Are Financial Markets Priced For Secular Stagnation? by Toby Nangle, ColumbiaManagement
- The idea that Western economies may have entered a period of secular stagnation has been attracting an increasing amount of attention.
- If this thesis proves correct, we would expect investment grade credit to produce decent excess and total returns as discount rates fell further in such an environment.
- Equity valuations look consistent with some probability of secular stagnation, and could therefore still deliver generous returns if global growth recovers.
By Toby Nangle, Head of Multi Asset Allocation and Zach Pandl, Portfolio Manager and Strategist
The idea that economies may be undergoing a long period of slow growth has been attracting an increasing amount of attention. Data shows fairly definitively that potential growth has been low for a number of years, including years preceding the global financial crisis. And so when Larry Summers, prominent economist and ex-policy maker, suggested that we may have entered a period of secular stagnation and that the neutral real rate* in a number of advanced economies may have fallen to zero, or even be negative to the tune of -2% to -3%, the world took notice.
A real life example of what Summers means can be seen in Japan over the past 25 years. Growth has been low and inflation absent, despite sustained zero or near zero interest rates, episodes of quantitative easing from the Bank of Japan, as well as large fiscal deficits. Summers leaves to our imagination factors that could push the United States into a similar state; however, we have examined some prominent candidates including demographic trends, changes in the level of income inequality, globalization, and technological advances that have slowed inflation and, at the same time, displaced workers. Many choose not to believe in the possibility of secular stagnation for the United States; they point instead to an historic output gap that is only slowly receding. And so as it stands, secular stagnation in the West has the status of a prophesy: the case for lower potential growth over coming years is by no means proven.
If the market believed that we were in a world of secular stagnation, how would we know? We could start by taking Japanese capital markets between 1989 and 2014 as a script. The Japanese script is one that features government bond markets rallying as the possibility of central bank rate rises and associated economic recovery was incrementally priced out. Corporate credit markets saw spreads tighten as the reach for yield spread from the JGB market. Plus, with no monetary policy uncertainty and the Japanese economy functioning like a company in solvent run-off, defaults stayed low. Equity markets crashed, crashed and crashed again, periodically rallying 40% before giving all this and more back as the heady profitability growth experienced in the run up to 1989 was priced out of the market.
To what extent is the U.S. running to this script? In summary, we believe that the U.S. Treasury market is priced consistent with a 50%-60% probability of secular stagnation, this outcome is largely absent from corporate credit markets, and there are some signs that equity markets, despite reaching new highs, are actually discounting a reasonable probability of secular stagnation.
The big financial market development of 2014 was the collapse of forward real yields in a manner consistent with an expectation of lower terminal central bank policy rates. Whether this was due instead to lower term premia in U.S. Treasuries that might be associated with international events is hard to disentangle. But the move is certainly consistent with a market that changed its view on the long-term equilibrium policy rate consistent with steady state growth.
Exhibits 1 and 2 benchmark recent U.S. inflation and government bond market developments to those experienced 18 years earlier in Japan. Exhibit 1 shows that in the period following the popping of the Japanese land price and stock market bubble in 1989, Japanese inflation fell to around zero percent, with only periodic jumps (associated often with tax effects). Other than a brief dip into deflation associated with the collapse in energy prices that coincided with the global financial crisis, U.S. inflation has remained broadly around the 2% target rate. Despite this though, forward interest rates have fallen faster in the U.S. than they did in Japan in the wake of their asset price bubble bursting in 1989 (Exhibit 2).
Exhibit 1: Japanese CPI versus U.S. CPI (18yr lag)
Exhibit 2: 5y5y JGB yields versus 5yr5yr U.S. Treasury yields (18yr lag)
There is still some way to go before U.S. markets fully discount a Japanese policy environment. As we have previously written, a rate cycle more in keeping with our near-term economic analysis could represent a rude shock to holders of interest rate duration. If we believe on the one hand that the Federal Reserve can be successful in delivering inflation to the economy (and the market is pricing inflation at around 2.2% in the medium term), and on the other that secular stagnation comes with a zero real (after inflation) Fed Funds rate, the U.S. Treasury market appears to be priced in a manner consistent with a 50%-60% implied probability of secular stagnation occurring.
Exhibit 3: Investment Grade Yield Spreads over U.S. Treasuries
Investment-grade credit spreads have narrowed steadily since the depths of the crisis, although this appears to be driven by higher levels of investor risk appetite, improved liquidity conditions and a lower expected path for corporate defaults. But investment-grade spreads are only back to around their average long-run levels (Exhibit 3). As such, it is not obvious that spreads are discounting a period of secular stagnation, and if this were to unfold, higher quality credit would likely attract significant further allocation driving yield spreads lower. It is true that the risk-absorbing capacity of investment banks has been regulated away; so it is up for debate as to whether investors should demand a higher level of liquidity premia than in a previous era. But notwithstanding this point, we believe that corporate credit markets are not priced for a protracted period of secular stagnation.
