Mr. Market, that manic depressive partner of all investors, yielded to despair last month – so much for a quiet August! Since my colleagues and I regularlyprovide topical, high-frequency commentary on market, economic and policy events on the PIMCO Blog, I’ll refrain frompiling on in this monthly missive and, instead, continue to focus on the framework and the big picture. In fact, your reactions to last month’s Macro Perspectives, which describedmy “Three Gluts” framework, suggest that the thesis (first formulated by Ben Bernanke a decade ago) that a secular global savings glut is largely responsible for low long-term interest rates remains controversial and may require furtherelaboration.
To recap, the term “savings glut” is short code for a situation in which the world’s desired saving exceeds desired investment. This depresses the naturalreal rate of interest, which is the mechanism that brings savers’ and investors’ diverging desires into equilibrium. While there is a host of potentialreasons for the ex-ante excess supply of saving over investment, here is my top-five list again:
- History (the long shadow of the financial crisis)
- Demography (we live longer but don’t necessarily work longer)
- Inequality (the rich save more than the poor)
- Technology (capital becomes cheaper and less of it is needed)
- Necessity (emerging market (EM) economies have to tighten the belt as capital pours back into developed markets)
Quantifying the impact of the global savings glut
Several of you have asked whether it is possible to quantify the impact of these and other potential factors. The unfortunate truth is that it is virtuallyimpossible to quantify the drivers of the global real long-term equilibrium interest rate or even come up with approximate estimates. An important reasonis that the equilibrium, or natural, real interest rate cannot be observed directly. The same holds for desired, ex-ante saving and investment,which, according to our thesis, show an excess of saving over investment (S>I). All we can observe in the data are actual, ex-post saving andinvestment, which, according to simple accounting principles, must be equal (S=I), unless we discover that capital flows to and from Mars.
Despite these almost insurmountable difficulties, two researchers from the Bank of England have recently made an interesting attempt to identify whichsecular factors have contributed how much to the approximately 450-basis-point (bp) decline in the global real long-term interest rate since the 1980s (seethe two-part post byLukasz Rachel and Thomas Smith on the Bank Underground blog). I’ll leave it to you to check out the methodology and details and will focus here on the main results: In a nutshell, the study finds thatsecular trends in desired saving and desired investment can account for 300 bps, or two-thirds, of the decline in real long-term interest rates overthe past three decades.Another 100 bps are attributed to slower expected trend growth, and only the remaining 50 bps remain unexplained.
To add some more detail, the study suggests that of the 300-bp contribution of the global savings glut to lower real rates, a bit more than half (160 bps)is due to higher desired saving: demography accounts for 90 bps, inequality for 45 bps and EM current account surplusesfor 25 bps. A little less than half (140 bps) is due to a decline in investment demand, with a lower relative price for capital goodscontributing 50 bps, lower public investment contributing 20 bps, and a wider spread between the risk-free rate and the rate of return on capital contributing 70 bps. As the authors note, these estimates are highlyuncertain. Yet they manage to explain the bulk of the decline in real long-term interest rates using evidence independent of the decline itself. And, takenat face value, the results support the thesis that a global savings glut has been the dominating factor behind the decline in the real equilibriumlong-term interest rate.
Why the savings glut is more likely to swell than ebb
Looking ahead, my hypothesis is that the global savings glut is more likely to increase than decrease in the coming years. This is chiefly for two reasons: negative “time preference” in the advanced economies, and macro adjustment in the emerging economies.
First, the demographics in the advanced economies are likely to remain a driver of higher desired saving for some time as people will want to build wealthto prepare for a longer prospective retirement period. We all hope and (at least on average) can expect to live longer, yet the increase in the averageretirement age is not keeping pace with the rise in life expectancy. There is even the intriguing possibility that, for many people in affluent societies, the rate of time preference has become negative. Inplain English, we may value future consumption during our retirement higher than today’s consumption.
To many economists, the notion of negative time preference sounds like heresy. In fact, it has been one of the subjects of a heated email debate on thecauses of ultra-low interest rates that has been raging for some time between more than 100 (mostly academic) German economists, in which I haveoccasionally participated on my weekends. The debate was initiated by the MIT-trained German economist Carl Christian von Weizsäcker, who proposed the ideaof a negative equilibrium rate of interest long before Larry Summers popularized it with his “secular stagnation” thesis. Most economists resist the notionof negative time preference because, ever since 19th century Austrian economists (like Eugen Böhm von Bawerk and Carl Menger) and early 20th centuryAmerican economists (like Irving Fisher) thought and wrote about the subject, the standard assumption has been that people are impatient and thus prefertoday’s consumption over tomorrow’s consumption. In other words, they display positive time preference. In the neoclassical Austrian capital theory, thisis one of the justifications for why interest rates must be positive – people demand compensation (interest) for deferring part of their currentconsumption into the future through saving.
It is easy to see why time preference should be positive in relatively poor societies where many people don’t earn much more than their subsistence leveland where life expectancy is relatively low – pretty much the 19th century world of the Austrian economists who came up with this notion. In that world,people can in fact be expected to be “impatient” and to value today much more than tomorrow. However, in our affluent societies where life expectancy keepsrising, many people seem to be very “patient” and prefer to save for all the nice things they plan to do once retired (think long cruises) – and also forthe potentially very high costs of medical and other care in the last years of their lives. One intriguing consequence of negative time preference is that it provides a theoretical explanation for negative interest rates. If timepreference was the only factor influencing interest rates (which of course it is not), negative interest rates would, in fact, be natural!
The second reason the global savings glut is more likely to swell than ebb lies in emerging markets. Recall that when Ben Bernanke first presented hissavings glut thesis 10 years ago, he put major emphasis on the impact of high and rising current account surpluses of emerging economies like China and therelated capital exports from these economies. More recently, he put more emphasis on the high euro-area current account surplus, particularly in Germany(see his April 1 blog post). Also,the former Fed chair concluded that with China transitioning to domestic-demand-led growth and the euro area recovering from its slump – with domesticdemand and, thus, import growth picking up – the global savings glut may ebb over time.
However, given the ever-more-apparent plight of many large emerging economies over the past several months, EM countries are more likely to become larger contributors to the global savings glut in the foreseeable future. One reason is that slowergrowth (China) or even recessions (Russia, Brazil) depress import growth and thus contribute to larger current account surpluses. Another is that capitalflight from many of these countries, including China, is on the rise. While it is true that official reserve acquisition by China has gone negativerecently as the yuan is no longer undervalued, this official contribution to the savings glut is likely to be replaced by private capital outflows fromChina. An important factor behind these capital outflows is likely to be the deteriorating demographics due to China’s past one-child policy, which islikely to lead to a structurally high savings rate. With domestic investment in China bound to decelerate over time, China will continue to be a large netexporter of capital. Moreover, virtually the entire emerging market complex has seen significant currency depreciation this year, and China has recentlyembarked on this route, too. This will improve EM cost competitiveness against the advanced economies and should lead to larger current account surplusesor smaller current account deficits in EM over time.
All this means the global savings glut is here to stay and, if anything, is likely to increase further in the foreseeable future. Whilecyclical forces – propelled by the other two global gluts, the oil glut and the money glut – may well push interest rates higher from their currentlydepressed levels, the savings glut is likely to limit the extent to which rates can rise. But this is unlikely to come as a surprise to you – after all,PIMCO has long condensed this view into the New Neutral concept. And regarding near-term forces, as usual, we will discuss and update our views at ourCyclical Forum this month. Stay tuned!