February 9, 2015
Expect a Decade of 1.7% Portfolio Returns from a Conventional Asset Mix
Friday’s employment report showed a 257,000 increase in January non-farm payrolls. This news was followed by a spike in Treasury yields up to 1.96%, a 4% plunge in utility stocks, a 5% plunge in precious metals shares, and took the S&P 500 within a fraction of a percent of December’s record high, before a late-day retreat. These frantic market movements smack of an investment climate dominated by one-dimensional “theme” based behavior - where asset prices have been amped up on yield-seeking speculation, but where the most marginal change in the outlook can trigger a race for the hills or a pile-on, depending on whether the asset has features that are consistent with that theme. On Friday, the knee-jerk reaction was that stronger employment will prompt the Federal Reserve to raise interest rates sooner, creating a scramble to get out of yield-sensitive Treasury bonds and utilities, to buy dollars, and to sell foreign currencies and gold. Of course, in equilibrium, there must be someone on the other side of those trades, so prices moved to the extent needed to find that match.
Even after last week’s volatility, we continue to observe dispersion in market internals, a recent widening of credit spreads, and other features of market action that – at least for now – convey a subtle but measurable shift toward risk-aversion among investors, in an environment where risk premiums remain razor thin.
It is in this context that I want to go back to our opening comment from last week (see Market Action Suggests Abrupt Slowing in Global Economic Activity):
“The combination of widening credit spreads, deteriorating market internals, plunging commodity prices, and collapsing yields on Treasury debt continues to be most consistent with an abrupt slowing in global economic activity. Generally speaking, joint market action like this provides the earliest signal of potential economic strains, followed by the new orders and production components of regional purchasing managers indices and Fed surveys, followed by real sales, followed by real production, followed by real income, followed by new claims for unemployment, and confirmed much later by payroll employment. Stronger conclusions, particularly about the U.S. economy, will require more evidence, but from a global perspective, these pressures are already quite evident.”
Notice that payroll employment is last in the sequence of economically informative indicators. Simply, payroll employment is well-known to be a significantly lagging indicator, reflecting the climate of economic activity that broadly existed several months earlier (the unemployment rate actually lags even more). The reason is inherent in how people are hired and fired. The point in time that a business decides to add or remove an employee is well-aligned with the actual business climate at that time, but the full process requires interviews, employment contracts, and negotiation of starting dates, or conversely, several weeks of termination notice, and additional lags in reporting the employment change. So what we saw last week in the employment figures tells us a lot about how the economy was doing in the September-October period. Indeed, more timely measures such as the new orders (including order backlog) and production measures of regional purchasing managers indices and Fed surveys are consistent with strong economic activity at that time. But those measures have also turned abruptly lower in recent weeks, following the initial deterioration that was evident in market internals and credit spreads.
In my view, the Federal Reserve may find it difficult, from the standpoint of economic activity, to justify raising interest rates even a few months from now – though I still believe that the Fed should immediately discontinue reinvesting the proceeds of assets as they mature, for reasons I detailed last week. Given the prospect that the Fed might defer hiking rates (via increases in the amount of interest it pays to banks on idle excess reserves), the very first reaction of many investors is that this would be bullish for stocks. It’s here where we need to be careful about what history, and even the experience of the past several years, has actually taught.
First and foremost, the response of the equity markets to Federal Reserve easing (and much other news) is conditional on the risk-tolerance of investors at the time, which we infer from observable market action such as internals and credit spreads, among other factors. Quantitative easing “works” by creating default-free liquidity in an environment where that liquidity is viewed by investors as an inferior asset. That is, if investors are risk-seeking, as inferred from the uniformity of market action across securities, sectors and asset classes of all risk profiles, then yes – Fed easing will tend to support further advances in stock prices regardless of the level of valuation. On the other hand, once investors have shifted toward risk-aversion, overvalued markets become vulnerable to abrupt free-falls and crashes, and monetary easing is not materially supportive for stocks because default-free liquidity is desirable.
