Every fiscal quarter, publicly traded companies and the securities analysts who cover them engage in a series of time-honored rituals. The rituals follow the elaborate pantomimes of a Japanese Kabuki drama, and the underlying message is almost always the same: Hope Springs Eternal. Consider the chart below, which shows quarterly earnings per share trends for the S&P 500.
How to Speak Kabuki
Here’s how to interpret this chart. The green line represents projected earnings per share for the next twelve months. Basically, analysts project what they think will happen to the companies they analyze in the year ahead, based on the economic context and all the usual industry- and company-specific competitive variables, and those assumptions boil down to the one NTM (next twelve months) EPS figure. This chart shows the composite NTM EPS for all the companies in the S&P 500. For example, on January 31, 2017, the composite estimate for what S&P 500 shares would be worth a year hence (i.e. at the end of January 2018) was $133.17 per share.
The red line illustrates the actual earnings per share for the twelve trailing months, known in finance-speak as LTM (last twelve months). So if the green line represents the hope, then the red line shows the reality, or what actually happened. As you can see, the actual LTM earnings per share for the S&P 500 on January 31, 2018 was $118.08. It wasn’t $133, as the analysts had forecast a year earlier.
Moreover, the chart shows that this “reality gap” plays out in virtually every quarter, year after year. What often happens as earnings season approaches is that analysts start to dial back their earlier predictions to bring them closer to reality. In so doing, they are playing another important role in the Kabuki drama: managing expectations. Earnings season is only partly about the actual growth number; it is also about whether this number is better than, or worse than, what was expected. The downward revision that customarily takes place in the weeks before a company reports has the effect of lowering the bar, so that whatever number the management team actually reports has a better chance of beating expectations.
This Time Is (Sort of) Different
Clear as a bell, right? In summation: expected earnings are usually rosier than the actual figures that come out later, and the elaborate play-acting between management teams (forward guidance) and securities analysts (downward revisions to expectations) contrives to deliver happy surprises to the market.
Except that the script is a bit different this year. Analysts are actually raising – not lowering – their expectations for 2018 earnings performance. Let’s consider the case for Q1 2018. This is the quarter we are currently in, so we don’t know how the companies are actually going to perform. What we do know is that the analysts’ consensus estimate for Q1 S&P 500 earnings as of today is 17.1 percent growth. When the same analysts made their Q1 forecasts at the end of last year, the estimate was 10.9 percent. That’s a big difference! And not just for Q1. The analysts’ Q2 estimate is also significantly higher today than it was seven weeks ago. In fact, the consensus estimate for full year 2018 earnings growth is 18.2 percent, which is 8 percent higher than on December 31, when analysts predicted growth of just 10.3 percent.
That difference – the difference between 18 percent and 10 percent – matters a lot for equity valuation. If companies in the S&P 500 grow their earnings by 18 percent this year, then a similar rise in share prices will not make stocks any more expensive, when measured by the price-earnings multiple variable. That would ease investors’ worries about a potential share price bubble. Double digit price growth matched by double digit earnings growth is arguably the best of all possible worlds for long equity investors.
Here is the one caveat. Much of the apparent optimism in the current earnings projections comes from one single fact – the tax cuts enacted in December that were disproportionately skewed towards corporate tax relief. You can see from the green line in the above chart just how dramatically expectations accelerated right around the time the relief package took shape. More than any other factor, the tax cuts help explain why, contrary to the usual practice, earnings estimates have been raised rather than lowered as reporting dates come closer (though we should also note that a weaker US dollar, should it persist, could also be an earning tailwind for companies with significant overseas activities).
More to Life than Taxes
Investors should take in this seemingly good news with a measure of caution. First, to the extent that the tax relief does make a strong impact on the bottom line (which would be the case for companies that actually pay something close to the previous statutory tax rate, by no means a majority of S&P 500 companies) it will be a one-and-done kind of deal. The growth rate will kick up for one fiscal year cycle of “comps” – comparisons to the previous year – and then stabilize at the new level.
