Investors often ask, “When should I rebalance my portfolio?” Most are looking for a perfect time of year. Others want to know if they should rebalance once a year, once a quarter, or during every full moon.

Vanguard back-tested a portfolio that was split 60/40 between U.S. stocks and U.S. bonds between 1926 and 2009. They didn’t bother with lunar cycles. But they did test four scenarios: rebalancing the portfolio monthly, quarterly, annually and never.

Rebalancing once a month would have taken more effort. It also yielded the worst results. It averaged a compound annual return of 8.1 percent. By rebalancing quarterly or annually, investors would have averaged an annual compound return of 8.6 percent per year. Not rebalancing would have earned top scores: a compound annual return of 9.1 percent.

1926-2009 60 percent stocks, 40 percent bonds

Average Compound Return Growth of $100
Monthly Rebalancing 8.1% $69,399
Quarterly Rebalancing 8.6% $102,259
Annual Rebalancing 8.6% $102,259
Never Rebalancing 9.1% $150,410

Financial writer, William Bernstein, did a similar study over 28 years. The portfolio was split 40 percent into the S&P 500, 15 percent in U.S. small stocks, 15 percent in foreign stocks and 30 percent in 5 year-government bonds.

He used 28 different starting and ending points. The first went from January 1969 until December 1996. He tested the second from February 1969 until January 1997. Each of his 28 scenarios started and ended one month apart. His findings were similar to Vanguard’s. More frequent rebalancing meant lower returns.

28-Year Average Portfolio Returns With Rebalancing

Average Compound Return
Monthly Rebalancing 12.03%
Quarterly Rebalancing 12.09%
Annual Rebalancing 12.091%
Rebalancing Every 2 Years 12.166%
Rebalancing Every 4 Years 12.267%

Vanguard doesn’t say that we shouldn’t rebalance–nor does William Bernstein. Rebalancing reduces risk. It prevents a rising asset class from dominating a portfolio. Lopsided portfolios can fall hard when last year’s darling stumbles at this year’s ball. During some time periods, rebalancing can also boost returns.

William Bernstein explained this in his 1997 article, When Doesn’t It Pay To Rebalance? He says that when the performance difference between asset classes is less than 4 percent per year, rebalancing will win. For example, if stocks average 6 percent over the next five years and bonds average 3 percent, Bernstein says that rebalancing a portfolio should boost returns.

To test his theory, I compared the performance of Vanguard’s S&P 500 Index Fund and Vanguard’s Total Bond Market Fund between December 31, 1992 and December 31, 2015. The stocks averaged a compound annual return of 9 percent. The bonds earned a compound annual return of 5.5 percent. There was a 3.5 percent annual performance difference. Based on Bernstein’s theory, rebalancing the asset classes should have beaten a portfolio that was made from the same parts, but not rebalanced.

Investing 60 percent in Vanguard’s S&P 500 Index (VFINX) and 40 percent in Vanguard’s Total Bond Market Index (VBMFX) would have averaged a compound annual return of 7.9 percent­–without rebalancing.

Vanguard’s Balanced Index Fund (VBINX) contains 60 percent U.S. stocks, 40 percent U.S. bonds. Vanguard rebalances the fund, keeping it close to its original allocation. Over the same time period, it averaged a compound annual return of 7.98 percent. Rebalancing won.

Not rebalancing would have also been riskier. By 2015, the portfolio would have had a much higher emphasis on stocks because stocks beat bonds. Investors would have taken more risk and they would have earned lower returns.

When Rebalancing Wins 1993-2016

Funds Strategy Annual Compound Return $10,000 Becomes…
Vanguard Balanced Index Fund Rebalanced by Vanguard 7.97% $58,304
60% S&P 500; 40% Total Bond Market Index Not rebalanced 7.88% $57,232

Does Vanguard, however, rebalance better than you or me? To find out, I looked at the returns of Scott Burns’ Couch Potato portfolio. The syndicated finance columnist introduced it in 1991. It called for a 50/50 split. Half of the money went into Vanguard’s S&P 500 Index. The other half went into Vanguard’s Total Bond Market Index. Investors manually rebalance the indexes once a year.

Between 1992 and 2013, the Couch Potato portfolio averaged a compound annual return of 8 percent per year. If its components weren’t rebalanced, it would have averaged a compound return of 7.7 percent. Once again, these results support Bernstein’s theory. He says that rebalancing will win if the performance difference between the index funds is less than 4 percent annually. Between 1992 and 2013, stocks averaged a compound return of 9 percent per year; bonds averaged 5.8 percent. That’s a 3.2 percent annual difference.

In the future, if stocks average 9 percent and bonds average 3 percent, rebalancing between the two might not juice returns. But rebalancing always reduces risk. That’s why it’s worth it–always.

Andrew Hallam is a Digital Nomad. He’s the author of the bestseller, Millionaire Teacher and The Global Expatriate's Guide to Investing: From Millionaire Teacher to Millionaire Expat.