CEPR - Center for Economic and Policy Research


En Español

Em Português

Other Languages

Home Publications Blogs Beat the Press Phantom Bubbles at FiveThirtyEight

Phantom Bubbles at FiveThirtyEight

Wednesday, 23 April 2014 04:38

FiveThirtyEight looks at the bubble horizon and concludes stocks and housing are safe, but we should be worried about bonds. The analysis here is seriously misguided.

First as a sidebar, contrary to what you read at FiveThirtyEight, real house prices are somewhat above, not below, their long-term trend levels. That doesn't mean we have a housing bubble, but anyone anticipating a future rise in nationwide house prices in excess of inflation is likely to be disappointed.

But the more important point is that the concern about a bubble in bonds is largely illusory. The piece constructs a case for a bond bubble that just is not there.

First, I was surprised to read that the size of the U.S. bond market is almost $40 trillion, which the piece rightly points out is considerably larger than the $28 trillion stock market or the $20 trillion housing market. When I checked the source for this number I discovered that the figure referred to the total size of the debt market, not just longer term debt that we would typically refer to as "bonds." The FiveThirtyEight figure includes 90-day T-notes and money market funds.

This is not just a question of semantics. Longer term debt (with a duration of five years or more) has large fluctuations in value in response to a change in interest rates. The price of shorter debt will also vary, but the size of the changes is trivial by comparison. This means that if we are worried about a bubble inflating bond prices, we should really only be looking at longer term debt. The size of this market would be roughly half as large, or less than $20 trillion. That's still big, but a considerably smaller basis for concern than the piece implies.

More importantly, the room for losses in this market is not nearly as large as it was in the case of the stock or housing bubbles. The stock market lost more than half of its value from its 2000 peak to its 2002 trough. House prices lost more than one third of their real value from the 2006 peak to the 2011 trough. By contrast, it is difficult to envision a scenario where the bond market loses even 10 percent of its value.


This can be seen with a simple bond calculator. The interest rate on 30-year mortgages is currently around 4.15 percent. Suppose it were to rise to 5.5 percent, a very large increase. This would imply a drop in the price of a newly issued 30-year mortgage of roughly 19 percent. That is considerably less than the drop in house prices or stock prices seen in the collapse of these bubbles.

Furthermore newly issued 30-year mortgages are a small fraction of bond market debt. Most of the bond debt has maturities of ten years or less. For these bonds the drop in price associated with a comparable rise in interest rates would be less than 10 percent. This could make for some unhappy investors, but would hardly lead to a financial or economic collapse.

In this respect it is worth noting bond prices already did take a very large hit following Bernanke's famous taper talk last summer. The interest rate on 30-year mortgages rose from less than 3.5 percent in the spring of 2013 to more than 4.5 percent in the summer. If there was any serious stress created by this fall in bond prices, the financial media neglected to mention it. It is unlikely that any future rise in interest rates will lead to as large a drop in prices.

The other evidence mustered in this piece for the bubble case is that spreads have fallen back to pre-crisis levels. Which invites the obvious "so?" And, we see that higher risk debt is increasingly being issued.

This is not the sort of stuff that should cause anyone to lose sleep. The collapse of the stock and housing market bubbles led to recessions because these bubbles were driving the economy. When they burst there was nothing else to replace the demand these bubbles had been generating.

Lower interest rates certainly help the economy, but does anyone believe that investment would freeze up or that housing construction would plummet if the interest rate on long-term debt rose by a percentage point from current levels? In short, the only reason to be concerned about a bubble in the bond market is that influential people are apparently taking the risk seriously and could pressure the Fed to needlessly raise interest rates and cause more unemployment.


Comments (14)Add Comment
Bond Bubble
written by Ed Brown, April 23, 2014 6:37
Dear Dean, I have read that bubbles are normally based on emotions, where people buy a thing (stocks, tulips, houses) in the belief that the price will be greater in the future. Most people start to think the thing (stocks, tulips, ...) can only go up, so whatever they pay today is irrelevant; it will be worth more tomorrow. Hence a bubble.

I do not believe that most people believe that interest rates will continuously drop from where they are today, as they are quite low already. This suggests that most people do not think the price will continuously rise, which suggests we do not have the necessary social precondition for a bubble in the bond market.

