Re-Defining the Standard of Value

John P. Hussman, Ph.D.
All rights reserved and actively enforced.

Just a note - The Hussman Funds generally pay their required capital gains distributions during November, which I expect to represent a few percent of each Fund's net asset value. The net asset value of each Fund, of course, declines by the amount of the distribution on the ex-day (thus reducing later capital gains liability). Last week's comment includes further information about Fund distributions.

---

While the fanfare accompanying each marginal new high in the market gives the impression that stocks are making important gains, the issue for a long-term investor must always be whether such gains are likely to be retained over the full market cycle.

Investors are eager to overlook the fact that stocks have lagged risk-free Treasury bill returns over the past 8 years, and are instead focused on the gains achieved during the current bull market. Yet even over the most recent 2.8 years since early 2004, the major indices have outperformed Treasury bills by only about 5% annually.

Now, 5% annually, retained over the full market cycle, is a respectable margin over risk-free rates. But here the 5% margin has applied to a bull-market-only portion of this cycle, where bull-market-only gains have historically outpaced T-bill yields by upward of 20% annualized, on average. Worse, with stocks strenuously overbought and trading at rich multiples on record profit margins, there's a very slim potential for investors to retain that 5% margin they've earned above T-bill yields during the past few years.

Think about it this way. Suppose that the market achieves a 15% total return over the coming year, finally followed in the next year by a shallow one-year correction of just 16%. The combination of that mild "cycle completing" decline, along with the passage of time, would still be enough to put T-bills ahead of stocks not only for the period since early 2004, but for what would then be a full decade since 1998. The only advantage over T-bill yields that this bull market has a reasonable shot of retaining is the gain that occurred between late 2002 and early 2004 (even that portion would largely vanish in anything but a mild "cycle completing" bear market).

It's tempting to think that the reasonableness of a defensive stance here depends on whether the market will make further new highs in the near term. But that ignores the fact that the only way to retain further gains here - without giving them back during the completion of this market cycle - is to exit at even higher prices. And the trouble is that if investors accept risk here, despite an overvalued, overbought and overbullish investment environment, they can't use any of those conditions as exit criteria. A move to overvaluation won't get them out, because the market is already overvalued.

Given that current valuations don't provide an adequate investment basis for taking market risk, and market action is already overextended, the only plausible exit signal for a speculative position has to be a deterioration of market action. Logically, then, the case for taking risk at present has to be that either a) the market will not experience any material correction in the coming years despite current valuations, or b) that the market is likely to move so much higher here that even a deterioration that triggers an exit will still provide a higher sale than at present. Those aren't impossible, but we certainly wouldn't risk a great deal of capital on them.

Still, the market's general action is favorable overall, and we have to be conscious that valuation does not hold much sway over short-term market fluctuations. Presently, Strategic Growth Fund holds a fully hedged investment stance plus just under 1% of assets in index call options. If stocks can clear their current overbought condition without a major deterioration in market internals, we'll have a reasonable basis for taking modest additional speculative risk (primarily by increasing our small call option position to perhaps 2% of assets). The reason to use calls is that implied option volatilities are very low, so option premiums are cheap (with little time decay). Also, I have no particular confidence that the market won't complement the other side of this "high pole" advance with a similarly linear decline.

In short, valuations remain rich, and despite my disagreement with the "Goldilocks" thesis, the fact is that market action currently suggests a willingness of investors to speculate. I do view the market as strenuously overbought, so it's likely that we'll observe a pullback at least on the order of a few percent at some point in the coming weeks, which will provide a better entry point for any speculative positions. It seems unlikely that further gains from these levels will be retained over the complete market cycle, but as long as market action is favorable we're willing to accept a modest speculative exposure to market fluctuations using limited-risk call option positions, and to increase those positions modestly on periodic market weakness.

Echoes

With the Dow and S&P 500 at similar levels to their 2000 peaks, we're about high enough to hear echoes across the canyon. The "Dow 36,000" bit is starting to get ink again from the press, which is kind of scary. We're also starting to hear murmurs about a shadowy "plunge protection team" - both from bulls hoping that someone is standing there to save them from potential losses (investors believed the same thing in '29), and from bears believing that the persistence of the recent advance is a sign of manipulation.