Second, there are other variables in flux that could at least partially offset the tax advantages. Corporate borrowing will be more expensive if (a) interest rates generally continue to rise and (b) credit spreads widen in response to higher market volatility. Companies in more price sensitive industry sectors may have to address issues of how much new inflation they can pass on to their customers. And, of course, corporate top lines (sales) will be dependent on the continuation of global economic growth leading to increased organic demand for their products and services.
Finally, the only way that companies can consistently grow their earnings above the overall rate of GDP growth is through productivity-enhancing innovations to their value chains. There may be a new wave of such innovations just around the corner – or there may not be. How these developments play out over the coming months will determine whether the current rosy predictions of the analyst community play out – or whether they quickly go back to that familiar old Kabuki script of hope and reality.
So, everything’s back to normal, right? The sharp pullback that began with the hourly wage number exactly two weeks ago has assumed its usual V-shape, with 5 straight trading days in the green following the technical correction level of minus 10.2 percent reached on February 8. Just like that silly Ebola freak-out back in 2014, this one looks like it will pass over, a brief squall yielding back to calm seas without so much as a full day spent below the 200-day moving average.
Yields Go North, Dollar Goes South
The good news, for those who prefer their equities portfolios neither shaken nor stirred, is that the continuing rise in bond yields is failing to inject fear into risk-on asset classes. The 10-year Treasury yield broke through 2.9 percent on Wednesday, even as the S&P 500 recorded yet another intraday gain of more than 1 percent. Those inflationary fears would seem tempered, even though Wednesday’s news cycle also served up a core CPI growth number a bit higher than consensus expectations. For the moment, anyway, the stock market seems comfortable enough with higher rates.
Which brings us to today’s big question: what is up with the US dollar? The chart below shows the downward trajectory of the dollar against the euro over the past six months, during which period the 10-year yield soared from just over 2 percent to the recent 2.9 percent.
All else being equal, rising rates should make the home currency more, not less attractive. Investors prefer to invest where returns are higher. Moreover, the dominant economic narrative around the US for the past half-year or so has been positive: above-trend growth, strong corporate earnings and high levels of business and consumer confidence. What’s not to love? Yet, even while the spread between the 10-year Treasury and the 10-year German Bund is 0.45 percent wider than it was six months ago the euro, as seen in the chart above, has soared against the dollar. And the pace has only accelerated since the beginning of 2018. Yes, the dollar jumped ever so briefly as a safe haven mentality took hold two weeks ago, but it fell back just as quickly – even while the 10-year yield reached new 4-year highs.
Supply and Demand, Yet Again
We’re starting to feel like Econ 101 professors around here lately, given how often the phrase “supply and demand” shows up in our commentaries and client conversations. We think it may be the single most important catchphrase for 2018, and in it lies a plausible explanation for that odd relationship between the dollar and bond yields. The supply-demand road inevitably leads back to China.
China’s central bank buys US government debt – lots of it. Chinese foreign reserves exceed $3 trillion, and the vast bulk of those reserves exist in the form of US government securities. Two things happen when Chinese monetary authorities (or any other foreign institution) buy US paper, all else being equal. First, the price goes up, and thus the yield, which moves in the opposite direction of the price, goes down. Second, the dollar goes up because Treasuries are a dollar-denominated asset. You see, it really is all about those supply and demand curves.
Recent evidence (including a weak Treasury auction last week) suggests that Chinese Treasury purchases are somewhat lower than they have been in recent years. That rumor back in early January, though quickly disputed, may have a kernel of truth to it. Foreign buyers indeed may have less appetite for Uncle Sam’s IOUs. Maybe they expect more inflation down the road (which would assume higher nominal rates). Maybe other factors are afoot. Whatever the reasons, reduced demand from non-US sources would indeed have the likely effect of pushing up rates and pushing down the dollar at the same time.