Your comments, and the thoughts of other commentators, on this suggestion would be appreciated.

Finally, I wish to thank you for your blog. I read it every day for several years now and I have learned a lot.
Bond Bubble?
written by Jerry Brown, April 23, 2014 8:18
I think Ed makes a good point. How can you have a bubble if people think that the price of the asset in question is not going to rise?
I am simply shocked.
written by Bill H, April 23, 2014 8:40
Five thirty eight is assumeed to be the ultimate authority on all things having to do with numbers and statistics. How can you dispute any of Nate Silver's conclusions? Shocking.
written by skeptonomist, April 23, 2014 8:58
It is very remarkable how oblivious the various "authorities" are to historical evidence. When interest rates were low all through the Depression and recovery (GDP growth averaged 10% from 1933 to 1937, and picked up again after the recession of 1937) and WW II, bond prices were extremely high. Debt/GDP reached over 120% at the end of the war. Did that "bubble" ever collapse? Or was the peak of interest rates around 1982 the collapse of the "bubble"?

Yes, as Ed Brown says there are bubbles because people expect prices to rise. Does anybody really expect the price of long-term bonds to keep rising (much) from where they are now? I saw a piece yesterday claiming that 100% of economists are expecting interest rates to rise within six months, that is prices of bonds to go down. As Dean says for somewhat different reasons, the idea of an asset bubble in bonds does not really make much sense.

Bond prices may go down when there are real fears of default, so if the danger of default is not recognized there may be a kind of bubble which collapses when the danger is discovered. Central banks can counter the collapse of such bubbles, and both the Fed and the ECB have done so since 2008.
written by JDM, April 23, 2014 9:25
Reading Dean's post, especially 538's goof about what constitutes "the bond market", made me think this is another manifestation of the idea that brought about the rise of the MBA, which I'd argue was disastrous. It's the idea that you can gave skills which apply to virtually anything without the need for context. So you don't need to be able to understand people to do person ell, you just look at some numbers. You don't need to have any deep knowledge about what a business does, because it's all just "product", and in 538's case, you don't need any of that pesky expertise, because your mad math skills trump that.

This makes it easy to bounce from one business to another, or one discipline to another in the case of 538, No experience necessary.
written by JDM, April 23, 2014 9:29
BTW, my apologies for not double checking my ever-helpful spell checkers "corrections", Specifically, "gave" for "have" (okay, typo there), and "person ell" for personnel.
I Just Don't Understand
written by Larry Signor, April 23, 2014 10:06
Unless bonds are being undermined as collateral for riskier ventures, where is the big scare? If I buy a 10 yr., $100.00 bond, paying %4 per year, at maturity I have made @ less inflation (and I get my $100.00 back). Someones calculator is broke or their fear factor is off the charts.
written by bananaguard, April 23, 2014 10:08
I was surprised to see that Flowers once worked for the Atlanta Fed. His casual use of the term "bubble" does not suggest the sort of disciplined thought I'd expect. I wasn't able to find a bio with small effort, so I don't know what he did at the Atlanta Fed.

I understand that he's blaming Fed Governor Stein for his views. Problem is, there are other Fed officials who disagree. President Evans this week, for instance, who said the bond market has gotten over last year's fever and is now priced pretty close to right. Picking one Fed official over another to quote is a choice, and the writer needs to be held accountable.

So, Mr. Flowers, what definition of "bubble" are you using? Aren't bubbles behavioral, rather than just configurations of price? Doesn't a low funding cost constitute a fundamental reason for low bond yields? Low stock and housing yields, too? If there is a fundamental cause, do we have a bubble in anything but the suspicion that some people will be sorry for their purchases, some day?
written by JSeydl, April 23, 2014 11:19
I'm certainly not saying the Fed should raise rates, but something structural maybe going on. The reality is that bond issuance has picked up dramatically. That's not driving the real economy, but it does beg the question: what are all these junk companies doing with the borrowed funds? Are they investing in new, capital-intensive technology? Well maybe, but equipment capital formation does not look unusually strong. Are the junk companies using the money to gamble in financial markets? Perhaps. Or maybe the money is going to investment projects in emerging markets? That seems plausible as EM investment continues to be strong, while advanced economy investment is weak:

Any other hypotheses?
written by JDM, April 23, 2014 12:12
Any other hypotheses?

