Scott Sumner  

Tyler Cowen on low interest rates

PRINT
Noah Smith on Milton Friedman... Robin's Turing Test...

Tyler Cowen has a new Bloomberg column:

Maybe Negative Yields Are a Sign of Prosperity

Just when it seemed that negative yields could not spread any further, they did. Corporate bonds paying negative interest rates now account for about $512 billion of market value, bringing the world close to a total of nearly $10 trillion in securities with yields below zero. Most are government securities.


Tyler then suggests that safe government bonds may reflect a search for insurance:

Maybe it's time we started thinking of negative securities as the equivalent of fire or earthquake insurance for that wealth. If there is truly $300 trillion in global wealth, is it so crazy to think that investors would pay a premium to buy $10 trillion dollars' worth of insurance?
In previous posts I've argued that low bond yields reflect slower NGDP growth. In the very long run, nominal interest rates often track expected NGDP growth. On the other hand, there are important exceptions to this generalization. During the period from 1934 to 1978, T-bill yields were often below (actual) NGDP growth rates, just as today. During the 1980s, T-bill yields were often higher than NGDP growth rates.

My hunch is that Tyler is at least partly correct in claiming that rising wealth and a demand for insurance play some role in the recent low rates. But I'd also like to push back a bit on that hypothesis, at least at the margin:

1. Keep in mind that while yields on government bonds are negative in Japan and in many European countries, they are still positive in the US. And yet US Treasuries are often viewed as the ultimate source of liquidity and safety, especially in a crisis. Of course the difference may reflect NGDP growth, which has recently been higher in the US than in Japan or the Eurozone.

2. Even Italian 10-year government bond yields (1.23%) are well below the yield on 10 year T-bonds (1.53%). I can't imagine anyone would regard Italian bonds as superior to T-bonds as a source of insurance. Italy has a massive national debt, an economy with near-zero NGDP growth since 2001, and a banking system under great stress. If they were to ever leave the eurozone, it seems plausible that Italy would not be able to repay its (euro-denominated) public debts.

Tyler continues:

As to when negative yields might go away or become less significant, there is now a straightforward answer: when Asia and other places become safer places for wealth. That would require less geopolitical risk, greater certainty that economic development will continue, better traditional insurance markets and, especially for China, greater ease of portfolio diversification into foreign equities.

Those developments are hardly around the corner but are plausible in the next few decades. Better insurance markets will eventually come along, the demands for super-safe securities may slack off and yields on the safer securities will rise. Just not yet.


I think there is some truth to this insurance argument, but I'd warn people that there are countervailing forces that may push in the opposite direction. Japan was the first to reach negative rates, and has three distinctive characteristics:

1. A high saving East Asian culture.
2. A mature, fully developed economy.
3. A population that leveled off, and is now falling.

Only one of the three applies to China, but in the future all three may apply. If 1.4 billion Chinese become rich, and keep saving a large share of their incomes. I'm not certain there are enough productive investment opportunities to absorb all that saving, at least at the sort of rates of return that we are used to. In my view, the growing Asia-fication of the global economy may put further downward pressure on interest rates over time.

Keep in mind that this discussion should be viewed as applying not to the overall level of nominal rates, but rather the level of interest rates relative to NGDP growth. The rate of NGDP growth is completely determined by central banks. Full stop. There's no point in looking elsewhere for explanations. That's the main reason for negative rates in Europe and Japan. But saving propensities and investment opportunities also matter at the margin, and in the future we may see a new normal of low interest rates, even relative to the already fairly low NGDP growth rates.

In the end, I'm not at all convinced that insurance is the primary story here. I expect the return on other assets to also slow sharply, at least after the level of asset prices has fully adjusted to the new world of low rates. I'm not sure that adjustment has occurred yet. But when it does occur, don't expect a continuation of the rate of return on stocks that we have been used to over the past century.

PS. The title of Tyler's column reminded me of a David Hume essay I reread last night:

Commerce encreases industry, by conveying it readily from one member of the state to another, and allowing none of it to perish or become useless. It encreases frugality, by giving occupation to men, and employing them in the arts of gain, which soon engage their affection, and remove all relish for pleasure and expence. It is an infallible consequence of all industrious professions, to beget frugality, and make the love of gain prevail over the love of pleasure. Among lawyers and physicians who have any practice, there are many more who live within their income, than who exceed it, or even live up to it. But lawyers and physicians beget no industry; and it is even at the expence of others they acquire their riches; so that they are sure to diminish the possessions of some of their fellow-citizens, as fast as they encrease their own. Merchants, on the contrary, beget industry, by serving as canals to convey it through every corner of the state: And at the same time, by their frugality, they acquire great power over that industry, and collect a large property in the labour and commodities, which they are the chief instruments in producing. There is no other profession, therefore, except merchandize, which can make the monied interest considerable, or, in other words, can encrease industry, and, by also encreasing frugality, give a great command of that industry to particular members of the society. Without commerce, the state must consist chiefly of landed gentry, whose prodigality and expence make a continual demand for borrowing; and of peasants, who have no sums to supply that demand. The money never gathers into large stocks or sums, which can be lent at interest. It is dispersed into numberless hands, who either squander it in idle show and magnificence, or employ it in the purchase of the common necessaries of life. Commerce alone assembles it into considerable sums; and this effect it has merely from the industry which it begets, and the frugality which it inspires, independent of that particular quantity of precious metal which may circulate in the state.