If this is the case, then we may be in for more bumpiness in equities. The S&P 500 digested the move to 2.9 percent very smoothly. We may see in the coming weeks whether the story plays out the same way at 3 percent or more. There will be a truckload of Treasuries coming down the road as we borrow to fund all that new spending and those tax cuts. More supply, in other words, potentially chasing less demand.
Watch the bond market: that was a core theme of our recent Annual Outlook and earlier commentaries in this brief, suddenly volatile year to date. Benchmark Treasury yields jumped on the first day of the year and never looked back. For the first month equities kept pace with rising yields, delivering the strongest January for the broad US stock market since 1987. Then it all went pear-shaped. Yields kept rising, while risk-on investors developed a case of the chills and sent stocks into a sharp retreat. The S&P 500 saw its biggest intraday declines since 2011, and the fastest move from high to 10 percent correction – 9 trading days – since 1980. Investors, naturally, want to know if this is just a long-overdue hiccup on the way to ever-greener pastures, or the start of a new, less benign reality.
The Expectations Game
The chart below shows the performance of the S&P 500 and the 10-year yield for the past 12 months through the market close on Thursday.
What caused that abrupt change in sentiment? Investors seemed perfectly happy to watch the 10-year yield rise from 2.05 to 2.45 percent last September and October, and again from 2.4 to 2.7 percent over the course of January. What was it about the move from 2.7 to 2.86 percent to precipitate the freak-out in stocks? The most widely cited catalyst has been the wage growth number that came out in last Friday’s jobs report; after growing at a steady rate of 2.5 percent for seemingly forever, the wage rate ticked up to 2.9 percent in January. According to this train of thought, the wage number raised inflationary expectations, which in turn raised the likelihood of a faster than expected rate move by the Fed, which in turn led to portfolio managers adjusting their cash flow models with higher discount rates, which in turn led to the sell-off in equities this week.
Algos Travel in Packs
There is a kernel of truth to that analysis, but it doesn’t really explain the magnitude or the speed of the pullback. For more insight on that, we turn to the mechanics of what forces are at play behind the actual shares that trade hands on stock exchanges every day. In fact, very few shares trade between actual human hands, while the vast majority (as high as 90 percent by some estimates) trade between algorithm-driven computer models. The “algos,” as they are affectionately known, are wired to respond automatically to triggers coded into the models.
On most days these models tend to cancel each other out, sort of like the interference effect of one wave’s crest colliding with another’s trough. But a key feature of many of these models is to start building a cash position (by selling risk assets) when a certain level of volatility is reached. Even before the selling kicked in last week, the internal volatility of the S&P 500 had climbed steadily for several weeks, while the long-dormant VIX was also slowly creeping up. The wage number may or may not have been a direct trigger, but enough of these models read a sell signal to start the carnage. Rather than waves canceling out, it was more like crests meeting and growing exponentially. More volatility then begat more selling.
The Case for Promise
So we’ve been given a taste of the peril that can come from higher rates. What about the promise? Here we leave the mechanics of short-term market movements and return to the fundamental context. The synchronized growth in the global economy has not changed over the past two weeks. The Q4 earnings season currently under way continues to deliver upside in both sales and earnings growth, while the outlook for Q1 remains promising. If wages and prices grow modestly from current levels – say, for example, so that the Personal Consumption Expenditure index actually rises to the Fed’s 2 percent target – well, that is in no way indicative of runaway inflation.
This should all be good news; in other words, if the current global macro trend is sustainable, it would strongly suggest that the current pullback in risk assets is more like a typical correction (remember that these normally happen relatively frequently) and less like the onset of a bear market (remember that these happen very infrequently). Higher rates have an upside as well, when they reflect positive underlying economic health. With one caveat.
