Foolish Workshop
Rebalancing Screens:
Should I or Shouldn't I?

By Moe Chernick (Workshop)

EL SEGUNDO, CA (September 21, 1999) -- One way that I have tried to enhance my returns is by "running winners" in my frequently traded screens. Running winners means that if a stock I already own is still "on the list" when it comes time to renew a screen, I hold onto all those shares instead of selling (or buying) some shares to make that position the same size as the new ones.

The theory behind running winners is based on two possible advantages:

1) It reduces trading costs
2) It can bolster returns by giving you larger positions in your hot stocks (i.e., stocks that have done the best for you).

Surprisingly, to me at least, a study that I have just completed with the PEG (traded quarterly) and the RS-26 week (traded monthly) screens shows that running winners does not improve your average returns.

The effect on the PEG 4 is slight. From 1987-1998, that screen was up an average of 48.1% with a CAGR of 42.1% using the data from my original backtest. If I had run the winners, the average decreases to 47.6% with a CAGR of 41.3%.

The differences were slightly greater when done with the RS-26 week screen. From 1987-1998, this screen had an average return of 48.1% with a CAGR of 43.6% using the data from Brian Malcolm's backtest. Changing the backtest to run the winners decreased the average return to 47.2% and the CAGR to 42.2%.

For each year in the study, I compared the returns and noted the difference between the traditional screen and a "running your winners" version. The following table shows the distribution of those yearly differences.

Difference    PEG 4         RS-26  
    0-2%        6             7
    2-5%        1             2
   5-10%        4             1
  10-20%        1             2
over 20%        0             0
The differences aren't great. The table shows that the difference was within 2% about half the time and was never greater than 20%. Only once was the difference over 10% for the PEG 4. In that year, 1996, running your winners actually paid off, with the running version returning 117% versus 100% for the standard version. However, three of the four times where the difference was between 5% and 10%, the return was in favor of rebalancing.

With the RS-26 week screen, only three times out of 12 was the difference greater than 5%. Those years were as follows:
Year     Running Winners      Rebalancing
1993          39.1%              52.3%
1996          40.3%              46.4%
1998         164.8%             149.0%
I think that from this data one can safely conclude that, over the long haul, it is slightly better to rebalance than to run your winners. Usually, the difference in returns will be insignificant. Occasionally the difference may be significant, but not consistently enough to lose sleep over.

Trading costs can be a different story. Commissions can eat into a smaller portfolio that is traded frequently. The slight differences in the long-term averages above lead me to conclude that the advantage of reducing trading costs by running your winners might be a good idea. My recommendation in that case is to rebalance only when a single position is taking up a significant portion of your screening strategy (i.e., a five-stock screen where 40% of your money is in one stock). That's a good idea for risk-management reasons as well.

Until next time: Fool On!

New Rankings | Workshop Returns