Again, as I noted in The Line Between Rational Speculation and Market Collapse, investors should remember that the Fed did not tighten in 1929, but instead began cutting interest rates on February 11, 1930 – nearly two and a half years before the market bottomed. The Fed cut rates on January 3, 2001 just as a two-year bear market collapse was starting, and kept cutting all the way down. The Fed cut the federal funds rate on September 18, 2007 – several weeks before the top of the market, and kept cutting all the way down.
What will matter significantly for investors is the condition of market internals, credit spreads, and other risk-sensitive measures in the event that U.S. economic activity begins to further reflect the downturn that is already evident abroad. It is that evidence of investor risk-preferences that will determine the proper response to any change in Fed policy.
Blurring the line between monetary and fiscal policy
Very often, the policy of central banks is described as “pumping money into the economy.” This is incorrect, and it’s important to understand what’s actually going on. When a central bank “creates money,” it does so by purchasing an already existing government liability (a bond) and paying for it with a different government liability. So what the Fed does is to substitute one government liability for another in the portfolio of investors. The reason this doesn’t automatically result in more spending is that the person who sold the bond was already inclined to hold that government liability as a store of value rather than as a means of payment. The monetary operation simply changes the form of that store of value from a bond to a currency unit. It’s incorrect to say that central banks “print money out of thin air” – what they actually do is change the form of existing government liabilities.
One might not think of currency as being a liability in the traditional sense, but if you look at the top of a dollar bill where it says "Federal Reserve Note," the accounting is still the same. The Fed is required to hold some asset (Treasury or government agency debt) on the other side of its balance sheet in order to back that "Note" in your hands. By contrast, fiscal policy by Congress involves an actual act of spending, and if it’s deficit-financed, that act of spending is coupled with the creation of a new liability.
What’s particularly interesting about Mario Draghi’s efforts at the European Central Bank is that he is actively trying to blur that line. Specifically, if the ECB buys an asset that then defaults in its hands, it has created a liability (in euro) that is unbacked. That is, it has engaged in fiscal policy. More precisely, if the ECB buys the debt of an individual government, and it goes bad, the ECB will have effectively created a joint liability of all Euro-area countries in a way that effectively relieves the debt of one particular defaulting government. That’s why Germany won’t allow it, and insisted that 92% of the Q-ECB program announced by Draghi must be done by individual national central banks.
Two points remain to be seen. First, I continue to suspect that at some point before March 15, it may turn out that the purchases to be made by each national central bank are voluntary and not mandatory. Second, it remains to be clarified what the legal mechanism is to compel those individual national central banks to retire euros created against debt that goes bad.
This second bit is interesting, because there are two possibilities. The first is that the national debt goes bad, and the country then creates new debt and gives it to the central bank in order to back the euros already created. In that case, the government must engage in an act of fiscal policy – essentially creating debt in order to make the central bank whole. This is consistent with Draghi’s statement that “each national treasury gives an explicit or implicit indemnity to its own central bank.” And while that may seem relatively painless, it also means that in the event of a default on government debt, it cannot be true that private holders of Euro-area government debt will be treated the same as the central banks (which is one of the ECB’s claims), as the central banks would be indemnified, while private holders would be out of luck.
The second possibility is that the country exits the Euro-system. In that case, the national central bank presumably could not have valid liabilities in the form of euros. Instead, it would have to retire those liabilities by buying-in euros (essentially canceling them out), paying for them by creating new liabilities in the form of the new national currency. This would also require a potentially complicated mix of market-determined and pegged exchange rate mechanisms in order to accomplish the switch. Fiscal policy would also be required, with the national treasury handing new debt securities to the central bank (with nothing in return) in order to back the corresponding liabilities. Meanwhile, just as those owing debt denominated in Swiss francs experienced a sudden increase in their debt obligations from the standpoint of the euro, existing debts in euros would likely be suddenly far larger from the standpoint of the new national currency. The resulting risk of widespread default is why an exit of weak nations from the euro could be so disruptive.