The Debt Factor
Call it the dark side of the “reflation-infrastructure trade” that caught investors’ fancy in late 2016. The US is set to borrow nearly $1 trillion in 2018, much of which is to pay for the fiscal stimulus delivered in the administration’s tax cut package. That borrowing, of course, takes the form of Treasury bond auctions. A weak auction of 10-year Treasuries on Wednesday is credited for pushing yields up (and stocks down) late Wednesday into Thursday. These auctions, of course, are all about supply and demand. Remember that brief freak-out in early January when rumors floated about China scaling back its Treasury purchases? Supply and demand trends stand to weigh heavily on investor sentiment as the year progresses.
Now, a great many other factors will be at play influencing demand for Treasuries, including what other central banks decide to do, or not do, about their own monetary stimulus programs. Higher borrowing by the US may be offset if overseas demand is strong enough to absorb the expected new issuance. Time will tell. In the meantime, we think it quite likely that the surreal quiet we saw in markets last year will give way to more volatility, and to sentiment that may shift several times more as the year goes on between the peril and the promise of higher interest rates.
Longtime readers of our research and commentary know that we spend quite a bit of time dwelling on the economic metric of productivity. Our reason for that is straightforward: in the long run, productivity is the only way for an economy to grow in a way that improves living standards. Curiously, the quarterly report on productivity issued by the Bureau of Labor Statistics generally fails to grab the kind of headlines the financial media readily accord to unemployment, inflation or GDP growth. So there is an excellent chance that today’s release showing a drop of 0.1 percent in productivity growth for Q4 2017 (and a downward revision for the Q3 number) didn’t show up in your daily news digest. And while one quarter’s worth of data does not a trend make, the anemic Q4 reading fits in with a larger question that bedevils economists; namely, whether all the innovation bubbling around in the world’s high tech labs will ever percolate up to deliver a new wave of faster growth.
Diminishing Returns or Calm Before the Wave?
The chart below shows the growth rate of US productivity over the past twenty years. A burst of relatively high productivity in the late 1990s and early 2000s faded into mediocrity as the decade wore on. After the distortions (trough and recovery) of the 2007-09 recession, the subsequent pattern has for the most part even failed to live up to that mid-2000s mediocrity.
There are two main schools of thought out there about why productivity growth has been so lackluster for the past 15 years. The first we could call the “secular stagnation” view, which is the idea that we have settled into a permanently lower rate of growth than that of the heyday of 25 years or so following the Second World War. The second school of thought is the “catch-up” argument, which says that scientific innovations need time before their charms fully work their way into the real economy. Readers of our annual market outlooks may recall that we closely examined the secular stagnation argument back in early 2016, while the catch-up philosophy occupies several pages of the 2018 outlook we published last week.
The most persuasive evidence made by the catch-up crowd is that both previous productivity waves – that of the late ‘90s – early 00s shown in the above chart and the longer “scale wave” that ran from the late 1940s to the late 1960s – happened years after the invention of the scientific innovations that powered them. Most economists ascribe a significant impact to the products of the Information Age – hardware, software and network communications – in explaining the late 1990s wave. But those products started to show up in business offices back in the early 1980s – it took time for them to make an actual impact. According to this logic, it should not be surprising that the potentially momentous implications of artificial intelligence, deep machine learning, quantum computing and the like have yet to show that they make a real difference when it comes to economic growth.
Productivity and Inflation
The economic implications of productivity tend to be longer term rather than immediate – that is probably why they don’t merit much coverage on the evening news when the BLS numbers come out. After all, the economy is not going to stop growing tomorrow; nor will millions of jobs disappear in one day if another productivity wave comes along with the potential to make all sorts of service sector jobs redundant (a point we make in our 2018 outlook if you’re interested). The lack of immediacy can make productivity debates seem more like armchair theory than like practical analysis.
But productivity (or its lack) does have a lot to do with a headline number very much in the front and center of the daily discourse: inflation. What the BLS is reporting in the chart above is labor productivity: in other words, the relationship between how much stuff the economy produces and how much it costs to pay for the labor that produces that stuff. If compensation (wages and salaries) goes up, while economic output goes up by a smaller amount, then effectively you have more money chasing fewer goods and services – which is also the textbook definition of inflation.
