Shun Rebalancing

Rebalancing is a bad idea.

Flawed Assumption

The argument in favor of rebalancing rests on a series of flawed assumptions. The most important is that there is no meaningful way to measure valuations. Our research has shown otherwise. Professor Shiller’s P/E10, the dividend yield equivalent of P/E10 and Professor Tobin’s q all work well.

The original assumption was based on research that required a 5% advantage per year for a decade before it would detect an improvement. It is the rare individual who can beat the market consistently by 5% per year. Yet, a consistent advantage of 1% or 2% is HUGE after compounding. It shows up in the data.

Flawed Criteria

Rebalancing optimizes one’s return when compared to risk, as measured by volatility. It can maximize return at a fixed level of volatility. Or it can minimize volatility for a given level of return. Or it can maximize the ratio of the return to the risk. (More precisely, this is with the Sharpe ratio, which uses the return in excess of what is available without risk.)

Advocates of rebalancing too often assume that accepting risk guarantees a larger (expected) return. This twists the important guideline to investors: you should never accept a risk without being compensated adequately. It is very easy to accept too much risk. Blindly accepting risk is folly.

When two or more investments have EQUAL returns (statistically), rebalancing can only help. If their price fluctuations are uncorrelated, rebalancing enhances annualized returns. [NOTE: The average of year-to-year returns remain the same, but the combined volatility decreases. This increases the annualized return even thought the average return remains constant.] If prices are negatively correlated, the advantage is even bigger.

When the returns differ, rebalancing results in a combined annualized return between the two or more independent annualized returns. Rebalancing offers an advantage only if it is impossible to measure valuations meaningfully. Otherwise, you do better by shifting the odds in your favor.

Making things worse, volatility varies with time. As a practical matter, Mean Variance Optimizers fail. It is not possible to optimize the return as compared to volatility, looking forward.

Predicting Stock Returns

Use the Stock Return Predictor. (There is a “Stock Returns” button on the left.) At today’s valuations, the expected real return of the stock market (S&P500) over the next ten years is below 1%. Stocks are a poor choice at this time. Because of volatility, however, real stock returns could be almost as good as a 7% gain or as bad as a 5% loss.

At this time, Ibonds and TIPS (with real returns of 1% to 2.5%) are more attractive than stocks.

Today, rebalancing makes a lot of sense for those who believe that the stock market always returns 6% to 7% regardless of valuations. By holding on to bonds, they protect their downside. For those who are willing to take valuations into account, cutting back on stock holdings makes a lot more sense.

Compare this with what happens at typical valuations, with P/E10=13 or 14. The worst case real return at Year 10 is 1%. At typical valuations, you should invest heavily into stocks. Rebalancing at typical valuations simply eliminates the upside potential.

Retirees and Accumulators Alike

The first ten to fifteen years have the greatest influence on retirement portfolios. Ten year returns are critically important in Safe Withdrawal Rate calculations. But ten years is a lot of time for accumulators as well. The relevant time frame is 10 to 20 years.

Preservation of capital is exceedingly important in today’s market. The best way to know when to preserve capital and when to invest heavily into stocks is to take advantage of valuations. Our retirement trainers show that it can take 5 or 6 years before valuations return to a normal range. Waiting is difficult, psychologically. Benjamin Graham’s advice, to have a minimum 25% stock allocation at all times, reduces regret.

Have fun.

John Walter Russell
August 23, 2007