Did we have a bubble in sovereign debt?

One of the most abused words in financial discourse is the word “bubble.” The problem is that the term “bubble” has been used as a stand-in for the word “overvalued.” We see people pointing to bubbles in residential real estate right now, and by all measures housing is about as unaffordable as it has ever been. But does that mean it is in a bubble? Probably not.

The best book on bubbles is Manias, Panics, and Crashes by Charles Kindleberger. Bubbles are first and foremost a psychological phenomenon. People generally believe that an asset is “special” and cannot fall in price in a sustained fashion. We saw this mentality in the late 1990s with stocks. The conventional wisdom on channels like CNBC was that all an investor needed to do was buy “good companies” and hold for the long term. The concept of price and value never really entered the equation, and why should it? If stocks go up over the long run you should make money regardless. Which is why people had no problem buying Cisco Systems at 70 times earnings in the late 1990s.

There were bestsellers like Dow 100,000 and people were quitting their jobs to become day traders. Everyone (banks, regulators, academia, the media) was caught up in the frenzy. CNBC supplanted ESPN in bars and barber shops.

Of course once the stock market bubble burst, a lot of companies did go bankrupt, but the stalwarts did survive. However, if you followed the conventional wisdom and bought Cisco Systems at the height of the bubble, you are still underwater by about 30% and the P/E ratio has collapsed to 17.5x earnings. And if you bought Cisco on margin, you probably lost most of your investment. So you can indeed lose money buying “high quality” companies, and that is the lesson of 1999. Which is why you aren’t about to see a bubble in stocks right now - the memories of 2000 are too fresh. Bubbles come along only rarely, and the last time we saw one prior to 2000 was the 1920s when Radio Corporation of America aka RCA was that generation’s Cisco Systems.

While there isn’t a lot of discussion going on about it, I think we have seen another bubble burst, and that is sovereign debt. We just exited an extraordinary period where interest rates were negative when you adjust for inflation. Indeed, some countries had negative nominal interest rates - you would pay above par for a German Bund to only get back par when it matures. A lender paying interest to a borrower is about the most upside-down phenomenon imaginable. It is the equivalent of putting cash in a safety deposit box and paying a monthly fee for the box.

This state of affairs wouldn’t have been possible without central bank involvement, specifically yield curve targeting and quantitative easing. That said, now that inflation is back, the bubble in sovereign debt has burst. The chart below shows the relative performance of different asset classes after a burst bubble. I compare the relative returns of stocks, real estate and sovereign debt. It certainly appears that we had a post-bubble-like sell-off in Treasuries.

The green line is the FHFA House Price Index after the peak of the residential real estate market in early 2007. The blue line is the total return index for the US 30 year bond after the peak in April of 2020. The red line is the Nasdaq 100 price index after the bubble burst in March of 2000.

If you had invested in a 30 year Treasury in April of 2020, you would have about half your money left after interest. It is hard to imagine that the “flight to safety” trade would get annihilated like that, but here we are. People lost 50% in Treasuries over the past 3 years. This is supposedly the worst bond market rout in 150 years, according to Mohammed El-Arian.

Part of bubble psychology is that regulators have to believe the asset in question is special as well. Certainly during the housing bubble, regulators were pushing Fannie Mae and Freddie Mac on affordable mortgages, and allowed Fannie to become one of the biggest mortgage arbitrage hedge funds on the Street with a highly leveraged subprime loan portfolio. Banking regulators allowed banks to mark their portfolios at par, when they were highly illiquid and wouldn’t fetch near par. Remember the “mark-to-model” days?

We see something similar now with sovereign debt, where deeply underwater bond and MBS positions are being marked at par and considered to be Tier 1 capital. It is striking that in all of the Fed’s stress tests, they never contemplated a scenario where interest rates would rise rapidly. In their minds, financial stress only occurs in recessions and rates fall during recessions. People were lulled into believing that Treasuries were risk-free assets, and they are “risk free” in that investors will get their principal and interest on time, but it doesn’t mean that there won’t be mark-to-market losses.

The Silicon Valley Bank collapse proved that mark-to-market losses on “risk free” assets are not a simple theoretical possibility, but they have actual effects in the banking system. Bank earnings season has just kicked off, and Treasuries were massacred in September. The big banks like JP Morgan were the beneficiary of these bank runs, however the smaller banks will be more vulnerable. This could be the catalyst to push the Fed to scale back quantitative tightening.

It does seem odd to not hear the term bubble when discussing the bond market in the business press, since the term does seem to fit. It is certainly more relevant to the bond market than it is to the current real estate market or the stock market.