the monthly screen selection, that stock would most likely not continue to be a top-five pick since the screen instructions require a high RS over whatever period one chooses to select. Thus, an automatic stop-loss is built into the monthly screen and new stocks which are now gaining RS when compared with the market, come to the top of the list, and we buy those stocks when we run a screen list at the end of a month. In many ways, this monthly approach is a kind of "survival of the fittest" approach to investing. And when one removes the emotional aspects out of the picture by "mechanically" selecting stocks which rely, for the most part, on statistical probabilites revealed in various screen defintions tested over long periods of time, one has a much better chance to make money, even in a down market.
HT |
