Assumptions about future returns are made every day by a wide variety of investors. Defined benefit pension plans rely on return assumptions to calculate funding status and necessary contributions while individual investors use them as a guide to determine if they are properly saving for retirement. Intuitively, higher return assumptions require less savings than lower return assumptions as investment gains are expected to be higher over time. Of course, the assumption of high returns in no way guarantees that will be the outcome. Rather, return assumptions should be grounded firmly in reality to avoid an investment shortfall in the future. Historical averages have been one popular way to formulate long-term return expectations. However, as we will see, these averages only tell part of the story. Furthermore, stock and bond market valuations today make it highly unlikely that a buy and hold approach will deliver returns anywhere near historical norms over the next 10 years. We believe that a flexible, active approach to asset allocation is the best way to provide investors with returns that allow them to achieve their financial objectives
over the coming decade.
Today many pension plans, both public and private, utilize long-term return assumptions in the 6% to 8% range. The chart above shows the distribution of return assumptions as reported by companies in the S&P 500 as of their last fiscal year end. The median return assumption of this group of companies is 7%. Many public pension plans have even higher assumptions. According to the 2015 survey of 126 public pension plans by the National Association of State Retirement Administrators (NASRA), the median return assumption among this group was 7.75%. A quick review of historical asset class returns explains why these return assumptions are so prevalent today. A simple blended portfolio consisting of 55% U.S. Large Cap stocks1 and 45% Intermediate U.S. Treasury bonds2 returned an annualized return of 8.5% since 1926.
However, these annualized returns are summary statistics that mask tremendous amounts of performance variation from year-to-year and even decade-to-decade. For example, since the start of 1926, in 20% of rolling 10 year windows, a 55/45 blended portfolio failed to achieve a return of even 6%. We believe that the next 10 years may be one such period where returns to a passive buy and hold balanced strategy are likely to come up short. The reason is simple—starting valuations both in equity and fixed income are not priced to deliver the returns that history has led investors to expect.
Valuations are notoriously poor predictors of stock and bond market performance over the short-term. Expensive can always become more expensive and cheap can always become cheaper, at least in the near-term. Over longer periods, however, the predictive ability of valuations becomes quite impressive. The table on the next page looks at the ability of the Cyclically Adjusted Price-to-Earnings ratio (CAPE) for the S&P 500 to predict equity market returns and the yield on the Barclays U.S. Aggregate Bond Index3 to predict bond market returns. In each case, linear regression models are built to measure the explanatory power (i.e., R-Squared) of valuations over various time horizons. As the time horizon extends, so too does the predictive power of valuation.
Looking out 10 years, starting valuations have explained over 75% of variation in stock returns and an incredible 84% of the variation in bond market returns since 1976. Unfortunately for investors, this means that the message valuations are sending today regarding long-term return prospects cannot be ignored. The scatter plots in the below charts show the historical relationship between starting valuation and ten year returns for stocks and bonds. A shaded vertical box has been added to highlight where we are today.
The S&P 500’s CAPE ratio of 26.36 as of 11/30/2015 translates to roughly a 5.6% estimated 10-year annualized price return for stocks. Factor in a dividend of 2% and this model suggests that a total return of 7.6% or so over the next 10 years is about as high as investors can reasonably expect. The story for bonds is even less compelling. The November ending yield to worst on the Barclays U.S. Aggregate Bond Index (i.e., U.S. investment-grade rated bonds) of 2.48% translates into an annualized total return based on our regression model of 3.3%. This number is exceedingly optimistic in our opinion as the model was based largely on a 30+ year bull run in bonds where both yield and price appreciation (i.e., falling yields) contributed to returns. At any rate, even if one uses the 3.3% estimate for bonds, a 55/45 stock/bond portfolio is still only estimated to deliver a total return of about 5.7% over the next 10 years:
The Case for Asset Allocation
It should be apparent by now that the strong relationship between valuation and long-term future returns spells trouble for those relying on a static buy-and-hold approach to achieve long-term returns anywhere near the assumptions used by many today. Investors in a situation like the one contemplated above have three options:
- Lower long-term return assumptions
- Raise target allocation to equities
- Embrace a flexible, active approach to asset allocation
Let us consider each option in turn:
Lower long-term return assumptions
Return assumptions should represent good faith estimates of the likely long-term return on investment. These estimates can have a significant impact on the estimated amount of savings required to reach an investment objective. A lower return assumption in a defined benefit plan increases the likelihood of greater contributions mandated by law. In other cases, a lower return assumption may require a foundation to reduce its spending targets to ensure that it remains financially viable well into the future. Participants in defined contribution plans such as a 401(k) must also contemplate increasing savings (if possible) in the face of lower expected returns or accept the prospect of a smaller balance to fund their retirement. While a lower return assumption may be necessary for those with unrealistic expectations, others should carefully evaluate their investment approach to determine if there is a way to increase the odds of successfully meeting their objectives.
Raise target allocation to equities
One seemingly logical solution to mitigate the impact of low expected fixed income returns is to reduce bond exposure and increase equity allocation. Depending on an investor’s current equity allocation, an increase may indeed be justifiable but as always, there are no free lunches. The greater expected return from stocks versus bonds is also accompanied by greater expected volatility. Greater volatility increases the likelihood that investment returns fall short of what was estimated over the long-term. In the shorter-term, it can also result in greater drawdowns that can adversely impact the ability of some investors to meet more near-term objectives.
Embrace a flexible, active asset allocation strategy
Long-term annualized returns can be very deceptive as they mask tremendous amounts of volatility that occurs over the course of a market cycle. For instance, consider an investment made in the S&P 500 on 12/31/2004 and held for 10 years. From start to finish, the annualized total return on this investment would have been 7.7%. However, the path to get there was anything but steady. In fact, in eight of the 10 years the calendar year return varied from the full period annualized return by 5% or more.