In fact, the BLS notes in its Q4 productivity release that higher compensation was indeed the driving factor behind this quarter’s lower productivity number. Bear in mind that unemployment is currently hovering around the 4 percent level (this is being written before the latest jobs report due out Friday morning), and anecdotal evidence of upward wage pressures is building. An upward trend in unit labor costs (the ratio between compensation and productivity) has the potential to catalyze inflationary pressures.
Keep all this in mind as we watch the 10-year Treasury inch ever closer towards 3 percent. As we noted in our annual outlook, it makes sense to watch the bond market to understand where stocks might be going. And anything related to inflation bears close monitoring to understand what might be happening in the bond market.
Something odd has been going on underneath the main headlines in investment markets over the past twelve months. The big story, of course, is how everything has gone up from US stocks to emerging market bonds to industrial metals, fossil fuels and more. All that is true -- and yet, there has been an unusual lack of correlation between the price movements of asset classes that normally track very closely.
Take equities, for example. In most years, the correlations between different equity asset classes, from US style classes to non-US developed and emerging markets, have been very close. Think of a horse race, with all the horses coming out of the gate together and basically running in the same direction at very similar speeds. This past year, though, the correlations between these asset classes have been very weak -- less like horses out of the gate, and more like letting a bunch of cats out of a box to wander around as each one sees fit.
2017’s Odd Math
Exhibit A is emerging markets. In 2017 emerging market equities rose, and so did the S&P 500. But statistically speaking, there was virtually no correlation between the two asset classes. The statistical measure of correlation is a spectrum from 1.0 (perfect positive correlation) to minus 1.0 (perfect negative correlation). A correlation of 0.0 suggests no tangible connection between whatever moved each individual asset over the time period measured -- each marched to its own set of influences.
In 2017 the correlation between the S&P 500 and the MSCI Emerging Markets index was 0.04 -- zero, for all intents and purposes. The average correlation over the past 5 years for these two asset classes was 0.59 -- statistically relevant positive correlation. 2017 was an anomaly. What the correlation tells us is that the fact of the S&P 500 and MSCI EM both going up that year was more coincidental than it was explained by similar driving factors.
An even odder pairing of wandering assets is seen in the US style classes of growth and value. The average correlation between the Russell 3000 Growth index and the Russell 3000 Value index over the past 5 years was 0.87 -- a very strong positive correlation. For 2017 the correlation was 0.23 -- still positive, but very weak. Correlation that weak between these two assets would make any covariance measures -- like alpha and beta -- statistically useless.
PMs Love Wandering Cats
An astute investor might say: So what? All those equity classes rose in 2017. Do I care how closely two assets are correlated? The answer is yes. It is important because portfolio managers make allocation decisions based in part on the correlation properties of the assets they include in a diversified portfolio. And for these managers, a bunch of wandering cats is actually much more attractive for portfolio inclusion than eight horses running in lockstep. The property of low correlation with other asset classes is value-additive; all else being equal, the manager would prefer combining emerging markets with the S&P 500 when the correlation is zero as opposed to when it is 0.6 or more.
So the key question is this: Did something change so structurally in 2017 as to suggest that correlations between historical birds of a feather (e.g. style and geographic equity asset classes) are moving to a lower plateau? Or are the wandering cats a one-off phenomenon, with the customary high correlations set to return in due course?
We’ll have to see what the numbers tell us as 2018 moves along. In the meantime, we will be testing out some hypotheses. One hypothesis is that the growth in passive investing -- primarily through ETFs -- is a catalyst for lower correlations. ETFs make it easier to trade asset classes wholesale, as opposed to the emphasis on individual stocks employed by traditional active managers. It’s entirely possible that this could lead to more pronounced variations between asset classes as the passive strategies, driven by more frequent short-term trading volumes, propel them in different directions on different days.
Now, why that would have only shown up in 2017 is another question, without a convincing answer. We’ll have a better sense of that a few months down the road. Expect to hear more about this from us as the year progresses.