GaveKal asks the quadrillion dollar question today. How do stock markets perform in inflationary environments?
Despite a roiling US bond market sell-off, equity investors had until Thursday taken comfort from the Federal Reserve saying it will stay easy for longer and, hey, what’s not to like about a “little” inflation? The idea of this paper is to test the notion that limited inflation can, in fact, benefit stocks.
To that end, let’s separate long-run US stock market returns into two clusters: the first will show returns in periods of structural disinflation (or deflation) and the second, returns in periods of structural inflation. The point of this exercise is to test if these two separate clusters show a statistically significant difference, as measured by equity returns over cash.
To answer the question, it is first necessary to define periods when structural inflation is falling, and also when it is rising. We do this by comparing the short-term trend (last 12 months) of the consumer price index to CPI’s longterm trend (last seven years). If the short-term trend is above the long-term trend (expressed in log terms), we deem it an inflationary period; in contrast, if the short-term reading is below the long-term one it is a disinflationary period. If the reader wants to test any other rule to separate structural inflation from structural disinflation or deflation, we will be happy to do so.
Using Robert Shiller’s market data and official CPI numbers—both going back to 1878—US stock market returns can be separated into two distinct clusters: it turns out that the US market has spent almost exactly half of the last 142 years in an inflationary period and half in disinflation, or deflation.
The black line shows US equity price returns in dollar terms since 1878. The red line denotes the same relationship but in periods of structural disinflation, while the blue line shows the return profile in inflationary times.
• All—and we mean all—of the excess returns from owning US equities in the last 142 years came in disinflationary periods, while no excess returns were achieved during the inflationary times.
• During inflationary periods, shares returned the same as cash, but with far higher volatility. To put it simply, shares are no inflation hedge. It should be noted the US has never faced hyperinflation, but if it did, shares would in this case likely work as an inflation hedge.
• In “normal” inflationary times, buy-and-hold strategies work terribly and investors should aggressively undertake tactical asset allocation.
• Inflationary—or “Keynesian”—periods are perfect for “value” managers and a disaster for “growth” managers since the Shiller price-to-earnings ratio collapsed in every inflationary period.
• Except for the 1929-33 bust, all of the US bear markets took place during periods of structural inflation. In fact, the 1929-33 collapse occurred with outright deflation, or prices actually falling.
The conclusion is simple: the stock market loves disinflation but hates both inflation and deflation for reasons well explained by our preferred Wicksellian methodology (see Stagnation Or Bust?). Consider the following:
• In a deflation, the market rate of interest rises in real terms, while the natural rate of interest collapses, causing the economy-wide return on invested capital to slide below the cost of money. At this point, the only sensible investment is to repay debt as fast as possible, which leads to a collapse in the money supply and further declines in CPI. This is what happened in the US in the 1930s, Japan in the 1990s and in Italy since the euro’s inception in 1999. Each time, banks have gone bust.
• In an inflationary period, one cannot compute the natural rate due to the replacement cost for capital being impossible to gauge, yet always higher than expected. As a result, earnings are overstated, too much is paid in taxes, P/E ratios fall, and by the end companies that are apparently making money go bankrupt.
• In a disinflation, the opposite phenomenon occurs as earnings are understated, not enough taxes are being paid, P/E ratios rise and the market rate, while going down structurally, remains all the time below an underestimated natural rate.
• Since the 1950s, the US CPI has never fallen in a sustained way, so all bear markets have been inflationary. In a previous paper, it was shown that since 1969 all the US bear markets occurred after monetary policy had been “Keynesian” for quite a while (see Managing Money In A Keynesian Environment). Hence, it is fair to ask a simple question: is there a causal link between a Keynesian monetary policy and inflation?
This seems a reasonable hypothesis since the goal of a Keynesian policy is to create a decline in the value of money, which amounts to the same amount of money buying fewer and fewer goods and services.
The next questions must be: where are we today, and where are we (possibly) going tomorrow?
• Where are we now? The US economy is transitioning from an environment of structurally falling inflation to one of structurally rising inflation—it is likely to stay in this in-between stage until June or July this year.
• Where are we headed? This is where the analysis gets properly interesting, for there are three distinct possibilities, as shown in the chart below that records increases in US CPI on a monthly basis:
1) The US inflation spike created by the low inflation rate of a year ago ends by this coming summer and we return to a disinflationary world.
2) The US inflation spike ends yet CPI begins to accelerate. It does so, however, by less than 0.5% on an annual basis, and we remain in a disinflationary world, but only just.
3) The US inflation spike ends, yet the inflation rate accelerates in the following year by more than 1%, giving an annual inflation rate of around 2.4% year-on-year. As a result, the US moves decisively into a structural inflationary period, which will last for the foreseeable future.
The fascinating part is that the expected annual increase for the US CPI in the next 10 years (using the spread between TIPS and classical bonds) has reached 2.23%.
• If expected inflation is rightly forecast by the spread, then we are moving into an inflationary period, which would mean contracting P/E ratios and maybe a bear market. In this case, investors should raise cash to deploy in “serious” currencies, which in this cycle are Asian ones, and in anti-fragile assets (gold), or robust assets (10-year Chinese government bonds). In that case, one should continue to buy TIPS, while selling classical bonds.
• If the expected inflation, as measured by the TIPS-bond spread, falls back to levels of recent years, investors should sell TIPS and stay fully invested in companies that have decent earnings growth. We would be back in a deflationary boom.
• Whether the US economy moves to situation (1) or (2), after August, the uncomfortable fact remains that from March to July, it will be in an “inflationary” period, and markets could be choppy. Buying a few shortdated put options or perhaps buying the VIX index outright, despite its very rapid decay, may make sense for these few months.
I still see the world as disinflationary once we get past this rebound pop:
- MMT is not yet embedded. See the failing attempt to lift US minimum wages.
- China is still on the path to lower for longer growth and this will wipe out commodity prices.
- Declining demographics are everywhere.
- Debt is everywhere.
- Zombie capitalism is here applying margin pressure everywhere.
The current pop in inflation is cyclical. Disinflation is still structural.