The Term Spread As Recession Predictor, Post

Using a plain vanilla term spread model (spread, short rate), what remains? From notes for tomorrow’s lecture. Figure 1: Estimated recession probabilities from probit on 10yr-3mo term spread, short rate over 1960-2018 (brown), over full sample (black). Red dashed line at 33%. NBER defined peak-to-trough recession dates shaded gray. Source: author’s calculations, NBER. Using a 33% threshold and the pre-2018 sample, the term spread catches every post-1960 recession (taking the pandemic recession was not a typical recession), but gives the false positive for 2024. The full sample regression (assumes no recession appears by 2026M08) no longer catches the 1990-91, 2001, and 2007-09 recessions, and still provides a false positive for 2024. In other respects, the full sample regression is less successful; the pseudo-R2 drops from 0.30 to 0.25. My personal experience with this metric showed a notable divergence during the 2008 financial crisis. This situation created a scenario where the predicted downturn did not align with the actual market behavior. Way off topic – U.S. withdrawal from Iraq: Anyone keeping up with the news is aware that U.S. military forces officially withdrew from Iraq this week, the final act in Shrub’s war of choice with Iraq. The timing strikes me as ironic as hell. News reports mention, often uncritically, the felon-in-chief’s edge to end U.S. “forever wars”. Anyone familiar with the recent history of the region knows that the two U.S. wars with Iraq grew out of the Iran/Iraq war of the 1980s. We supplied Saddam with weapons and egged him on to attack Iran. Saddam learned impunity, which got him killed. Iran learned how much war it could survive, and it was a lot of war. Our withdrawal from Iraq stands in comic contrast with our new own-goal war with Iran. Iran-backed militias have celebrated in the streets all across Iraq. We and Israel have been trying to wipe out Iranian allies across the Middle East and we very likely just turned Iraq over to Iran’s allies. Meanwhile, the Houthis control more of Yemen than ever before, a direct consequence of our war on Iran. I’m not saying we should have stayed in Iraq. Hell, we never should have fought Shrub’s Oedipus-complex war. We should have learned the lessons of Shrub’s war, but the very guy who campaigned to be president by condemning Shrub’s war(s) just had to have a war of his own. Iran and Iran’s allies have lost their grip on Gaza, much of Lebanon and parts of Syria, but have tightened their grip on Yemen and Iraq, and grabbed the Straits of Hormuz and Bab-el-Mandab. Looks like Iran is winning this round, and winning big. I see many good reasons to leave Iraq – but doing so in the midst of a conflict with Iran makes little sense. At a minimum it could have been offered as a part of deal. But it’s possible that Iran was getting ready to bombard our troops in Iraq with drones – and better leave on your own before it looks like another humiliating retreat under fire. This is probably bad: https://thehill.com/policy/defense/6124637-pentagon-deploys-marines-middle-east/ Three amphibious assault ships, 4,400 marines and sailors, headed for the Mid-East. Remind me, what are amphibious assault ships good for? Oh, right…putting troops onto the soil of foreign countries. When might they arrive? Probably before November 3. My guess is the Houthis – they are small enough for a quick successful invasion. However, there is a fairly high risk that they have surprises prepared that will make it take longer and cost more casualties than Kegsbreath has imagined. And now another aircraft carrier: https://www.aljazeera.com/news/2026/10/2/new-aircraft-carrier-10000-us-troops-is-the-iran-war-about-to-escalate Republican polls must really be bad in Iowa and Alaska, formerly reliable red states. You’ve heard Trump boasting about a new $15 billion steel plant for — Iowa?? — from a shady company that has been bankrupt and kicked out of Minnesota as an “unreliable partner.” So now Trump is crowing about a $45 billion investment in Alaska for an LNG pipeline by South Korea. Except South Korea today said that it is only beginning to review the possibility of the project. Seoul has said that “no decision has been made” on any investment. And the kicker: Seoul’s Industry Minister also expressed regret to Commerce Secretary Howard Lutnick, saying the White House announcement went beyond what had been agreed. A polite way of saying that they regret that the White House is lying. No country trusts the US anymore. It’s coming right after that $2 per gallon gasoline that he also promised. It may be part of that huge buildup of US manufacturing that his tariffs would induce. And South Korea already has a real, signed deal with Canada for 3.4 million tons of LNG annually from a port in Kitimat, British Columbia. South Korea said that Canada, with substantial hydrocarbon reserves and a stable geopolitical profile, meets the criteria Seoul is looking for. I hope nobody tells Trump. Under the current circumstances, I would be very careful about using the traditional term spread model. As I understand it, the US government is financing itself through money market funds via hedge funds. All that is short money, effectively uninsured demand deposits. Similarly, the government appears to be benefiting from stablecoins, which are backed by US Treasuries. The deficit, therefore, looks to be financed by the growth in stablecoins. Thus, the government is again borrowing demand deposits, this time primarily internationally, from the likes of illegal immigrants, Philippino nannies in the UAE, guestworkers in Russia, Nepali restaurant owners, Argentine landlords and South African farmers. This is both brilliant and appalling, but be that as it may, the US is again leaning heavily on the short end of the curve. Therefore, under the current circumstances, we may see unusually upward pressure on short term rates due to the disappearance of traditional institutional buyers of long-term US debt. The term spread may not be telling you what you think it is. I’m curious as to the source of your information. Treasury securities are Treasury securities, and are sold at auction. Even non-competitive auction bids are priced during the auction. Secondary market trade in Treasury securities is the source of interest rate quotes, not atablecoin prices. It’s not clear to me how stablecoins or nannies figure unto secondary market trade of Treasury securities. OK, I may have m8ssed your point underneath the stuff about illegal immigrants and nannies. If what you mean is that hedge funds are a larger than normal holder if Treasuries, that’s true. Something like 7% of outstanding Treasuries – or mayne 7% of publicly held. Anyhow, that does make the Treasury market riskier, but I’m not clear as to how it distorts the signal from the curve. Hedgies can operate anywhere along the curve they want, so unless there is a specific curve strategy in play, there’s no reason to look for a distortion to the curve from hedge fund participation in the Treasury market. As a distraction from Iran and the election and talk of AI killing us all, here’s a new look at the future of AI spending and revenue from Bain: https://www.bain.com/insights/new-innovation-is-required-to-fund-ais-6-trillion-buildout-technology-report-2026/ In the mildest language umaginable, the report says the majority of the revenue necessary to make continued AI buildout workable has to come from an as-yet unimagined well of economic value. Productivity gains can only provide a fraction of the revenue necessary to pay for expected investment, with the rest coming from new markets for AI goodies. Either we find magic new uses for AI, or so long, and thanks for all the chips. Here we go – just 29,000 new jobs in September, with the jobless rate at 4.2% vs a prior 4.1%. The 3-month average gain in jobs was 51,000, same as in August. Hourly earnings up just 0.1% on the month. Sad, but not completely bad.