Paying outrageous executive salaries and comps?

p.s. I'd missed that Flowers once worked for the Fed. That makes his apparent lack of expertise (given his uses of terminology and his assumptions) rather more frightening than merely foolish.
corp profits versus s and p
written by jim, April 23, 2014 10:13
although after tax corp profits measure more companies than those listed in the s and p, by using a simple fred graph you can see from the troughs corporate profits have risen 282 percent versus the stock market at 270 percent. looking at the relationship between the two, it is quite evident a bubble was forming in late 1990s as s and p got completely away from corp profits. however at this point they are rising in lock step. i believe the great moderation of the 1990s has lead to a different world in investing, further aided by the imf sticking it to em countries, and keeping our interest rates suppressed. even more important is the average real earnings yield of the s and p, which has averaged 2.5% during the past 50 years. currently it is 4%, which is why i believe that equities will become "overvalued" versus historical norms. The environment for equities has never been better in terms of interest rates, inflation, margins, corporate profits as % of gdp, and most importantly all other investment opportunities. why wouldn't the market pay more today for a dollar 10 years out at 1.5% real rates? but i do agree with Dean that the bond market cannot be called a bubble, because as Gavyn Davies pointed out in his FT op-ed, nominal rates should be about 3.25-3.5% for the foreseeable future. i still have trouble putting Emerging markets, europe, etc into context because the valuations depend so much on how the dollar does versus these currencies, what share of growth wages account for in corporate profits, and how good a job the government does creating social safety nets.
Fed Governor Jeremy Stein
written by Riley, April 24, 2014 10:40
It surprises me that you could comment and critique the FiveThirtyEight article without even mentioning Jeremy Stein. The entire argument is basically re-articulating the argument that Fed Governor Jeremy Stein has been making over the last couple months. His argument is about incorporating measures of financial market stability into a monetary policy framework. Therefore, he is dealing with the interest rates for the whole market, not just long-term debt. Furthermore, I think his argument has interesting implications for equities as well! i.e. If the risk premium shoots up in fixed-income, what happens to the required rate of return for equities?

Anyways, just wanted to clarify that this whole discussion is not really about a "bond market bubble" but about financial market stability and the implications of that for monetary policy (specifically asset purchases).

written by moneyBob, April 24, 2014 11:47
A few thoughts.

1. You may want to check the dv01 on a 30-year mortgage. They amortize, and your loss on a 150 bps increase in rates would not be 19%, unless you a paying way too much for specific collateral with unique prepayment characteristics. So before you dump on Nate, check your own facts...(even though this only bolsters your argument by saying the thresholds to a 20% loss in bond markets are extraordinary).

2. All of you guys miss the real point - leverage. A "bubble" in asset valuations is only important relative to the leverage employed to own that asset. What is the haircut on stocks, real estate, and bonds? Stocks are largely held directly. Homes, 20% downpayments, with the crisis predicated on higher levels of leverage. Bonds, well...haircuts on a 30-year Treasury bond are ~5%. Bonds are scary because their are more turns of leverage than equities. Assuming two turns on equity and twenty on bonds, what is the equivalent equity absorbing loss of the respective market? Seems like your equity 20% loss gets a little smaller, no?

The Size of Historical Rate Rises
written by Matthew Wittman, April 24, 2014 1:02
Freddie Mac has 30-year mortgage rates going back to 1971. The data is available here:
In the 1980s rates spiked up to above 18%. That was a large movement, and one specifically caused by the Fed raising rates to reduce inflation, which, given the current run-away money creation, may be happening again in out future. The 5.5% the author suggests as being large is not. A 30-year mortgage originated at 4.15% would lose 70% of its value.

Write comment

(Only one link allowed per comment)

This content has been locked. You can no longer post any comments.


Support this blog, donate
Combined Federal Campaign #79613

About Beat the Press

Dean Baker is co-director of the Center for Economic and Policy Research in Washington, D.C. He is the author of several books, his latest being The End of Loser Liberalism: Making Markets Progressive. Read more about Dean.