Dec. 29, 2019

Portfolio rebalancing has been around for ages, whether it be a Sharpe ratio, Markowitz rebalancing, equal weighting of securities held, or something else; most were done at pre-scheduled time periods. This made them program time-dependent. As such, it was almost independent of whatever the market was doing at the time. If you opted to rebalance at the start or end of the month, it was an arbitrary program choice that carried with it its own set of trading circumstances.

You design an automated trading strategy and because of its inherent structure, you also design what will turn out to have been a statistical problem. For instance, a fixed fraction monthly scheduled rebalancing of equity on equal weights with a bond or cash switcher as a protective measure, like many strategies found on the Quantopian website, will technically determine their long-term outcome. (Quantopian was shut down in October 2020.)

Your stock selection will matter, and how much capital will be at stake. Just as the bond or cash switcher will determine the method of interaction with real market data.

On a 100-stock portfolio, you will get at most (100 ∙ 12 = 1,200) trades in a single year if rebalancing monthly. Running the program for 200 months should provide at most (100 ∙ 200 = 20,000) trades. However, due to the structure of the program, that is not what will happen. It will trade less. The reason is simple: not all stocks, at a scheduled rebalancing time, will need to have their weights “adjusted” and, therefore, might not trade.

Furthermore, the rebalancing has some issues that will gather, with time, some secondary effects that would not be so desirable over the long term. Stocks on the rise are gradually sold off to the point that they have less and less impact on the overall portfolio performance. While underperforming stocks will see their share count increase the more they fall in price.

This is creating, with time, a portfolio that has subdued its winners while doubling down on its poorest performers. A prelude to a disaster in the case of a major drawdown. Hopefully, the protective measures will be there to alleviate what could be a severe drawdown if not a portfolio destroyer.

The rebalancing, since it is coupled with its own stock selection process, can alleviate the downside risks by concentrating its selection on what it perceives as the best momentum or rising stocks out there. A simple premise would be to limit the selection to stocks that are already performing above market averages. A set of stocks performing above the market average will beat the average by definition.


Dec. 29, 2019, © Guy R. Fleury. All rights reserved.