Long-Term Portfolio Projections Update
Let me first point out what this article is not: I will make no effort to predict the 3-month, 6-month, or 12-month returns of any market. Indeed, although I will later present long-term risk and return outlooks, and they are presented as point estimates, I want to make it very clear that these are not predictions but rather statements of relative risk and return possibilities. For many types of instruments, the error bars around the average annual performance are so large as to make point estimates (in my view) nearly useless. The numbers come from models of how markets behave when they are priced “like they are now” in terms of several important metrics. They are not prescient. However, that is what investing is really all about: not making the “right” bet in terms of whether you can call the next card off the deck, but making the “right” bet with respect to the odds offered by the game, and betting the right amount given the odds and the edge.
I also will not make portfolio allocation recommendations here. The optimal portfolio allocation for you depends on more variables than I have at my disposal: your age, your career opportunities, your lifestyle, your goals, any insurance portfolio and your risk tolerance, to name just a few.
What I will do here, though, is to give top-down estimates of the long-run returns and risks of some broad asset classes, and make some general observations.
Compared to last year, 10-year projected real yields are lower. For all relatively low-risk investments, expected real yields are negative. This is not terribly surprising given the current landscape: risk-aversive behavior is expensive behavior at the moment. This exercise helps to remind us of the fact that in avoiding short-term risks, we are sacrificing considerable long-term returns.
However, with that said it must be also observed that the preceding statement only is strictly true if a fire-and-forget investor makes a single portfolio decision at the beginning of the 10-year period. For an investor who may exit safe securities and enter risky securities once the compensation for taking risk is adequate, it could well be prudent behavior to eschew short-term investments in relatively risky securities. While I do not make 1-year forecasts explicitly, as investors we must always take into account the fact that we have to pay to wait for a better entry point, but at times this waiting is justified.
I don’t analyze every possible asset class. For this exercise, I limit the universe to stocks, TIPS, nominal bonds (both long Treasury and corporate bonds), commodity indices and (since many of us already own it) residential real estate. My estimates and some notations about the calculations are in the table below.
|Inflation||2.40%||Current 10y CPI Swaps|
|TIPS||-0.10%||Current 10y TIPS. This is not at equilibrium, but it is what we can lock in today.|
|Treasuries||-0.38%||Nominal bonds and inflation-linked bonds ought to have the same a priori expectation, but Treasuries trade rich to TIPS because of their value as repo collateral. Current 10y nominal rate is 2.03%, implying -0.38% real.|
|T-Bills||-0.50%||Is less than for longer Treasuries because of liquidity preference.|
|Corp Bonds||-1.52%||Corporate bonds earn a spread that should compensate for expected credit losses. Chart 1 below suggests that Moody’s “A”-Rated Corporate yield is about 40bps rich to where it should be for this level of Treasury yields. And these yields, remember, are being artificially held down at the moment. If Treasuries were at 3%, Corp AA fair values would be another 75bps higher. I think corps are very rich although I admittedly don’t use an actual default model.|
|Stocks||2.57%||2.25% long-term real growth + 1.98% dividend yield – 1.66% per annum valuation convergence 2/3 of the way from current 20.8 Shiller P/E to the long-run mean. Note that I am using long-run growth at equilibrium, not what TIPS are implying. See here for more on this method.|
|Commodity Index||4.78%||Various researchers have found that commodity futures indices have a long-run diversification return of about 3.5%. To this we add 1-month LIBOR to represent the return on the collateral behind the futures, and a ‘relative value’ factor to reflect the performance (relative to the expected model) of hard assets relative to currency.|
|Real Estate (Residential)||1.38%||The long-run real return of residential real estate is around +0.50%. Current metrics (see Chart 2) have Existing Home Sales median prices at 3.25x median income, versus a long-term average of 3.55x. Converging to the mean over 10 years would imply aan 0.88% per annum boost to the real return.|
The results, using historical volatilities calculated over the last 11 years (and put in terms of ‘real annuitized income,’ a term that means essentially the variance compared to a fixed 10-year real annuity, which in this analysis would be the risk-free instrument), are plotted below. Return as a function of risk is, as one would expect, positive. For each 0.30% additional real return, an investor must accept a 1% higher standard deviation of annuitized real income.
There are some surprises here (indeed, some surprises for me since I don’t do this calculation every day). The biggest of them is that the way I do the analysis, home prices are actually slightly cheap to other available assets. Please remember, this isn’t a forecast of 2012’s return, but after a long round-trip it appears to me as if we no longer need to regard our homes in general as albatross assets!
Treasuries, not surprisingly, are extremely rich and should probably be avoided by most investors who are not hedging fixed nominal liabilities. Fixed-income allocations are better in TIPS and T-Bills, although neither offers very exciting real returns. Corporate bonds look quite rich, with yields that are even lower than they should be given the level of Treasury yields, and the latter are (as I’ve just pointed out) themselves rich. Yield-seeking investors are extending credit to issuers on terms that seem not to adequately reflect the very risky business environment that currently exists.
Stocks are rich, but not as rich as they have been in the past. Last year, my 10-year projection was 2.58% annualized real return, so equities have held steady while many other asset classes (TIPS, Treasuries, Corporate bonds) have gotten richer. Be careful about whether you read that statement as “stocks have cheapened,” or “stocks are cheap.” The former, a relative statement, is true, the latter, a (more) absolute statement, is false. As investors, we clearly would prefer to make the latter statement. As a long-only investor, you should care about the value of equities relative to equilibrium, not relative to other overvalued markets, anyway. Right now, if you compare in the chart above Stocks (which are overvalued) to Treasuries (which are more overvalued), stocks look ‘cheap.’ But making such a statement is analogous to someone in Holland a few hundred years ago remarking ‘This tulip is less-overvalued than that tulip, so it’s a good investment!’ Of course, no tulips were good investments, even the ‘cheap’ ones.
As I have been saying, Commodity Indices are the most-attractive investment in my little universe. While they are historically as risky as stocks, they offer much greater value given their performance over the last year (-11.8%, basis the DJ-UBS index) compared to both equities (S&P +0.61% as of this writing) and the money supply (M2 +9.5%). The “best” portfolio depends on a lot of characteristics of the investor as well as of the portfolio, but one could do worse with a portfolio consisting of TIPS, T-Bills, and Commodity Indices as the liquid assets balancing the illiquid investment in one’s home. If you are afraid of missing the next bull market in equities and not being able to look people in the eye at the next cocktail party, then buy out-of-the-money calls on stocks. Implied volatilities have fallen sharply over the last week or two (based on the VIX), and so protecting the ‘downside’ from a regret standpoint is much more affordable than it has been.
Finally, one caveat to all of this: I have not mentioned the economic backdrop, nor incorporated my view of that backdrop into these forecasts (with the exception of choosing to value corporate bonds off of a higher Treasury yield than is actually extant, recognizing that the Fed has artificially depressed long Treasury yields). The “risk” spoken of above is based on the historical standard deviation of the various asset classes, and is independent from the macroeconomic risk we all face at the moment. From a strategic perspective, abstracting from the tumult of the time makes sense; however, every investor must also be cognizant of tactical considerations. I hope that this column helps raise some useful questions about those tactical considerations, even if it doesn’t often provide the right answers to those questions.
That’s all for 2011. Thanks to all followers and one-off readers of this column! Have a happy new year!