Campbell Harvey, Inventor of Inverted Yield Curve Recession Signal, On The Bond Market, The U.S. Economy, And Federal Reserve Policy

Episode Summary

Executive Summary: Cam Harvey revisits his yield curve recession signal, explaining how he discovered it in the 1980s and why he still believes the 10-year minus 3-month Treasury spread has a strong recession track record. He argues the 2022 inversion remains meaningful, though effects may now be partly causal as businesses delay investment and hiring. While he sees a likely slowdown in 2024, he thinks structural factors could still produce a soft landing.

Main Topics: Discovery of the yield curve signal (Priority: 5/5): Harvey explains how, as a young intern forecasting U.S. GDP for a copper miner, he rejected complex econometric models in favor of a simple asset-price-based indicator and found that yield curve inversions preceded recessions. Why bonds beat stocks as a forecasting tool (Priority: 4/5): He argues bonds are cleaner than stocks for recession prediction because coupon and maturity are defined and interest rates are more stable, reducing noise and false signals. Track record and validation of the model (Priority: 5/5): Harvey defends the signal using in-sample and out-of-sample evidence, emphasizing an 8-for-8 recession hit rate with no false positives across modern U.S. recessions. 2022 inversion and the question of false signals (Priority: 5/5): The conversation focuses on whether the 2022 inversion is a false alarm. Harvey says it may have been if the Fed had stood down, but the continued tightening made a slowdown more likely. From indicator to causal mechanism (Priority: 4/5): Harvey suggests the yield curve may now influence behavior directly: firms see inversion, cut investment and hiring, and that risk management itself can slow the economy. Fed policy, inflation, and neutral rates (Priority: 5/5): He criticizes the Fed for keeping rates too low for too long, then hiking too aggressively, and argues inflation is overstated due to lagged shelter data. Soft landing vs. recession in 2024 (Priority: 5/5): Harvey discusses offsets such as job openings, housing equity, and consumer resilience, but warns weakening savings, credit stress, and bank deposit flight may still trigger a recession or at least below-average growth.

Key Arguments: The yield curve is a simple, publicly available indicator that outperformed complex forecasting models because it captures forward-looking market expectations. Bonds are better recession predictors than stocks because they have fixed coupons, fixed maturities, and more stable discounting assumptions. Harvey’s preferred signal, 10-year Treasury yield minus 3-month T-bill yield, has been correct for all eight U.S. recessions since the 1960s with no false signals. The 2022 inversion was not automatically a false positive; it was still consistent with recession risk, especially after the Fed kept hiking. The signal may be partly causal today because widespread awareness changes corporate behavior, leading firms to delay capex and layoffs. The 2008 inversion reflected rising systemic risk and a flight to safety as investors bought Treasuries amid leverage and financial fragility. The Fed’s rapid post-2022 tightening was historically severe and likely increased recession odds, while its inflation diagnosis over-weighted lagged shelter costs. A soft landing remains possible because labor demand still exceeds supply, household balance sheets are stronger than in 2007, and consumers had extra savings to spend. Consumer spending was the main reason 2023 avoided recession, but that support may fade as savings run down and delinquencies rise. Bank deposit flight from low-yield savings accounts into money market funds could tighten credit and act as a 2024 headwind.

Data Points: Yield curve track record: 8 out of 8 - Harvey says the 10-year minus 3-month yield curve correctly predicted all eight U.S. recessions since the 1960s, with no false signals. In-sample accuracy: 4 out of 4 - His dissertation sample covered four recessions, all preceded by inversions. Out-of-sample accuracy: 4 out of 4 - After publication, the model correctly anticipated the next four recessions as well. Yield curve inversion year: 2022 - The transcript notes the yield curve inverted in 2022, prompting debate over whether the indicator still works. Fed funds rate: 5.5% / 5 3/8% - The discussion cites the overnight/fed funds rate as being above the 10-year and 2-year yields, producing an inverted curve. 2-year Treasury yield: 4.3% - Current yield curve example given during the interview. 10-year Treasury yield: 3.9% - Current yield curve example given during the interview. 30-year Treasury yield: slightly above 10-year yield - Used to illustrate that the curve is downward sloping across maturities. Lead time from inversion to recession: 13 months average - Harvey says the last four recessions had an average 13-month gap between inversion and recession start. Current time since inversion: 14 months - He says the economy is already around the historical average lead time, so it is too early to call the signal false. Longest lag from inversion to recession: 22 months - Harvey cites the longest observed gap as almost two years. Job openings minus unemployed: 6 million to 2 million - He says the gap narrowed from about 6 million in March 2022 to about 2 million, still leaving more openings than job seekers. Housing inflation share in CPI: 35% - He argues shelter is the largest CPI component and is measured with a lag. Shelter inflation rate in CPI: 6.2% - Harvey says official shelter inflation overstated current conditions because it reflects past data. Alternative current CPI estimate: 1.8% year over year - He claims a more real-time housing measure would put CPI near 1.8%. Alternative shelter inflation estimate: ~2% or lower - He says true shelter inflation is closer to 2%, maybe 1% or 0%. Expected upcoming inflation print: 2.5% - He forecasts a large drop in the next inflation report due to base effects. Fed losses: about $180 billion - He says the Fed is making record losses from paying interest on reserves. Federal debt service: over $700 billion - He warns debt-service costs are rising sharply and could become the second-largest spending category. U.S. public debt: $34 trillion - He references headline debt and argues the true burden is higher once unfunded liabilities are considered. U.S. GDP: about $28 trillion - Used in discussing debt-to-GDP and fiscal burden. Savings deposit rate at large bank: 2 basis points - He cites his own bank as an example of banks paying almost nothing on savings. Savings deposit rate average across U.S. banks: about 70 basis points - Used to illustrate low deposit compensation versus money market rates. Money market fund yield: well over 5% - Shown as the alternative pulling deposits out of banks. CFO recession pessimism: 70% - He says Duke CFO survey data in late 2019 showed about 70% expected recession in 2020. 2023 Q3 real GDP growth: 4.9% - He attributes strong third-quarter growth largely to consumer spending.

Pivotal Quotes: "I think it is likely that we do see much slower growth in 2024." — Cam Harvey: His updated view after the 2022 inversion and continued Fed tightening. "So the cost of your forecast is the 25 cents at the time it cost for a Wall Street Journal." — Merton Miller (as recounted by Cam Harvey): Harvey recalls a dissertation committee comment highlighting how cheap and accessible the yield-curve signal was. "The yield curve is a model that is very good at predicting recessions, and it also predicts economic growth." — Cam Harvey: His closing summary of why the indicator still matters.

Implications: Listeners should treat the inverted yield curve as a serious warning, but not a certainty. Harvey sees 2024 growth slowing, with soft-landing odds still alive, yet rising policy, fiscal, and credit stresses could turn the warning into a recession.