Thinking about the puzzling action in bond markets last week. Here are constant maturity yields from Treasury; they're based on market rates ~3:30 PM. The Fitch downgrade came after that on Aug. 1, although news may have leaked out before 1/
What's notable is that the spike took place mainly at the very long end; it has already been fully reversed for 10-years, although not for 30-years. This is really very odd 2/
Suppose you believe, as Fitch suggested, that US political dysfunction is the problem — and God knows we have that, although why you should be more worried now than a year ago, or why you need Fitch to tell you, is unclear 3/
If you think about plausible scenarios for bad events, they should have *less* effect on very long-term than medium-term rates, for two reasons 4/
First, a debt ceiling crisis that leads to a technical default shouldn't have much impact on very-long-term investors, as long as the debt is eventually repaid 5/
Second, even if you think a future government might actually repudiate debt, on the grounds that bondholders are people like George Soros or something, this should reduce all debt prices more or less equally — but a given rate increase is a bigger price fall at the long end 6/
So, a hypothesis: very long-term debt is a small fraction of total US debt, and probably mostly held by a relatively specialized set of investors. And long-term rates may therefore be subject to Shleifer-Vishny type limits to arbitrage 7/ scholar.harvard.edu/shleifer/publi…
That is, if very long-term bonds become cheap, the investors who should be buying more have just suffered significant capital losses, and may not be able to step in 8/
What I'm suggesting is that last week's bond action may have been a more diffuse version of the insurers/gilts death spiral that hit Britain during the Liz Truss moment 9/
I don't know if this is at all right, and would welcome input from people who know more about what happened. And maybe any attempt to make sense of erratic markets is a fool's errand. But I do think it's worth trying to figure this out 10/
I meant pension funds/gilts death spiral 11/
• • •
Missing some Tweet in this thread? You can try to
force a refresh
With the recent rise in consumer sentiment, time to revisit this excellent Briefing Book paper. On reflection, I'd do it a bit differently; same basic conclusion, but I think partisan asymmetry explains even more of the remaining low numbers 1/ briefingbook.info/p/asymmetric-a…
The Michigan sentiment index has two components: current conditions and expectations. It's kind of legitimate to have partisan diffs on expectations, if you think your party has better policies. It's the gap on current conditions that's startling 2/
Michigan doesn't provide partisan breakdowns every month until 2017 (hence the limited range of that chart). A quick and dirty approach is to use annual averages, with whatever months they do provide for each year, which lets us push back to 2006 3/
Recent economic news has been extremely good. But there's a strange meme among some D consultants that Biden shouldn't boast about it, because it seems out of touch — that people aren't feeling the good economy. But they are! 1/
The venerable Michigan survey has rocketed up lately 2/
It's true that consumer sentiment is still weaker than you might expect given the economic numbers. But that's largely partisanship. Using Civiqs numbers, Democrats have more or less fully accepted the good news 3/
Immigration is looming larger in the campaign, partly because it's becoming harder for Republicans to run against Biden on the economy. But there's a strong case that immigration has been a key part of Biden's economic success 1/
Inflation has come down so easily in part because of strong labor force growth. How much of that growth can be attributed to foreign-born workers? All of it 2/
Some people might look at that and say that foreigners have stolen 3 million jobs from Americans. But we have full employment, indeed a very tight labor market. Look at what the Conference Board survey says 3/
A tale of two inflation measures. Some analysts are still citing the blue line, when they should be citing the red line. This is professional malpractice 1/
Using annual core CPI puts you way behind the curve, for 2 reasons. First, annual: even core CPI was 4.6 in the first half of 2023, 3.2 in the second half. Second, known lags in official shelter prices lagging far behind market rents 2/
So annual CPI creates a spurious impression of stubborn inflation, with a difficult last mile to cover. PCE puts a lower weight on shelter, and on a shorter-term basis tells us that we've already traveled that last mile 3/
The debate over Fed policy, especially about when to start cutting rates, seems to have become disconnected from the reality of rapid disinflation. Thinking about it reminded me of ... an experience I once had on a cycling trip 1/
In 2015, I think, I went on a week-long cycling tour in Vermont. I was in pretty good shape, but hadn't done a trip like that for a while, and was slightly worried about my stamina 2/
On the second day, I was getting close to what the notes from Discovery Tours said was a major climb. As I approached, there were a series of short, sharp uphills, and I thought "if this is the approach, the real thing must be really hard" 3/
Debates about the causes of inflation and disinflation are getting strangely tangled, partly because some people don't seem to recognize that both aggregate demand and aggregate supply can shift. Here, using standard textbook pictures, is what I think happened 1/
During the pandemic and early aftermath, we had a lot of fiscal stimulus. This sustained growth and employment despite an adverse supply shock, but doing so involved a temporary surge in inflation. Then the supply shock reversed, and we got immaculate disinflation 2/
Doing it this way avoided one risk — long-term scarring from a persistently depressed economy — while running another — inflation getting entrenched. Given how things have actually turned out, it seems obvious that policymakers made the right call 3/