While timing the market perfectly is a fool’s errand, history shows that it is possible to identify periods where the odds of a particularly good or bad outcome are elevated. An active strategy seeks to take advantage of these periods by allocating capital away from areas that have an elevated risk of sustained capital loss and toward those where investors are being paid to take returns. To demonstrate the feasibility of such an approach, consider a simple allocation model driven by the shape of the yield curve.
The yield curve measures the spread between long and short dated U.S. Treasuries. The yield curve tends to go negative (i.e., inverts) late in the business cycle as short-term rates rise above those of longer dated maturities as the Federal Reserve raises rates to rein in an overheating economy. Conversely a very steep (i.e., upward sloping) yield curve tends to occur early in a business cycle as short rates are low due to central bank efforts to stimulate economic growth. Our yield curve model, as described below, presents an alternative to a 55% stock/45% bond portfolio (55/45 portfolio). Specifically, it takes a 30% weight (i.e., underweight position) in stocks when the yield curve is inverted and maintains that position until the curve returns a positive reading of at least 150 bps, at which point 80% is allocated to stocks (i.e., overweight position).
Starting in 1976, $100 invested in this simple active asset allocation model grows to $6,750 by October 2015 while that same $100 invested in a 55/45 portfolio only grows
In reality, one would never want to rely on a single factor such as the yield curve to drive asset allocation. However, the fact that a greater outcome would have been possible using a simple model suggests that there is indeed merit to an active approach to asset allocation. While we do not know what the future holds, having the flexibility to overweight and underweight particular asset classes can provide investors with the tools necessary to potentially outpace a static allocation approach.
Active Asset Allocation Today
The backdrop today for active asset allocation is challenging both from a near-term and long-term perspective. Having discussed the longer-term challenges at length above, it is now time to shift our attention to the outlook for 2016.
At any point in time, equity market returns are driven by one or more of the following factors: fundamentals, valuation, and sentiment. Our research finds that, historically, unsustainable extremes among these factors are often associated with market regimes that have a heightened risk of sustained loss or high return potential, depending on the nature of the extremes. For instance, by the end of the 2008-09 bear market, stocks changed hands at multi-decade lows in terms of valuation, sentiment was so negative that it seemed the sun might never come up again, and a global economic recession wreaked havoc on corporate bottom lines. It was out of these extremes that the strong bull market environment of the past several years was born. The late 1990s provides a great example of the other sort of extreme—sky high valuations, wildly speculative sentiment, and economic excesses, which all brought the great bull market to a screeching halt in early 2000.
Those seeking a point estimate forecast for the U.S. equity market in 2016 will have to search elsewhere. We believe that such efforts are generally a fruitless exercise that diverts attention from a true assessment of the risks and opportunities present in the market. Rather, we prefer a framework that seeks to identify regimes where the risk of sustained capital loss (i.e., a bear market) is high, or conversely market environments that strongly favor owning stocks. Such a framework also acknowledges that the market is likely to go through periods where neither regime is overwhelmingly likely. This is precisely where we find the U.S. stock market at the outset of 2016. While valuations are somewhat elevated, the U.S. economy remains far from overheating, and sentiment is anything but speculative today. In an environment such as this, investors can be rewarded by taking advantage of market volatility to increase their exposure to attractively valued and well positioned markets/sectors/companies. History suggests that the volatility of the opening trading days of 2016 is hardly unprecedented. If this volatility continues, investors may be presented with an opportunity sooner rather than later this year.
Active investing, of course, is more than just seeking to take advantage of market volatility. Valuations may be far from compelling in the United States today, but there are opportunities for those willing to invest globally. The next page shows that the CAPE ratio (using next 12 month earnings-per-share estimates) for many countries is significantly more attractive than for the United States. Valuation, of course, is just one piece to the overall asset allocation puzzle, but it does suggest that there may be opportunities outside of the United States today.
Nevertheless, successful asset allocation is not just about what an investor owns, it also is about what one chooses not to own. Investors in passive strategies must own whatever allocations make up the index, regardless of the merit of those investments. This can prove disastrous when an index becomes heavily skewed toward overvalued segments of a market. Examples of such distortions from recent decades include the allocation to Japanese stocks in the MSCI EAFE4 in the late 1980s, the allocation to technology stocks in the S&P 5001 in the late 1990s, and the allocation to financials in value indexes such as the Russell 1000 Value5 heading into the global financial crisis.
Today, we believe this phenomenon is occurring in aggregate U.S. fixed income indices. Weights in these indexes are usually based on the amount of debt outstanding such that entities with the largest borrowing needs receive the largest index weights. As a result, the amount of U.S. Treasury debt in these indexes has steadily moved higher in recent years. At the same time, the duration (interest rate sensitivity) of these indexes has also increased meaningfully. We believe that this dynamic will be problematic for those pursuing a passive fixed income strategy, as current yields do not compensate investors for the duration risk they must take on to own these indexes. However, investors pursuing an active strategy with the flexibility to deviate from the benchmark can better manage the risk of higher rates today, and may be better positioned to take advantage of higher yields should they rise in the future.
Valuations suggest that over the next 10 years a static asset allocation approach is likely to fail to meet the long-term return targets of many investors. In a low-return world, investors can ill-afford to operate without what we believe is the most effective tool for increasing the likelihood of achieving their long-term return objectives: an active approach to asset allocation. As 2016 unfolds, there are likely to be many twists and turns in the financial markets. Active asset allocation may provide investors the opportunity to respond to these developments and shift the odds of long-term success more in their favor.