Hurricane Irma was first born as a loose collection of clouds that gathered into a tropical storm off the African coast. Thousands of miles west, it was blue skies and sunshine in South Florida, where I was taking a break. Days later, Irma was slamming into the Caribbean, and I was scrambling to get out of dodge. In retrospect, Irma will go down as the most powerful Atlantic storm in recorded history. Likewise, in the current market environment, it feels like things just couldn’t be better. And it is no fun to be a Cassandra about stock prices when they continue to hit new highs. Nonetheless, considering risks that may seem very far away is sometimes the prudent thing to do.
Investment professionals worldwide are well aware that equity valuations are overextended. So are corporate executives. Deloitte’s Q2 CFO Signals report found that 78% of surveyed CFOs believe that equity markets are overvalued. Some professionals, including MRP, are increasingly concerned that the risk of a major cyclical market decline is growing. On top of extended valuations, the list of disturbing signs includes the aged bull, extensive retail / high net worth portfolio leverage, trading automation technology, overly bullish sentiment, insider selling, and uncertainty about the Fed.
While the exact end date of this long bull market is not knowable, there are many good reasons to be concerned about a developing storm beginning in the fourth quarter of this year, and extending well into 2018. But, as storm-trackers would say, the chronological “cone of uncertainty” is wide. MRP believes the catalyst will most likely be a change in a key variable that drives valuation. But first, a quick look at how high is high.
The most widely followed measure of valuation is the P/E of the S&P 500. Price is always the current price, but analysts point to different P/E numbers depending on what E is used as the denominator. There is the trailing P/E, wherein E equals the sum of each EPS for the past four quarters, the forward PE which references the sum of the EPS projections for the next four quarters, and the Cyclically Adjusted P/E (CAPE) which utilizes the average of 10 years of earnings adjusted for inflation.
The nominal trailing 12-month P/E ratio has an historic average of approximately 14-15. But nominal averages can be misleading. The numerator (price) and denominator (earnings) can fluctuate, distorting the historic data for the mean P/E. For instance, the ratio can become skewed upward even in a bad market when earnings may be declining faster than price; a good example is May 2009 when the S&P’s P/E reached a maximum level of 124 in the depths of the Great Recession. The converse of this, a P/E becoming skewed downward by earnings growth outpacing price in a good market, also happens. By identifying and removing the outliers via the interquartile range rule, in this case, the highest 4.5% of observations as well as an equal portion of the lowest values for balance, we calculate a trailing P/E ratio of about 24 against a normalized mean of 17.7, almost exactly the same as the median of 17.8. The current trailing P/E exceeds this mean and median by more than 25%.
The forward P/E is elevated as well. Wall Street is predicting a forward P/E just shy of 19 for this year and close to 17 in 2018, based on EPS estimates of $127.14 and $144.88, respectively. The historic mean forward P/E is 15.9, but from 1960, with outliers accounted for, the mean is 17.4.
Also known as the Shiller PE ratio, the CAPE is currently around 30, nearly double the long-term average of 16.7. The 5-year CAPE is equivalent to 23.6 compared with the long-term average of 18. In MRP’s model, containing data since 1960, the CAPE’s normalized mean is equivalent to 20.27. In a recent interview, Robert Shiller himself drew comparisons between our current market and the crash of ’29: “The market is about as highly priced as it was in 1929… from the peak to the bottom, it was 80 percent down. And the market really wasn’t much higher than it is now in terms of my CAPE ratio.” Shiller also noted that “The U.S. has the highest CAPE ratio of 26 countries. We are number one.”
Another valuation ratio worth looking at is the total market cap to Gross Domestic Product. A ratio between 75% and 90% is typically considered to represent a fair valued market. As of September 13th, the ratio is currently over 135%.
The table below shows the decline that would result from a full mean reversion of the market’s valuation stats.
So, the case for valuations being historically high is solid. A mean reversion of valuations could result in one of the more severe market corrections in the record books. But experience shows that richly valued markets can stay richly valued for long periods of time. Indeed, they can become even more richly valued. To be too negative about the market at this point simply because of overvaluation could be very costly. As John Maynard Keynes pointed out, “The market can stay irrational longer than you can stay solvent,” which gets to another issue on the horizon: this market in particular is indeed testing simple longevity records.
