July 6, 2004
Stronger Earnings Are Likely. So What?
Many analysts would be embarrassed of the predictions they regularly make, and the indicators they regularly cite, if they would take ten minutes to analyze the historical data. But then, portfolio managers usually cut their teeth by analyzing companies and the economy based on relationships they are taught are important, rather than questioning whether those relationships actually exist in the numbers. That's not really their fault – they were taught by professors and analysts who don't look at the data either. If you poll 100 academic economists, I'd put money on it that fewer than half could tell you the current GDP of the U.S. within $3 trillion of its actual value*.
This is why analysts still believe that the Federal Funds rate has anything to do with the quantity of bank lending except during liquidity crises (the link between bank reserves and lending was eliminated in the early 1990's when reserve requirements were dropped on everything but demand deposits).
It's why analysts believe that government deficits predictably raise the level of interest rates, and that money growth is closely correlated with inflation (it's growth in unproductive forms of spending, regardless of how that spending is financed, that forces structural shifts in the level of interest rates and inflation on account of government policy).
It's also why analysts believe that short-term S&P 500 earnings growth is actually correlated with returns on the S&P 500 (the correlation between year-over-year earnings growth and annual total returns on the S&P is literally 0.0). That's not to deny that stock prices and earnings are related over the long-term (otherwise P/E ratios would be useless). Rather, it's to emphasize that annual fluctuations in S&P 500 earnings growth are entirely meaningless in driving the S&P.
If you're looking for considerations that have a large impact on investment returns, here's the dynamite: changes in market P/E ratios (particularly from extreme levels to historical norms or opposite extremes) can strongly affect the returns that investors earn over periods as long as 10-30 years. Fluctuations in earnings growth? Not so much. Whether one looks at S&P 500 earnings over 10, 20, 50 or 100 years, earnings have fluctuated within a very well-defined 6% long-term growth channel. Stocks are not a claim on a single year of earnings, and short-term earnings fluctuations around their long-term trend (e.g. peak-to-trough or trough-to-peak movements within that growth channel) have very little relationship with changes in the S&P 500 over the same period.
It's true that stock price changes have some ability to predict earnings. Over the past 50 years, there's been a slight positive correlation (0.12) between the annual return on the S&P 500 and the growth in S&P 500 earnings over the following year. But other indicators are much better for anticipating earnings. For example, the ISM Purchasing Managers Index has a meaningfully positive but not reliably tight 0.56 correlation with S&P 500 earnings growth over the following year. Interestingly, the PMI has a negative correlation with subsequent S&P 500 returns (-0.36). The fact that these correlations go in opposite directions underscores that short-term earnings strength and stock returns don't reliably go hand in hand.
[Geeks note – one might infer that the opposing correlations with the PMI imply that strong earnings accompany weak stock returns. In other words, if A implies both B and C, then B must accompany C. That inference would only be true if both correlations with A were 1.0 or –1.0].
As for valuation metrics, the price/peak-earnings ratio outperforms other commonly used ratios (price/reported earnings, price/operating earnings – though you have to estimate these if you want to go back beyond a couple of decades, price/dividend, price/book, and price/revenue), displaying a –0.34 correlation with S&P 500 returns over the following year. Even this is a fairly weak relationship and is not strongly predictive. Valuations are generally important only for their implications about long-term market returns.
In my view, the quality of market action (an indicator of investors' willingness to take risk) matters profoundly to whether an overvalued market declines, or simply becomes more overvalued. This is why I don't make forecasts. Even at high valuations, there is always the potential for investor risk preferences to shift. While I believe we can adequately measure those shifts, we can't forecast them. So we align our investment positions with the prevailing condition of valuations and market action at each point in time.
Prospects for earnings growth
Last week, the Federal Reserve opted for the expected quarter-point hike in the Federal Funds rate. Looking at history, S&P 500 earnings grew by an average of 19.6% in the 12-months following an initial rate hike, with negligible returns in the S&P 500 over the following 3 and 6 months, and below average returns over the following year. If the Market Climate was unfavorable at the time of the initial hike, there were no net gains on average for the S&P 500 over 3, 6 or 12 month horizons (though a shift to a favorable Climate occasionally made market risk worth taking over some intervening period). As usual, none of that should be taken as a forecast, but it does underscore the fact that earnings growth does not directly translate into investment returns.
Given the historical record, the still relatively strong ISM figures, and other economic factors, it's probable that S&P 500 earnings have not peaked. Though reported S&P 500 earnings (don't get me started on operating earnings) are at a new high of about $52, the current peak of that 6% long-term growth channel is closer to $62, which would require robust profit margin expansion from here, but is not unattainable. Again, that's about a 19% difference, so one doesn't need extraordinary assumptions to accept the prospect of further earnings growth here.
Though the current price/peak earnings multiple is over 21 here, it's tempting to ask, “Well then, if attainable earnings are $62 on the S&P, isn't the real price/peak earnings multiple just 18?” The issue here is that if you base multiples on “attainable” or “projected” earnings, you'd better compare the resulting multiples with historical norms calculated the same way. While the average price/peak earnings multiple is about 14 historically, the average price/“top of channel earnings” multiple is just 12.
