A Who's Who Of Awful Times To Invest

Excerpted from John Hussman's Weekly Market Comment:

Unless we observe a rather swift improvement in market internals and a further, material easing in credit spreads – neither which would relieve the present overvaluation of the market, but both which would defer our immediate concerns about downside risk – the present moment likely represents the best opportunity to reduce exposure to stock market risk that investors are likely to encounter in the coming 8 years.

Last week, the cyclically-adjusted P/E of the S&P 500 Index surpassed 27, versus a historical norm of just 15 prior to the late-1990’s market bubble. The S&P 500 price/revenue ratio surpassed 1.8, versus a pre-bubble norm of just 0.8. On a wide range of historically reliable measures (having a nearly 90% correlation with actual subsequent S&P 500 total returns), we estimate current valuations to be fully 118% above levels associated with historically normal subsequent returns in stocks. Advisory bullishness (Investors Intelligence) shot to 59.5%, compared with only 14.1% bears – one of the most lopsided sentiment extremes on record. The S&P 500 registered a record high after an advancing half-cycle since 2009 that is historically long-in-the-tooth and already exceeds the valuation peaks set at every cyclical extreme in history but 2000 on the S&P 500 (across all stocks, current median price/earnings, price/revenue and enterprise value/EBITDA multiples already exceed the 2000 extreme). Equally important, our measures of market internals and credit spreads, despite moderate improvement in recent weeks, continue to suggest a shift toward risk-aversion among investors. An environment of compressed risk premiums coupled with increasing risk-aversion is without question the most hostile set of features one can identify in the historical record.


Recent weeks mark the first time in history that our estimates of prospective 10-year returns on all conventional asset classes have simultaneously declined below 2% annually. We don’t expect a portfolio mix of stocks, bonds and cash to achieve any meaningful return over the coming 8-year period. The fact that the financial markets feel wonderful right now is precisely because yield-seeking speculation and monetary distortions have raised security prices today to levels where they are likely to stand years from today – with steep roller-coaster rides in the interim.


The chart below shows historical instances where overvalued, overbought, overbullish conditions matched current extremes, and where bubble-tolerant overlays (based on measures of market internals and credit spreads) were also unfavorable, and where the S&P 500 had established an all-time high. On less stringent criteria, there would also be additional instances shown in 2010 and 2011.

As I’ve noted before, if one thing was truly different about the yield-seeking speculation induced by QE in the recent half cycle, it is that QE reduced the overlap between overvalued, overbought, overbullish conditions and periods of deteriorating market internals. The current set of overextended conditions is the most severe variant that we define, and unlike several instances in the half-cycle since 2009, we do not have indications from market action and credit spreads that these extremes are likely to be tolerated for long.

Extremes in observable conditions that we associate with some of the worst moments in history to invest include:

  • Aug 1929 (with the October crash within 10 weeks of that instance),
  • Aug-Oct 1972 (with an immediate retreat of less than 4%, followed a few months later by the start of a 50% bear market collapse),
  • Aug 1987 (with the October crash within 10 weeks),
  • July 1999 (associated with a quick 10% market plunge within 10 weeks),
  • another signal in March 2000 (with a 10% loss within 10 weeks, a recovery into September of that year, and then a 50% market collapse),
  • July-Oct 2007 (followed by an immediate plunge of about 10% in July, a recovery into October, and another signal that marked the market peak and the beginning of a 55% market loss),
  • two earlier signals in the recent half-cycle, one in July-early Oct of 2013 and another in Nov 2013-Mar 2014, both associated with sideways market consolidations, and the present extreme.

It’s notable that while these extremes sometimes occurred at the exact market high, other instances were only proximal to those highs, resulting in brief retreats and a final push higher. We don’t know whether the current instance will have consequences similar to the 1929, 1972, 1987, 2000 and 2007 ones, but suffice it to say that these conditions were more notable for their outcomes over the completion of the full market cycle than they were for their immediate outcomes.

*  *  *

We’re quite aware, and slightly entertained, by critics who offer me as a poster-child against evidence-based market analysis or markedly unconventional views – half-way into a market cycle, with prices at record highs, and with reliable valuation measures at obscene levels. These will be great fun to save for later, as they were in 2000 and 2007.  It should be obvious that our challenges in this half-cycle began with my 2009 decision to stress-test our methods against Depression-era data, having correctly anticipated the tech crash, the financial crisis, both of the associated recessions, having moved to a constructive outlook early in the intervening bull market, and when the long-term benefits our discipline were rather indisputable. Given an incomplete half-cycle at speculative extremes, my sense is that assessments of our discipline will change considerably by the completion of this cycle.

*  *  *

When investor preferences are risk-seeking, overly loose monetary policy can have a disastrous effect by promoting reckless speculation and enhancing the ability of low-quality borrowers to issue debt to yield-starved investors. This encourages malinvestment and financial distortions that then collapse, as we saw following the tech and housing bubbles. Those seeds have now been sown for the third time in 15 years. All of this has been in pursuit of what is in fact a nearly nonexistent cause-effect relationship between monetary policy and economic activity.