All of this will be fascinating to watch. Fortunately, we don’t need to make pointed forecasts. Our focus remains on both valuations and on factors (market internals, credit spreads, and other risk-sensitive measures) that convey information about investor risk-preferences. For now, we observe extraordinarily thin risk premiums in an environment where we still infer a measurable shift toward investor risk aversion. That will change with time, and we’ll take that evidence as it arrives.
Expect a decade of 1.7% portfolio returns from a conventional asset mix
Speculative yield-seeking has driven asset prices higher in recent years to the point where many asset classes now provide no risk premium at all. The most historically reliable equity valuation measures we identify (having a correlation of about 90% with actual subsequent 10-year total returns) remain about 112% above – more than twice – their pre-bubble norms. We presently estimate 10-year S&P 500 nominal total returns averaging just 1.6% annually. Yes, even less than the 1.9% yield-to-maturity on 10-year Treasury bonds. Even here, we’re assuming historically normal future nominal growth in revenues, nominal GDP and so forth of about 6% annually – see Ockham’s Razor and the Market Cycle for the arithmetic and long-term historical record of these estimates.
As a sidenote for those inclined to dismiss these estimates because of our awkward transition in other aspects of our work from 2009 to mid-2014 (see A Better Lesson Than "This Time is Different" for a full discussion of that transition), recall that these same methods identified the market as undervalued after the 2008 collapse, with the valuation measures falling below historical norms and our estimates of 10-year total returns for the S&P 500 moving above historical norms. We've maintained a very open narrative during our awkward transition since 2009 (which started with my insistence on stress-testing our methods against Depression-era data following a market collapse that we had fully anticipated). We've detailed where our challenges have been, and exactly how we completed that transition in mid-2014. But our valuation methods were not part of that difficulty, and should not be dismissed on that basis.
Now, there are several 10-year periods in the historical record where actual market returns in the preceding decade overshot (undershot) the return that one would have projected a decade earlier. For example, by the 2000 peak (when we correctly projected negative total returns for a decade), the S&P 500 had overshot the return that one would have projected in 1990. Conversely, by the 1974 low (when the same methods projected total returns near 20% annually for a decade), the market had undershot the return that one would have projected in 1964. Those "errors" are the main reason why the correlation of these estimates with actual subsequent returns isn't closer to 100%. Not surprisingly, these periodic "errors" have invariably been associated with major valuation peaks (troughs). In order to establish an overvaluation extreme, the market must overshoot prior expectations that manner. As I detailed in Do The Lessons of History No Longer Apply?:
"Given the full weight of the evidence, it should be clear that one can’t just say 'well, look, the S&P 500 has done better than these models would have projected a decade ago,' and use that as a compelling argument that this time is different and historical regularities no longer hold. Quite the opposite – the overshoot in S&P 500 total returns since 2004 – relative to the prospective returns one would have estimated at the time – is highly informative that stocks are strenuously overvalued at present. That conclusion has strong statistical support. In fact, when we examine the historical evidence, we find that there’s a -68% correlation between the error in the projected return over the past decade and the actual subsequent total return of the S&P 500 in the following decade. That is, the more actual 10-year S&P 500 returns exceeded the return that was projected, the worse the S&P 500 generally did over the next 10 years. Notably, the “Fed Model” has a correlation of less than 48% with actual subsequent 10-year returns. It’s sad when a valuation measure that is so popular is outperformed even by the errors of better measures."
One can certainly decide to embrace equity market risk in the belief that a 1.6% 10-year prospective return on stocks is acceptable given currently low interest rates. That’s fine, provided that one also understands that returns over the coming decade are likely to be no less dismal in light of that justification. One might also assume away the risk, and console oneself that the next decade will simply represent a kind of stable, low-return equilibrium across every kind of risky asset. And maybe armadillos will suddenly discover the ability to fly. It’s just that evidence is not kind to those fantasies. With the exception of the 2002 market low, no market cycle in history ended with prospective equity returns below 10%, even in periods where Treasury yields were similar to the present (as they were for most of history prior to 1960).
