Annually rebalancing client portfolios hurts performance: Pinnacle’s Michael Kitces

The simple and straight forward ritual of annually rebalancing client portfolios is actually hurting performance, according to Michael Kitces, director of research at Pinnacle Advisory Group.

Speaking Monday morning in Orlando at the IMCA annual conference, Mr. Kitces laid out the case for rebalancing based instead on the volatility of the assets making up a portfolio, which could leave portfolios untouched for multiple years at a time.

Citing numerous studies on how rebalancing brings a portfolio back into it original position, and helps manage risk, Mr. Kitces underscored the obvious impact of selling winning assets and buying more losing assets.

“Rebalancing your portfolio is one of the few free lunches out there,” he said. “But we know that markets can remain irrational longer than a lot of people can remain solvent.”

In one example, he compared the performance of a hypothetical portfolio split evenly between stocks and bonds that was rebalanced annually with the same 50/50 portfolio that was left alone over the period.

The portfolio that was left alone outperformed the rebalanced portfolio by 20%.

“It’s the power of compounding; stocks out-compound bonds,” Mr. Kitces said. “A lot of rebalancing doesn’t enhance returns, and in some scenarios it increases risk. In the long run there isn’t much question, because systematically rebalancing can chop about 20% off the returns over the same 30 years.”

However, the portfolio that never got rebalanced resulted in a weighting of 80% stocks, and 20% bonds, representing a glaring problem associated with just leaving a portfolio alone for three decades.
“I do not want to be standing in front of a judge trying to explain why my retired client is 80% allocated to stocks,” Mr. Kitces said.

His research showed that rebalancing hurts performance in both bull and bear market periods. And, while annual rebalancing underperforms zero rebalancing, more frequent quarterly and monthly rebalancing performs even worse.

“By rebalancing too frequently we clobber ourselves with transaction costs,” he said. “Rebalancing dials down volatility, but it also dials down returns, and we all know that people don’t eat risk-adjusted returns.”

During both the 2000-2002 bear market and the 2003-2005 bull market, the buy-and-hold strategy outperformed all forms of rebalancing.

The alternative to calendar-based rebalancing is rebalancing when any asset class reaches a pre-determined level of gain or loss, known as tolerance band rebalancing.

While this band can be set at virtually any level, Mr. Kitces said a 20% band is optimal. Key to using a band to trigger rebalancing is near-constant portfolio monitoring, he said.

“Daily monitoring is best, but that could be dialed back to every other week without losing much,” Mr. Kitces explained. “With tolerance band rebalancing you’re trying to figure out the maximum stretch point. You’re not trying to figure out the frequency of when asset classes bounce back. This is not about caring how long it takes, just how long you can stretch it before it bounces back.”

For portfolios that have money coming in, he said the money should be invested in the asset class that is down the most. “It becomes an incredibly easy way to keep on track,” he said.

And for retirees who are taking money out of a portfolio, Mr. Kitces said you should sell the asset class that is up the most.
“Rebalancing doesn’t help retirees,” he said. “Just sell what is up the most and you’re accomplishing the same thing.”