The most perplexing thing I read this week was an odd column by Paul Merriman at MarkeWatch, "Why Rebalancing Could Be a Huge Mistake." He makes the counterintuitive claim that portfolio rebalancing doesnt really work: Conventional wisdom holds that regular rebalancing is a sound practice to control investing risk. But Ive concluded that some of that conventional wisdom is wrong.
I dismissed Merrimans claim offhand because what he described was not true asset class rebalancing. He was looking within the same asset class -- U.S. equities -- and subdividing them into four market cap weighted, style-based subgroups: Large-cap, small-cap, large-cap value and small-cap value stocks.
His conclusion that rebalancing among these four didnt work was a bit disingenuous. If you went on an all-Chicago-deep- dish-pizza diet, and didnt lose any weight, you would not exactly be justified in declaring that diets dont work. But that is functionally the equivalent of Merrimans critique of intra-U.S.equity rebalance.
But rather than rely on my own insights into rebalancing, I tracked down two experts who have studied rebalancing for years. Their criticism was much stronger than my own.
Robert Arnott runs Research Affiliates, where he helped craft the idea of fundamental (as opposed to market cap weighted) indices. RAFI, as the firm is known, generates returns through a combination of Smart beta and asset allocation strategies.
Arnott critiques Merrimans rebalancing claimas a "flimsy and meaningless straw man. He shows through a series of examples what occurs if you chose not to rebalance a portfolio: The investor loses the benefits of diversification, as the portfolio over time morphs into a riskier, 99 percent stocks, mostly small-cap value fund.
The obvious risk (at least to me) with such a portfolio drift is the behavior of its holders. Any un-rebalanced portfolio eventually would become a gnarly mass of volatility and drawdowns. Imagine what that does to the psyche of the investor saving for retirement. It invites bad behavior and poor decision-making. Meanwhile, a diversified portfolio will have different sectors doing better or worse each year, producing more stable returns.
James Picerno takes a different critical approach to reach a somewhat similar conclusion. He starts with the caveat that the optimal rebalancing strategy is unknown in advance. Instead, it is designed specifically because we never know what the future holds. Rebalancing what is essentially an all equity portfolio of using only U.S. equities is not the same as rebalancing a broad asset allocation model. We should not expect an equities-only portfolio to have the same results via rebalancing than what we can achieve via a regularly rebalanced multi-asset-class strategy
We may not know in advance the precise improvements rebalancing produces over time, but we do know they are substantial. The alternative is a massive drift from ones original asset weighting.
Any asset allocator who is comfortable with what will eventually become all equity portfolio should skip rebalancing. The rest of us continue to take advantage of what may be the only free lunch on Wall Street.