Now, we’ve seen negative expected risk premiums on stocks before, but never in an environment where interest rates provided no way out of the situation. For example, in December 1968, these same valuation methods would have projected no risk premium in stocks versus 10-year Treasury bonds. But at the time, 10-year Treasuries were still yielding 6.2% (as it happened, the S&P 500 went on to achieve a 10-year annual total return of 2.9%). Similarly, in August 1987, the estimated risk premium on equities also went negative relative to 10-year Treasury bonds, but those bonds were yielding 9.5% (as it happened, the S&P 500 ended the following decade with a total return of 8.6%). The projected risk premium also went negative during the tech bubble, reaching two troughs. The first was in April 1998 when 10-year Treasuries were yielding 5.7% (the S&P 500 ended the following decade with a total return of 0.3%). The other was in January 2000, setting the most negative estimated risk premium in market history, when the 10-year Treasury bond was yielding 6.8% and our 10-year estimate for S&P 500 total returns was below zero. As it happened, the S&P 500 ended the following decade with a total return averaging -0.9% annually. In each case, bonds offered an adequate return even when stocks were relatively uncompetitive.
The problem for investors here is that risk premiums are compressed in equities at a time when bonds offer no way out. Normally, one can shift allocations between conventional assets depending on prospective returns. Here, conventional asset allocations (particularly equity-heavy ones) create not only the risk of severe losses, but the prospect of dismal returns even if one accepts those risks. When risk premiums are compressed across the board, conventional asset allocations are very much like trying to squeeze water from a stone. This is an environment where alternative asset classes (such as hedged equity) might be of benefit, as they clearly were in periods following other valuation extremes such as 2000 and 2007.
By our estimates, never in history, prior to the past 5 weeks, have the prospective 10-year nominal annual total returns of both stocks and Treasury bonds been below 2% at the same time. We currently project a 10-year nominal annual portfolio total return averaging only about 1.7% annually for anything close to a standard portfolio mix of equities, bonds and cash – regardless of how much diversification one has within each of those asset classes.With regard to the near-term, our risk concerns are always most pointed when a hostile market climate is coupled with an advance that is overbought or at prior resistance. But that’s also the point where continued strength – if sufficiently uniform – could allow us to infer a fresh shift toward risk-seeking investor preferences. We can’t rule that out, and at current valuations our probable response would be to encourage an outlook along the lines of “constructive with a safety net.” At present, however, we don’t observe that evidence. With obscene overvaluation joined with continued evidence of investor risk-aversion and an overbought advance pressed against upper Bollinger bands, our immediate concern about the potential for an air-pocket or free-fall remains high.
The foregoing comments represent the general investment analysis and economic views of the Advisor, and are provided solely for the purpose of information, instruction and discourse. Please see periodic remarks on the Fund Notes and Commentary page for discussion relating specifically to the Hussman Funds and the investment positions of the Funds.
The foregoing comments represent the general investment analysis and economic views of the Advisor, and are provided solely for the purpose of information, instruction and discourse.
Prospectuses for the Hussman Strategic Growth Fund, the Hussman Strategic Total Return Fund, the Hussman Strategic International Fund, and the Hussman Strategic Dividend Value Fund, as well as Fund reports and other information, are available by clicking "The Funds" menu button from any page of this website.
For more information about investing in the Hussman Funds, please call us at
513-326-3551 outside the United States
Site and site contents © copyright Hussman Funds. Brief quotations including attribution and a direct link to this site (www.hussmanfunds.com) are authorized. All other rights reserved and actively enforced. Extensive or unattributed reproduction of text or research findings are violations of copyright law.
Site design by 1WebsiteDesigners.