So you can compare 21 with 14, or 18 with 12. Either way, valuations are well above average, and earnings multiples are by far the most charitable of all valuation multiples because profit margins and return on equity are currently quite high (making P/Es less extreme than price/revenue and price/book ratios) while dividend payout ratios are quite low (making P/Es less extreme than price/dividend ratios).
June employment probably won't derail the Fed
Last week's employment report came in decidedly weak. As a result, the notion immediately surfaced that the Fed might discontinue its program of rate hikes altogether, or at least postpone it meaningfully.
Don't count on it. If you look at the year-over-year CPI inflation rate, it's already above 3% (it's above 4.5% on the PPI). Even with fairly benign numbers in the upcoming July and August reports, that inflation rate will probably rise to 3.25% and 3.4%, respectively. In other words, the simple change in the inflation rate is already set to outstrip the pace of Fed tightening, even without velocity effects which will probably accelerate short-term inflation pressures. The hawks on the FOMC won't like that, and they were already pushing for more aggressive action months ago.
In short, be careful about what you believe, and when in doubt, stare at the data. At present, the risk of contraction in valuation multiples (particularly given still unfavorable market action) outweighs the very probable growth in earnings here. For those inclined to make predictions, the same strength in the PMI that argues for earnings strength also argues for subpar stock returns, as do the high level of valuations, unfavorable Market Climate, heavy insider selling, advisory sentiment, and other factors.
I don't like making predictions either way (and they're not required anyway). Very simply, if we observe a favorable shift in market action, that potential pressure on valuations will become less important, and despite high valuations, we would be willing to accept at least some speculative exposure to market risk. At present, we remain fully hedged in stocks.
As of last week, the Market Climate in stocks remained characterized by unusually unfavorable valuations and modestly unfavorable market action. Despite the fact that I view the fundamentals of both stocks and the economy with suspicion, those views don't compel other investors to agree (if they did, there would never have been a dot com bubble). If investors adopt a robust preference to accept risk, there is very little in the way of evidence and intellectual debate that will dissuade them.
As always, it's the quality of market action, not the extent or duration of a particular market fluctuation, that conveys information about risk preferences. We've seen very dull market action for a while here, but if it resolves into a broad pattern of accumulation across a wide range of industries, particularly on expanding volume, we would be inclined to lift a portion of our hedges. That's certainly no sort of prediction, but it is important for shareholders to understand that we are not attached to any particular market outlook, and that there is always evidence that could emerge to shift the Market Climate we identify.
For now, the Strategic Growth Fund remains fully invested in a broadly diversified portfolio of stocks, with an offsetting hedge to remove the impact of market fluctuations from the portfolio. What remains is “active risk” – the risk (and potential) for our stocks to behave differently than the broad market. When we are in a fully hedged position, this is always our primary source of risk, as well as our primary source of expected returns.
In bonds, the Market Climate is characterized by modestly unfavorable valuations and unfavorable market action. I continue to view Treasury Inflation Protected Securities as a relatively attractive alternative to straight bonds. In my view, inflation risks and pressures remain much higher than widely assumed, and while I certainly believe that we are likely to observe higher defaults and eventual economic softness that could make Treasury securities a safe haven, I don't believe that we are at that point yet, and we don't observe enough evidence on Market Climate considerations to take a significant exposure to straight Treasuries.
The fact is that the economic environment is bifurcated depending on whether you look at income statements or balance sheets. From an “income statement” perspective, we're likely to observe near term growth in earnings and GDP, but also inflationary pressures as monetary velocity presses higher. From a “balance sheet” perspective, the massive U.S. current account will continue to weigh on longer-term growth prospects for gross domestic investment, and heavy levels of consumer and corporate debt create real risks for a fresh “deleveraging cycle” in the economy, with its attendant defaults. The issue is one of timing, and at this point, the evidence continues to lean toward near term economic strength and inflation pressures, with balance sheet problems exerting themselves on a longer horizon. So while I do believe we'll find ourselves in a substantially more aggressive bond market position perhaps several quarters from now, we still have insufficient evidence to take that position here.
The Strategic Total Return Fund continues to carry a duration of about 3.25 years, primarily in TIPS. Our investment position in precious metals shares is now about 14% due to a combination of appreciation and modest purchases. As such, that position will probably continue to account for much of the day-to-day fluctuation in the Fund. Our strongest source of potential return here would be downward pressure on the U.S. dollar. Our primary risk would be a substantial increase long-term real interest rates, which might help to support the dollar despite large current account imbalances and revaluation pressures on the Chinese and Japanese currencies. If robust, sustained, and long-term GDP growth were likely in the U.S., the risk of higher real interest rates would be more pointed than it appears at present.
In short, though I don't believe that straight Treasury securities are priced to deliver substantial returns in the near term, our ability to hold various bond market alternatives – primarily TIPS and precious metals shares here – substantially expands our flexibility to navigate even an environment featuring inflation pressures and Fed tightening.
* $11.5 trillion
|Home | The Funds | Open an Account | Account Access | Research & Insight | Site Map|
For more information about investing in the Hussman Funds, please call us at
513-587-3440 outside the United States
Site and site contents © copyright 2003 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.