Thus an encrease of commerce, by a necessary consequence, raises a great number of lenders, and by that means produces lowness of interest.


A few years ago I would have dismissed this claim. Faster RGDP growth tends to lead to higher interest rates. But if one thinks in terms of the total stock of wealth, augmented by this GDP growth, then there may be some truth to Hume's claim.


Comments and Sharing


CATEGORIES: Finance , Macroeconomics




COMMENTS (6 to date)
Brian Donohue writes:

Scott, great post. Surely there are lots of factors conspiring here.

Real interest rates have been falling for a long time, tho it's hard to see through the confounder of uneven monetary debasement. I think Hume is right on, though he is only describing part of the story.

Demographics (young Boomers push rates up, old Boomers push rates down) are also part of the story, but this should 'wash out' in 20-30 years (nice euphemism, eh?)

China, which will get old fast, is trying to accomplish in a generation or two what the West has spent centuries at. I reckon this will show up as lower rates for the foreseeable future too.

There is definitely an insurance/guarantee element to this right now too though. Psyches are still fragile given the painful stock market we've endured this century. Whenever things go "risk off", everyone piles into Treasuries.

Real returns on stocks will be lower, sure, but...

The REAL annual return on the S&P 500 during the second half of the 20th century (1950-1999) was 8.6%. OK, even I'll concede that maybe the 90's ended on a bit of a 'bubbly' note, but even for the 40 years 1950-1989, the REAL return was 7.3% per year.

(All REAL rates calculated using CPI-U).

Anybody who expected a continuation of these trends was nuts. Buffett, who rarely comments on the macroeconomy, said at the turn of the century that stock investors shouldn't expect more than 6% or so NOMINAL based on then current stock prices.

In the first 15 years of the 21st century (2000-2015), the S&P 500 has earned a REAL annual return of 1.9% (4.2% nominal, lower than Buffett's prognostication.)

And people think we're still in a stock market bubble.

Given the paltry yields on safe assets, the S&P 500 (a "non-insurance" product) still looks pretty cheap right now. There is plenty of room for future returns to be a lot lower than the 20th century and still a lot higher than alternative investments today.

Scott Sumner writes:

Brian, Yes, those are good points.

Matt writes:

I am somewhat in agreement about the insurance explanation but skeptical that 'better insurance markets' will solve the problem (assuming low rates are a "problem" in need of a "solution"). The reason is that insurance companies are mostly just intermediaries that manage diverse, low-risk portfolios more efficiently than people on their own. There will absolutely be an efficiency improvement with better insurance markets in Asia, but it won't change the basic overall picture of real capital being saved by people that need it invested somewhere.

I can't find a link to it, but a couple months ago one of the investment banks published a piece of research arguing that the portfolio shift trends among the boomers (shifting from equity to fixed income) throughout early/mid retirement will result in continued increased demand for fixed-income assets for another 8-10 years in the US. Only around 2025 or so would there begin to be any actual drawdown in boomer savings that would reduce the amount of fixed-income purchases. It's a bank piece (might have been PIMCO), so it's not necessarily academic-level rigor and inherently self-interested, but it was nonetheless a fairly convincing piece to me. I'd have to imagine similar dynamics are at work and/or more advanced in OECD countries with even grayer populations.

B Cole writes:

Yeah, I think low interest rates are a function of supply and demand. The globe has capital gluts.

Print more money to boost demand.

Jason writes:

I think the article misses the point of "insurance". Insurance pays off well when a bad state of the world happens. Long-term bonds perform this role in a way that short-term bonds do not.

The key empirical observation in the article is not that short-term rates are negative, that has been true for awhile. However, more long-term rates are becoming negative. How does insurance explain a shrinking term premium in some countries?

In deflationary countries, stock markets tend to fall when bad news about further harmful deflation happens. In other words stock prices fall when nominal yields fall even further. As a result, long-term bonds rise and act as a hedge against the risk that people care about.

People can be willing to buy negative yielding long-term bonds when bad events result in even more negative long-term yields.

James Alexander writes:

For the expected return on stocks think of inverting the forward PE ratio line on page 25, lower chart:
https://www.factset.com/websitefiles/PDFs/earningsinsight/earningsinsight_7.29.16
Still above 5% ...

Comments for this entry have been closed
Return to top