I don't really give much weight to these arguments. It's certainly possible to manipulate a thinly traded security or even spike a futures market from time-to-time. But in a market that regularly trades nearly 2 billion shares a day just on the NYSE, it would require implausibly large concentrations of capital to support the market with purchases in the face of a sustained willingness of investors to sell. Even the total amount of bank reserves overseen by the Federal Reserve is just $43 billion, in an economy with a GDP of $13 trillion and a similarly large stock market capitalization.

What we observe as persistence and uniformity is a standard feature of markets where speculators have acquired a "taste" for risk-taking. We can measure it, we're seeing it at present, and it's capable both of ending abruptly and of continuing for a while. My impression is that this speculative mood is less reflective of manipulation than it is the adoption of a common "theme." We saw it with the "new era / dot.com" theme 6 years ago, and we've got it with Goldilocks today. Still, none of that changes the fundamentals, nor the fact that markets have a strong tendency to fluctuate around fundamentals - both above and below - over time.

Re-defining the standard of value

Let's talk about those fundamentals. Currently, the S&P 500 trades at over 18 times record earnings on record profit margins. Earnings for the S&P 500 are also at the top of the long-term 6% growth channel that connects prior earnings peaks going back nearly a century. Historically, when earnings have been anywhere near that 6% peak-to-peak trendline, the P/E multiple on the S&P 500 has averaged about 9 or 10.

Similarly, the S&P 500 trades at a price/revenue ratio of about 1.5 currently, compared with a ratio that has historically fluctuated between about 0.5 and 1.0, with an average of about 0.8. With few exceptions, the valuations that we've observed over the past decade are valuations that have only been observed over the past decade.

So shouldn't we "adapt" our valuation measures to reflect the higher level of recent valuations? More generally, rather than looking at say, the price/peak earnings multiple on the S&P 500, wouldn't we get a better measure of valuations if we adjusted that ratio based on its average level over the most recent decade or so?

Short answer: No. If we look at the price/peak earnings multiple on the S&P 500, for example, we find an 82% correlation since 1950 between that multiple and the subsequent 10-year total return for the S&P 500. Higher valuations predictably, reliably, and strongly imply lower long-term returns for stocks. In contrast, if we "adapt" the level of valuations to recent history by measuring the P/E ratio relative to its prior 10-year average, we immediately wipe out about four-fifths of our ability to explain subsequent 10-year market returns. If we "normalize" the P/E ratio using its 20-year or 30-year average, it still loses explanatory power. The attempt to "adapt" the standard of value to recent history simply destroys information.

Similarly, if we redefine value using the "Fed Model" approach, and divide the corresponding earnings yield by the 10-year Treasury yield, we end up entirely destroying the power of P/E ratios to explain subsequent long-term returns. That's not a surprise, because for most of history, interest rates were at or below current levels, earnings growth was at least as strong, and yet valuation multiples were persistently lower.

The Fed Model does not describe a historically nor theoretically valid relationship between interest rates and earnings yields, but is instead the artifact of an "omitted variable": a two-decade collapse in the risk premium that stocks are priced to deliver. Most of the decline in earnings yields from the early 1980's to 2000 (on which the Fed Model is based) represents a move from extreme undervaluation of stocks to extreme overvaluation, not a "fair value" relationship between earnings yields and interest rates. It's obvious from the longer historical record (as well as basic finance theory) that the relationship between earnings yields and interest rates is not anything close to one-to-one, as the Fed Model implies.

As I've detailed in prior weekly comments, the Fed Model (which analysts increasingly call the "Capitalized Earnings Model" despite having no basis in sound finance), doesn't correlate well with subsequent returns, isn't anything close to a proper discounting model, ignores the extreme gap between stock and bond durations at present, is corrupted by an omitted variables bias, and is based on a data series - forward operating earnings - that didn't exist prior to about 1980 (and for good reason).

That said, suppose we allow for the possibility that 18 is the new 11. Historically, the price/peak earnings multiple on the S&P 500 has had a median of 11, a level that we observed, for example, at the 1990 market trough. But suppose that valuations simply deserve to be higher here. Suppose the current P/E of 18 is "just right" (Cough). Well, even in that case, we would still expect stocks to fluctuate around that norm over time.

