When Should I Rebalance?: A Tactical Approach
Start with the Strategic Asset Allocation Mix
Asset allocation determines most of the return of a portfolio1, and by far the most important decision investors need to make is how much to allocate to stocks versus bonds. Higher stock allocations usually lead to higher long-term returns, but at the cost of higher return volatility. Bonds tend to have lower but more stable returns. And unlike stocks, which can grow their dividends and thereby keep up with or even outpace inflation, bond yields are fixed, so bond investors suffer during periods of inflation.
Most experts agree that an investor’s risk tolerance (emotional willingness to take risk) and risk capacity (financial ability to take risk) should largely determine his or her strategic asset allocation (the expected long-term average allocation to each asset class over time). The problem is that assessing risk tolerance is very difficult. There are some good tools available that attempt to apply some objectivity and rigor to the process2, but the same investor with the same circumstances may be given wildly varying recommendations by different software, or even by different advisors using the same software. Risk assessment is still mostly an art rather than a science.
Perhaps it should not be surprising that, given the subjectivity of the risk assessment process, and the enormous importance of the outcome, most investors end up close to the conventional 60/40 stock/bond asset mix. As Keynes observed, “it is better for reputation to fail conventionally than to succeed unconventionally.”3
Many investors believe that, once determined, the strategic asset mix is somehow sacrosanct. That is, they think that a portfolio should be periodically rebalanced to return to the strategic asset mix if the allocations stray from their targets by a sufficient amount. Typically, current market conditions are not considered. That is, it does not matter what expected returns on stocks and bonds may be, the objective is to blindly return to the strategic asset mix.
This strikes me as somewhat ludicrous. Why ignore the expected return information for stocks and bonds given by current pricing and fundamentals? Even if an investor should have a long-term average mix of, say, 60% stocks and 40% bonds, why wouldn’t current market conditions influence what the tactical asset mix should be today?!
In fact, given that current pricing and fundamentals directly drive expected returns for stocks and bonds, and do so based upon solid financial theory and purely objective information (unlike the risk assessment process), I believe that current market conditions should always determine the extent to which current allocations will be above or below their long-term strategic targets.
Use the Current Expected Returns for Stocks and Bonds to Tactically Allocate Between Them
Financial assets are subject to a lot of return variability, which can mask the fact their expected returns (the mid-point of a distribution of possible future returns) may be derived from current prices and fundamentals. This is particularly true in the case of bonds. Forecasting the expected return for bonds is very straightforward. A bond is a loan with specified repayment terms. Held to maturity, unless the seller defaults, the return is the yield-to-maturity (YTM), which takes into account the current price, the interest coupons, and the principal repayment at maturity. Thus, the expected return for any bond is its current yield-to-maturity. End of story.4
Stocks, like bonds and all other financial instruments, have an intrinsic value equal to the present discounted value of future cash flows. Unlike bonds, where the cash flows are known in advance, cash flows for stocks (dividends) are not contractual, and tend to grow over time, which makes forecasting stock returns much more difficult than forecasting bond returns.
The total return for stocks can be separated into two components: dividend income and capital appreciation (change in price), which over the long term is based upon the dividend growth rate. The reason that capital appreciation ultimately depends upon the dividend growth rate is that dividends are the cash flows that give a stock its value. If I hold a stock forever, my return is the dividends I collect. If I sell the stock to someone, the price the buyer is willing to pay is determined by the long-term dividend flow the stock will generate discounted at an appropriate rate. One method to estimate the future dividend growth rate is to use the historical average dividend growth rate over a reasonably long time period, say 10 years.5
Some would argue that cash dividends have become less important as share buybacks have grown in importance. 6 Fortunately, using the growth rate in dividends per share may indirectly incorporate the effects of share buybacks through the shrinking share count.7
Determining the tactical allocation between stocks and bonds would logically be influenced by the expected return spread between stocks and bonds. One method for estimating the expected return for stocks is by adding together the dividend yield and the growth rate in dividends per share. The expected return for bonds is estimated by the yield-to-maturity. A higher-than-average expected return spread for stocks over bonds would result in a higher-than-average allocation to stocks, and vice-versa.8
This methodology I call the “Dynamic Model” approach to managing the asset mix. In a follow-on article, I will explain the model’s methodology and compare its backtested performance to a static 60/40 stock/bond asset mix.
1The classic study is Brinson, et. al., 1986, updated in Brinson, et. al, 1991. Generally confirmed by Ibbotson, 2010 among many others.
2Michael Kitces has an excellent article on this subject.
3John Maynard Kenyes, The General Theory of Employment, Interest and Money, Chapter 12. This quip may be particular applicable to investors and (especially) advisors who do not want to risk potential criticism for taking too much “maverick risk”—the risk of being different. In economics, advisors are what are called “agents.” They act on behalf of the owners of capital, who are termed “principals.” The “agency problem” arises when agents act more in their own best interest than in the best interest of their clients, the principals.
4For some bonds and bond indexes, if the bond gives the issuer the right to call the bonds—which is common for corporate debt—a better measure is the “yield-to-worst” or YTW. Schwab has a nice short article explaining different types of bond yields here.
5Valuation results are sensitive to variations in the method for estimating growth. I have used the 10-year compound growth rate in live models for years, but for more stable and robust results I blend it with forecasts of expected real GDP growth and expected inflation derived from the Treasury-TIPS spread. For the purposes of this and future articles, I kept it simple and used only the DPS growth rate.
6S&P Dow Jones Indices has good information on buybacks, but their data includes only gross buybacks rather than net buybacks (subtracting stock issuance) and is published with a delay. Yardeni Research also publishes on S&P 500 dividends and buybacks, and includes a graph on net issuance, but the numerical data behind the graph is not given, and like S&P DJ data, has a considerable lag.
7Indirectly capturing the effect of net buybacks by using growth in dividends per share assumes that corporations will continue to spend the same amount of cash on dividends in aggregate, which may or may not be the case.
8Other methods for estimated the expected return of stocks may be based on earnings rather than dividends. However, earnings are highly cyclical and require the use of some sort of smoothing technique. One well-known methodology used by Robert Shiller is to simply average earnings per share (EPS) over the previous 10 years. After adjusting for inflation, the resulting price/earnings ratio, or “cyclically adjusted P/E (CAPE), has been shown to be predictive of future stock returns over the long-term.