The old age of this bull market against the historical record of the economy has been impressive. The average complete stock market cycle lasts about 69 months with the average period of expansion lasting 58 months and the average period of contraction about 11 months. This current bull, the second longest in history, has now avoided bearish territory for 101 months. It is the 12th bull market since the end of World War II, and the other 11 market expansions have ranged in duration from a little over a year to 153 months. The latter record holder occurred in the ’90’s, and it is now the only previous bull market run that this one has so far not surpassed.
It is also noteworthy that, like human life expectancy, bull markets have been lasting longer. Prior to the mid-70s, they lasted, on average, only 42 months. Subsequently, the average has expanded to 73 months. But the current bull now is already 1/3 longer than that updated average.
None of this has deterred many investors though, who remain very bullish according to the latest AAII Sentiment Survey report, released on September 14. Investor optimism has only slightly retreated from its third-highest level of the year and pessimism recovered from its 17-month low. The proportion of bullish investors increased by 10.8% percentage points to 41.3% (38.5% is the historical average). On the other hand, the bearish group fell by 8.6 percentage points to 27.2%. The broad shift reflects high expectations of further gains, despite the ongoing drama in Washington. Interestingly, equity markets remain one of the smallest external risk factors to CFOs in Deloitte’s aforementioned CFO Signal report. While this seems to be a vote of confidence for markets, it could very well be dangerously misplaced when contrasted against current technical levels and historic precedent.
But it appears not all corporate insiders are feeling so bullish, or at least some are more bearish than they are leading on. According to the Wall Street Journal, the ratio of insider buying to selling fell to a 29-year low earlier this year.
Yet another cause for concern is the volume of assets being held and managed by robo-advisers. These automated advisers assess an investor’s risk appetite and allocate funds based on algorithms and quantitative analysis. Because of their popularity, Vanguard has almost quintupled its assets since 2015, rising from $17B to $83B, while Betterment has seen growth just shy of 700% over the same time frame. Meanwhile, algo trading by hedge funds has grown exponentially. The most prominent risk is that algorithmic trading at such scale remains untested in a true financial crisis. This technology-driven strategy has flourished during the long bull market, but could just as easily be the death of it, or at least exacerbate a correction if algorithms unexpectedly trigger a chain reaction. One relevant precedent to this scenario is the “Quant Quake” of 2007, when a sudden liquidation by a Goldman Sachs multi-strategy fund or proprietary trading desk caused algos to run awry, igniting fire sales in similar quantitatively-constructed portfolios. Goldman saw three quarters of its assets wiped out within a five-year span, as a result.
Meanwhile, debt used to buy stocks has soared to levels last seen in the periods leading up to the last two market crashes. At the end of July 2017, NYSE margin debt was at an all-time record high of $549.9 billion, or 2.86% of GDP. Similarly, the margin risk indicator reached 2.78% in the year 2000 before tumbling as the dotcom bubble burst, and spiked again in mid-2007, foreshadowing the eventual market crash. Relatedly, non-margin loans collaterized by stocks and bonds have surged to at least $100 billion for the biggest brokerages, up exponentially since 2009’s financial crisis. These additional factors could exacerbate a fundamental-driven correction if markets tank and the value of existing collateral shrinks. The borrowers would have to liquidate in order to meet their margin calls.
Many have pointed out that bull markets do not die of overvaluation or old age. This is true. And the various other signs cited above are just that – signs. Individually, they are not likely to cause a big market drop, although they could make a correction worse than might otherwise have been. Usually some disruptive force triggers the reversal from a bull to a bear market environment. More often than not, it is a shift in monetary policy, from stimulative to restrictive.
The Fed conducts policy based on its dual mandate: promoting price stability and maximum sustainable employment. The Fed’s actions, as always, depend on economic data. In that regard, the latest numbers on real growth should be showing the Fed that concerns about renewed economic weakness are misplaced. US GDP growth was just revised upward for the second quarter to a 3.1% annual rate versus 1.6% in the first quarter.
The most recent ingredient added to the mix has become the economic impact of Hurricanes Harvey, Irma and Maria. The combined destruction is estimated to cost well over $200 billion, and subsequent economic slowdown in impacted areas of Texas and Florida (the second and fourth largest state economies in the US, respectively), as well as Puerto Rico will likely cause a mirrored negative reaction in the 3rd quarter GDP figure. Goldman Sachs has already downgraded their estimate by as much as 1% in light of severe energy, manufacturing, and financial disruption. The Atlanta Fed’s GDP Now model is indicating a 2.2% growth rate for the third quarter versus the peak forecast of 4.0% on August 3rd.