Suppose then, that earnings continue to grow along their peak-to-peak growth trend of 6%, but that perhaps 4 years from today the P/E on the S&P 500 will briefly dip to 16 times those future record earnings. Well, let's do the math. Given the current dividend yield of 1.8%, the annualized total return on the S&P 500 over that 4-year period would be just [(1.06)(16/18) ^1/4 + .018(18/16+1)/2 - 1 = ] 4.84%. That's still less than T-bill yields or 4-year Treasury yields.

In other words, it's not enough to argue that "the standards of value have been raised." Unless investors are also willing to assume that the market will no longer fluctuate around those standards, it's very difficult to conclude that stocks are priced to deliver satisfactory long-term returns.

There is, in my view, no compelling evidence to support such a change in the standard of value. What we do observe is a market that has attained high multiples on record profit margins, and now requires the permanent maintenance of both in order to achieve even moderate long-term returns. It's probably safer to assume that markets will continue to fluctuate.

We certainly don't require a return to historical norms like 11 times peak earnings in order to take an un-hedged investment stance. Indeed, we removed 70% of our hedges in early 2003 even though we never reached historical average valuations (much less the lower medians). Still, unless we assume away the tendency of markets to fluctuate, it's difficult to make a case that further market gains will be retained over the complete market cycle.

Again, market action does suggest that investors have little aversion to risk (for now), so we are willing to accept modest and carefully limited speculative exposure (particularly call option positions). Those positions have to be constantly managed as day-to-day market action changes, in order to offset the impact of time decay and to capture "gamma." If the market experiences a reasonable short-term selloff without much deterioration in market internals, we may boost those positions to as much as say, 2% of assets. In any event, it's important to always distinguish between investment positions and speculative ones. It's unfortunate when people take aggressive positions at overvalued market highs, in the belief that they are investing.

Market Climate

As of last week, the Market Climate in stocks was characterized by unfavorable valuations and moderately favorable market action. Short-term, the market remains overvalued, overbought and overbullish - a combination that has historically made stocks vulnerable to negative short-term returns, on average, even when market action has been favorable.

Adding to the general impression of an overextended stock market, Vickers notes that corporate insiders of stocks traded on the NYSE ramped up their selling activity last week to 7.81 shares sold for every share bought. On the Nasdaq, the insider sell/buy ratio shot to 6.36. Both ratios have more than doubled over the past few weeks. One wonders, if corporate insiders have so little optimism about their own shares, why should investors?

In any event, we're holding our risks fairly close to the vest. At present, we are much more willing to accept "basis risk" - a difference in composition between the stocks we hold and the indices we use to hedge - than we are to accept exposure to general market fluctuations. That potential for our holdings to perform differently than the major indices is currently our primary risk as well as our primary source of expected returns. I continue to expect stock selection to be the primary driver of Fund returns in the months ahead. The performance of our stock holdings is broken out separately in our most recent annual report.

In bonds, the Market Climate last week was characterized by modestly unfavorable valuations and modestly favorable market action. The overall shape of the yield curve appears to price in a modest decline in interest rates about 2-years out, followed by a gradual normalization. Statistically, however, the current shape has historically been associated with a fairly parallel upward shift in yields across all maturities - in this case, toward a relatively flat yield curve at a level of about 6%. So there is a clear difference here between the happy scenario that market participants seem to expect, and what has actually happened when market participants have expected that happy scenario.

In my view, it is still far from clear that the economy will enjoy a "soft landing" with muted inflation. But we need not take a view on that. The prevailing valuations and market action in stocks and bonds are sufficient to guide our investment stance. The Strategic Growth Fund remains well hedged, but continues to manage a small position in index call options to provide a modest exposure to market fluctuations. The Strategic Total Return Fund continues to carry a duration of just under 2-years, mostly in Treasury Inflation Protected Securities. On last week's strength in precious metals shares, I clipped off a few percent of our precious metals holdings in order to hold our overall exposure to about 20% of assets in the Strategic Total Return Fund.