Moving to stocks we might expect that the Japanese script is pretty clear: when secular stagnation arrives, sell everything and don’t look back. But the valuation of Japanese stocks in 1989 was a world away from the valuation of U.S. stocks: the price/earnings ratio of the Topix index was 60X in 1989, and this compares to a trailing price/earnings ratio for the S&P 500 today of 18X. By far that majority of the collapse in Japanese stocks post-bubble was associated with an unwind of this heady valuation. Furthermore, when we examine the U.S. equity market from a top down cross-asset perspective, we find a valuation that is low enough — and not inconsistent with some prospect of secular stagnation. We observe this when looking at stock market valuations from two different perspectives.
Exhibit 4: S&P500 forward earnings yields and U.S. high yield bond market yield to worst
The first perspective can be seen in Exhibit 4, which shows the forward earnings yield of the S&P500 for different calendar year consensus earnings expectations (the inverse of the forward PE), and compares these to the overall U.S. high yield bond market yield (black line). We are certainly not comparing apples with apples. High yield corporate bonds tend to be issued by companies with either significant levels of financial risk or by undiversified smaller capitalization businesses; bondholders benefit from seniority in the capital structure, but have fixed rate cash claims rather than the chance to participate in growing equity dividends that emanate from profitability growth over time. And yet despite these crucial differences, we observe that a tentative relationship appears to exist between discount rates for these differing claims on corporate profitability.
Prior to the global financial crisis, fixed rate lower quality credit tended to yield more than forward equity earnings; since the crisis this relationship has been reversed. One explanation for this reversal could be that equities are now discounting lower long-term growth rates in profitability. To elaborate, the lower short-term earnings yields that equity investors accepted in the pre-crisis period may have been accompanied by expectations of faster levels of profitability growth in the long-run (so these lower earnings yields were just the ‘down-payment’ on long-term growth), and this relationship may, post-crisis, have been reversed. This is not inconsistent with some probability of secular stagnation being discounted by equity markets.
The second perspective can be seen in Exhibit 5, which shows the level of long-run dividend growth required for stocks to deliver comparable returns to long-dated corporate bonds. To produce this chart, we applied to stocks an idea that is used to value inflation-linked bonds. When comparing an inflation-linked bond to a fixed rate bond of comparable term, liquidity and credit, the difference in the yield is called the breakeven inflation rate. This number is equal to the level of inflation that is required to allow the inflation-linked and conventional bond to deliver the same return to the investor. Exhibit 5 shows an analogous breakeven dividend growth rate that approximates the level of profitability growth that will allow owners of equities to experience the same long-run returns as long-dated bondholders (given a few basic assumptions). Over the long term, this framework has proved relatively fruitful, explaining periods of strength and weakness in equity markets with reference to long-term positive and negative dividend growth surprises (Exhibit 6).
Exhibit 5: Expected dividend growth rate implied in equity and corporate bond markets
Exhibit 6: Dividend growth forecast errors and equity excess returns over bonds
The bottom line from Exhibit 5 though is that the growth hurdle rate has been falling for equities for some time, in keeping with the secular stagnation story, and stands today just a little over 2%. Assuming that the Fed is able to deliver inflation to the economy, with the market discounting inflation at around 2%, this aggregate breakeven dividend growth rate appears very low in real terms at around zero and might be thought of as another measure of the equity market discounting a reasonable possibility of secular stagnation.
The good news for investors not bought into the likelihood of permanently lower profitability growth and an inflation environment requiring almost perpetually low rates is that equity markets have plenty of returns in store for the patient investor.
Summary and conclusions
- Government bond yield curves discount terminal central bank policy rates that may be partially consistent with the secular stagnation thesis. Without deflation in the U.S. it appears unlikely that an investment strategy concentrated in mid-curve government bonds would perform strongly even if this thesis proves correct. If it proves a mirage, government bondholders could face sizeable draw-downs as the rate-rise cycle commences.
- Credit markets do not appear to discount secular stagnation: spreads are at long-term averages, and we would expect meaningful contraction in the event that secular stagnation came to pass. We would expect investment grade credit to produce decent excess and total returns as discount rates fell further in such an environment.
- Despite the Japanese precedent of collapsing equity markets, we find evidence that markets are priced consistent with some reasonable probability of secular stagnation unfolding. As such, a similar collapse in U.S. equity markets looks unlikely in the event that low growth and zero rates are here for a prolonged period. We would, however, anticipate a higher level of volatility in the process of transitioning to an environment of secular stagnation.
*The real funds rate may be said to be neutral (“neutral real rate”) when it is at a level that, if maintained, would keep the economy at its production potential over time.