You’ve probably heard or read that you should periodically rebalance your investment portfolio.  That means selling assets from asset classes that currently exceed your target allocation and buying more in asset classes that lag, in order to maintain your target allocation over time.  The presumed benefit is higher returns, reduced portfolio risk, or both.  Unfortunately, there is no mathematical basis for either assumption.  Why, then, do it at all?

Michael Edesess, an economist with experience in the investment, energy, and environment fields, wrote several years ago that one of the justifications often given for rebalancing is the concept of “mean-reversion.”  When the value of an asset or asset class swings too far from its historical mean, it will at some point revert back towards that mean.  Unfortunately, he went on to point out that the theory on which mean-reversion is based was developed for a single period of time, not for an investment lifetime.  And it fails to address the fact that in an efficient market (such as the U.S. Stock Market) you cannot assume historical patterns will continue for the simple reason that if prices were to follow a predictable pattern, investors would arbitrage that pattern until it disappeared.

So what’s the alternative?  It’s the well-known “buy and hold” strategy.  Is there any empirical evidence that either approach is better?  In 2011 Forbes magazine compared two portfolios over a 25-year period starting in 1985, both with an initial target allocation of 60% stocks and 40% bonds.  The difference was that one portfolio was never rebalanced, while the other was rebalanced annually back to its 60/40 target.  Not surprisingly, the “unbalanced” portfolio performed better during periods of strong market growth.  But over the entire period the balanced portfolio had a better return: $97,000 vs. $89,000 for an initial investment of $10,000.  Furthermore it suffered less during steep market declines.  In the downturn of 2007-2009 the unbalanced portfolio lost 37% while the balanced lost only 30%.  And although the unbalanced portfolio gained 48% during the rebound from March 2009 through the end of 2010, the balanced portfolio did almost as well (44%).  This turned out to be a pretty consistent pattern during much of the overall holding period: when the market went down, you would have been better off had you rebalanced each year.  When the market rose, you would have profited less, but the upside difference overall was much smaller than the downside.  In other words, rebalancing the portfolio provided better downside protection at a relatively small cost to returns.

To be fair to buy-and-hold proponents, this is only one analysis of one allocation model over one 25-year period, not a lot of data from which to reach a strong conclusion either way.  But there are a couple of benefits that rebalancing provides that do not depend on numerical proof.  The first involves investor behavior.  Studies consistently demonstrate that investors change their behavior from aggressive to conservative after markets have fallen, and vice-versa, largely based on emotion.  This causes them to sell assets when prices are low and buy them when prices are high, exactly the opposite of what’s needed to sustain investment growth.  Rebalancing effectively forces an investor to do the right thing: sell when prices are high and buy when they are low.  And it is formulaic and easy to implement.  There is no emotion or market guesswork involved.

The second benefit is simply the very reason why you rebalance in the first place: to maintain your target allocation.  This is something everyone needs to do as part of a good investment strategy.  Surprisingly, many investors (and even advisors for that matter) do not even have one.  They simply invest to grow their savings without any thought as to how much growth they really need and how much risk they are actually incurring.  Modern Portfolio Theory teaches us that the more return you are targeting, the more risk you will incur.  If you only need to grow your savings by five per cent annually to support your goals for the rest of your life, why take on the added risk of striving for ten?

Even without a definitive theoretical basis, periodically rebalancing a portfolio to a target allocation is strategically sound and leads to more rational investor behavior.  That’s a good enough argument for me!