But when the stimulative effects of rebuilding finally kick into gear, the third quarter slump could be made up for by resilient GDP growth in Q4 2017 and Q1 2018. These numbers, though, will not be available until the end of the current year, well after the key and final 2017 Fed meetings of December 12th and 13th. Meanwhile, several other sources have been raising their forecast for worldwide economic growth.
Inflation, on the other hand, has remained below the Fed’s 2% target. The recently reported “Core” PCE deflation now stands at 1.3% versus a year ago. The latest Consumer Price Index (CPI) for August is up just 1.9%, down from the 2.7% hit in February – but still above June’s 1.6% reading and 0.0% two years ago. MRP has characterized this inflation lull as an “intermission”. We believe inflation acceleration will soon resume, as the Q4 rebound in economic activity, higher oil and commodity prices, and the weaker dollar kick in. This increase in growth may even continue expanding beyond 2017, as the possibility of major tax reform legislation is looming in congress.
Meanwhile, the labor market remains tight. Even as the most recent jobs data reflected wage growth lag and non-farm payrolls increasing less than expected, unemployment has continually stayed below 5%, defying the Philip’s Curve model used to illustrate the inverse relationship between inflation and unemployment.
Recently, following this month’s FOMC meeting, Fed Chairwoman Janet Yellen expressed some uncertainty while addressing both indicators, stating that the Fed “may have misjudged the strength of the labor market, the degree to which longer-run inflation expectations are consistent with our inflation objective, or even the fundamental forces driving inflation.” While this seems somewhat foreboding for interest rate hikes, she followed it with her belief that low inflation is “probably temporary” and that while the share of prime-age workers (age 25-54) is below optimal levels, this may be a new normal. Her most telling statement was an affirmation that she still sees the Fed on track to continue down it’s planned route: “Given that monetary policy affects economic activity and inflation with a substantial lag, it would be imprudent to keep monetary policy on hold until inflation is back to 2 percent.”
William McChesney Martin, a Fed chair back in the day before most of today’s traders were even born, once remarked that the job of the Federal Reserve is “to take away the punch bowl just as the party gets going.” Albeit, more of a “slo-mo” meaning than a sudden taking away. The rate hike cycle is already underway. The central bank began a program of interest rate normalization with four rate hikes beginning in December 2015. As of September’s FOMC meeting, 12 of the 16 Fed members still anticipate a fifth hike before the end of this year, and 11 of the 16 believe there will be three more in 2018.
According to CME’s Fed Watch tool, futures traders are starting to accept the Fed’s plan for one more hike this year, indicating a 71.9% likelihood that the Fed executes a rate hike in December. However, assuming the Fed continues its pattern of raising rates by only 25 basis points, the futures market is still skeptical of three hikes in 2018, anticipating only one. This disconnect is nothing new. For the last several years, the Fed has repeatedly laid out a path for interest rate normalization, which traders bet would never happen. Time and again, the traders were proven right as the Fed postponed rate hikes.
Inflation could get a boost from the Harvey aftermath. The aforementioned rebound may continue to be buoyed through September by persistently high gasoline prices. In the second week of the month, 9% of US refining capacity remained offline. A factor that must now also be weighed is the possibility of ethical pushback against raising the cost of borrowing while major disaster recovery is underway.
So far, the Fed has not backed off from its stated goal of shedding $2 trillion worth of assets from its $4.5 trillion balance (25% of GDP) by 2022. Most observers agree that the overvaluation of stocks and bonds, and the extended duration of this bull market have been partly driven by the quadrupling of the Fed’s balance sheet and a prolonged period of low interest-rates. The Fed will begin this balance sheet reduction in October. No one knows how this unprecedented development will affect market valuations.
One factor that may induce a more hawkish stance is the Fed’s own concern about the valuation of financial markets. FOMC minutes from the July meeting revealed that many officials at the central bank believe the stock market is too expensive. If the Fed feels that stocks are running hot, it may be compelled to act to stifle a growing equity bubble. Regardless, the markets and the Fed have clearly been growing more and more disconnected. The disconnect could lead to a shock for traders who have grown used to the central bank continually capitulating and resetting its trajectory for interest rate normalization. But as the GDP numbers bounce back from the storm-depressed third quarter and inflation continues to move higher, the Fed is likely to reaffirm plans to continue raising rates, changing further the discount rate used in the present value equation. That could ultimately be the catalyst for the correction that triggers a mean reversion of P/Es and a stormy market environment in the year ahead.