Threshold Rebalancing

This is the third and the final part of Rebalancing series. Previously (here and here), I have checked how often one should rebalance an investment portfolio for a traditional 60% stocks and 40% bonds portfolios as well as for a multi-asset, broadly diversified portfolio. In both cases, annual rebalancing performed better than semi-annual and quarterly.

Today, I am going to analyze a predetermined threshold rebalancing method and summarize which rebalancing method works best and which one you may use to reduce portfolio risk and enhance performance.

What Is a Threshold Rebalancing?

A threshold rebalancing method assumes rebalancing your portfolio once any asset class or individual stock in the portfolio deviates from its target weight by a predetermined level. It can be a 1%, 5%, 10% or any other level your are comfortable with. Fundamentally, this method significantly differs from a time-based, where one rebalances a portfolio periodically based on a chosen time interval. Depending on your threshold level you may end up rebalancing either too often (multiple times during one year) or infrequently (a few times every 3-5 years).

In order to see the benefits of the threshold rebalancing, I’ve conducted a test for 1%, 3%, and 5% threshold levels on a multi-asset class portfolio. When any asset class in the portfolio drifted above or below the threshold level on any given day, all asset classes would be rebalanced back to their target weights.

Consistent with my previous analysis, analysis was based on the following assumptions:

• Testing period: 12/31/2003 – 3/20/2015
• ETFs used: “VTI” for US stocks, “AGG” for bonds, EFA for developed markets stocks, EEM for emerging markets stocks, IYR for real estate and GLD for gold
• Rebalancing threshold levels: 1%, 3%, 5%, and  no rebalancing
• Starting portfolio balance $100,000
• Trading transactions and taxes are not accounted for

A sample diversified portfolio that I used in my tests has the following allocation:

• US Stocks – 35%
• Bonds – 20%
• International developed markets stocks– 15%
• Emerging markets stocks – 10%
• Real Estate – 10%
• Gold – 10%

As we can see on the charts below, doing threshold rebalancing significantly improves portfolio performance when compared to “no rebalancing” portfolio.

When we breakdown by the threshold levels, we can see that rebalancing at 5% level increases portfolio balance to $253,331 which is 2.3% higher than 1% threshold, 2.1% higher than 3%, and 10.3% higher than “no rebalancing” portfolio. That is a big difference which you cannot simply leave on the table. Risk-wise, rebalancing each time at a 5% level decreases risk as measured by standard deviation to 13.1%, which is 0.8% lower in comparison to “no rebalancing portfolio.” The other factor I would like to focus your attention on is the number of rebalancing events. For a 1% level you would have to do 109 rebalancing events, while at 5% level just 7. Taking into account transaction costs, the advantage of 5% level would be even more.

A quick takeaway: without any doubts threshold rebalancing works, reduces portfolio risk and increases return when compared to “no rebalancing” portfolio.

Time-Based Vs. Threshold Rebalancing

The last thing I would like to do is compare a multi-asset portfolio with time-based rebalancing to the same portfolio but using threshold level rebalancing.

Last week I showed that an annually rebalanced portfolio performed better that quarterly or semi-annually both return and risk-wise. But how does annual rebalancing compare to threshold rebalancing? Which method is better?

To answer this question, let’s look at the table below.

As can be seen from the table, 5% threshold level enhanced the portfolio return by almost 2% over the annual rebalancing method, or 0.4% per year. However, risk was 0.1% higher 5% threshold level. The difference in risk is negligible and I would not pay much attention to it. However, 0.4% annual threshold level advantage is a meaningful, in my opinion, and if you can obtain it by simply switching between the methods, why not to do it?

Conclusion

I’ve shown that whatever rebalancing frequency or threshold level you choose, rebalancing enhances overall portfolio performance and, most importantly, decreases the level of portfolio volatility when compared to “no rebalancing” portfolio. For most individual investors the best approach would be to do it at a 5% threshold level, or at least annually. Despite better performance and less risk, another good reason for less frequent rebalancing is avoidance of transaction costs. The more often you rebalance your portfolio the higher the negative effect of transaction costs is, which can eat up a significant portion of rebalancing profit.

Another important thing to note is that rebalancing is a time-consuming process and heavily depends on how diversified your portfolio is. The more asset classes/stocks you have in it, the harder it is to rebalance. If you have no time nor clear understanding which method you should choose, I would recommend hiring an investment adviser who will identify the best method and perform all the “dirty” job for you. The only thing I am confident is that rebalancing works, enhances returns, and reduces risk. That’s why you should never ignore it.