Investing is complicated, but it begins rather simply. How it ends is the question.
There are infinite possibilities for reassembling the major asset classes. The challenge is finding the one that satisfies your particular set of expectations, risk tolerance and financial situation. The basic choices boil down to choosing some combination of stocks, bonds, REITs and commodities. We can further subdivide those broad categories and we can also employ any number of asset allocation strategies to manage the mix through time. That includes leaving out one or more asset classes, holding some cash, selecting individual securities and venturing into the shadowy realm of alternative betas.
How should we begin? We can start by considering a government bond. The benchmark 10-year Treasury Note offered a 3.72% yield as of yesterday. That’s a good place to launch our analysis because we have a high degree of confidence—a virtual certainty, in fact—that we’ll receive a 3.72% total return from a 10-year Note if we buy and hold till maturity. Deciding if the rate du jour will suffice depends on various factors, starting with a strong estimate of one’s future liabilities.
To keep things simple, let’s assume that an investor’s time horizon is exactly ten years. If earning 3.72% on current assets will match the anticipated liabilities over that time frame, buying and holding a 10-year Note looks like a good deal. In effect, the investment strategy is solved. On the other hand, if a 3.72% return on the given assets falls short of expected liabilities, we need to consider alternative possibilities.
Of course, even if 3.72% looks like a good deal, the investor may be more ambitious than settling for current yield. And, of course, we haven’t yet factored in inflation in the years ahead. As a result, looking at prospective returns and liabilities should be analyzed in real terms, i.e., inflation-adjusted returns. Indeed, the money game has already become more complicated and we’ve yet to leave the relatively safe terrain of government bonds.
With that in mind, let’s reconsider the example above with a modest assumption of 2% inflation for the next 10 years. That means that our 3.72% return on a 10-year falls to a 1.72% real yield. Will that suffice?
For some perspective, let’s consider that a 10-year inflation-indexed Treasury offers a real yield of 1.80% as of yesterday. Perhaps we’re better off with the inflation-indexed bond since it offers slightly higher real yield than its nominal counterpart, assuming 2% inflation.
Could we do better? Maybe. For instance, in the July issue of The Beta Investment Report, we advised that our conservative long-term equilibrium risk premium estimate for our Global Market Index was roughly 2.4%. For global equities, our equivalent forecast is 4.0%. (By risk premium we mean the return after subtracting a risk-free rate, such as the return on T-bills. If we assume a 2.0% risk free rate, GMI’s expected risk premium of 2.4% becomes a 4.4% total return.)
Like a game of chess, the first few moves in asset allocation are easy and even average investors can do well. But the level of complication quickly rises, as do the risks, since we must factor in a rising array of forecasts. Considering historical returns can help shape our outlook, but they’re only half the battle. As a result, we need to make assumptions as well to answer such questions as: Should we hold a 10-year alone? How will that fare under a moderate inflation outlook? And what about owning a passively allocated mix of the major asset classes? Should we try to second guess that? If so, how? What’s the reasoning?
Yes, the market offers some clues about which asset classes look relatively attractive, or not, but we need to proceed cautiously. The more we move away from benchmarks and passive allocations, the higher the risk, which implies that we should only engage in such activity to the extent that our confidence is above average.
How will we know if our confidence is better than average? There’s no easy answer, although we can develop some perspective by projecting risk premia on a regular basis if only to understand the embedded challenge. We engaged in an abbreviated version of just that in March, and we go into more detail on a regular basis of the pages of BIR. Depending on the asset class and the prevailing conditions, sometimes we have a relatively high degree of confidence in our projections, which inspires adjusting the market’s weight for the asset class. At other times, we’re not so confident, in which case we favor a market weight.
There are no easy answers in designing an asset allocation, all the more so when you consider that we must continually manage it through time. There is, however, lots of work to do. Inevitably, we’re going to be wrong sometimes, a fact that keeps us humble and raises the bar for what constitutes above-average confidence about projecting risk premiums.
The future, in short, is generally unclear, although it doesn’t appear that returns are totally unpredictable at all times. A few clues, it seems, can go a long way.