Rebalancing Taxable Portfolios

As noted above, most of the research on the benefits of rebalancing was performed on investment portfolios that were exempt from taxes. Because the tax costs associated with rebalancing taxable accounts can be very substantial, this research is suspect in the private client world. As a result, rebalancing of family portfolios needs to be more of a qualitative than a quantitative process, although of course it should be both.

As a review, we need to keep in mind the purpose of periodic rebalancing. Let's assume that we start with a portfolio strategy that is exactly aligned with our target asset allocation. As time goes by, our portfolio will drift away from the target allocation, as various sectors of the markets rise or fall faster than other sectors. Over time, our equity allocations will tend to grow much faster than our cash and fixed-income portfolios, but over the short term the opposite phenomenon could occur. In either event, the risk level of our portfolio will have changed, eventually substantially.

Suppose, for example, our target allocation has 55 percent in stocks, 35 percent in bonds, and 10 percent in hedge funds. In the absence of rebalancing, we may wake up some day to find that our portfolio is now invested 70 percent in stocks, 15 percent in hedge, and 15 percent in bonds. This is a vastly more risky portfolio than we set out to own, and in a bad market environment it will be hit hard—far harder than we are likely to tolerate without flinching badly.

Or assume that the markets have been in a bear phase. In the absence of rebalancing we may wake up to find that we are invested 40 percent in stocks, 15 percent in hedge, and 45 percent in bonds. This is far too cautious a portfolio for us. When the markets recover we will be largely left behind.

Instead of allowing our portfolio to drift with the whims of the market, we should be rebalancing periodically. But simply saying so raises a whole host of issues. Let's examine a few of the more important ones.

Setting Strategic Ranges

As with institutional investors, family investors will want to start with a target asset allocation strategy and then establish strategic ranges around these targets within which the portfolio will be allowed to move before rebalancing is considered. But a family's strategic ranges should generally be wider than an institution's. Allocation ranges that are too narrow will result in too-frequent rebalancing, and for taxable investors the cost of rebalancing will likely outweigh the benefits.

How wide should the “bands” be within which the portfolio can fluctuate? Too little attention has been given to this issue. As mentioned previously, most advisors use the same ranges for families that they have been using with institutions. But because the cost of rebalancing taxable portfolios is greater, the strategic bands should be wider. Even advisors who advocate wider bands for families tend simply to add and subtract 5 percent or 10 percent to or from the target allocation, resulting in silly outcomes. For example, an asset class with a target allocation of 40 percent might be allowed to fluctuate between 35 percent and 45 percent (12.5 percent either way), while an asset class with a 10 percent target allocation will be allowed to fluctuate between 5 percent and 15 percent (100 percent either way!)

A better approach would be to set the bands to describe a range of fluctuation that is consistent with the expected volatility of the asset class. For example, if an asset class has an after-tax standard deviation of 20 percent, why not start by setting the strategic ranges at plus or minus 20 percent? If an asset class has an after-tax S.D. of 8 percent, why not start with that as the strategic range? We can adjust these ranges upward or downward depending on how we feel about the tax issue, but the adjustments should be consistent with the expected volatility of the asset class, as well as its tax treatment.

Rebalance Back to What?

Let's assume that we have decided to rebalance our portfolio because our U.S. large-cap exposure, which has a target of 25 percent and a range of 20 percent to 30 percent, is sitting at 32 percent. Should we rebalance back to the target allocation (25 percent) or merely back to the top of the range (30 percent)? There is no purely quantitative answer to this question. If rebalancing can be done inexpensively, the better answer is to rebalance back to the target. But if rebalancing is going to be expensive, the better answer (assuming we are going to rebalance at all) is to rebalance only back to the top of the range. We need to keep in mind, however, that if we rebalance only back to the top of the range, a strongly rising market will require us to rebalance frequently.

A second issue to consider is whether valuations in the U.S. large-cap sector are high or low by historical standards. If we are at a 32 percent position in the sector but valuations remain low, we should be less eager to pay taxes and rebalance. But if valuations are high, rebalancing will be a more compelling idea.

How Often to Rebalance?

The issue here is highly qualitative, almost intuitive. My view is that taxable investors should set quantitative targets, not calendar targets. If an asset class has a 25 percent target exposure and a strategic range of 20 percent to 30 percent, we should consider rebalancing when the range is exceeded on either side. The default position should be to rebalance, but that default can be overcome if the cost of rebalancing is very high or if market valuations suggest that rebalancing won't prove rewarding. We should not rebalance automatically on a quarterly basis, as many institutional investors do.

The default position can also be overcome for valuation reasons, as noted above. If we are out of balance on the upside, and if valuations for the asset class remain low or reasonable,⁵ we might postpone rebalancing, being reasonably confident that a price collapse is unlikely. If we are out of balance on the downside, there will be no tax consequence associated with buying into the asset class. The question will be where we get the cash to make the buys and what the tax consequences, and valuation conditions, are in the cash source market class.

We should also have in place an over- or underexposure level that requires rebalancing, regardless of the cost and regardless of what we think of market valuations. That level might be set at double the normal range, or at whatever seems reasonable (but probably not more than double). When we are out of balance by that amount we must rebalance. Otherwise, our enthusiasm or pessimism will seriously compromise our investment results.

Thus, the targets, strategic ranges, and maximum and minimum ranges for two hypothetical asset classes (the first with a 20 percent S.D. and the second with a 10 percent S.D.) might look something like this:

We would create the same targets, normal ranges, and maximum ranges for each asset class.