Ep. 360: Dirk Willer on Trading Global Macro Regimes, the End of QE, and Navigating Equity Bubbles

Summary

Dirk Willer argues that macro investors should avoid applying QE-era assumptions mechanically. During the QE period, markets often treated shocks as deflationary and expected policy easing to support risk assets. Going forward, shocks may be inflationary or deflationary, so investors need to identify the shock and the prevailing regime before choosing a trade.

Trading framework

• Express the view “close to the fire.” Use the asset most directly tied to the thesis, minimizing unrelated risks. For example, express an equity-bubble view in equities rather than the dollar, an easing-cycle view at the front end of the rates curve, and a fiscal view through swap spreads rather than simply trading the yield curve.
• Control the entry and manage risk. Willer says exits are often difficult to control, so a trade with a tighter stop can improve risk/reward. If this conflicts with being close to the thesis, he prioritizes the latter.
• Use regimes when the best instrument is unclear. Their nearest-neighbor framework compares current conditions with historical periods and assesses asset-class Sharpe ratios in similar regimes. Inputs typically include growth, changes in growth, inflation, changes in inflation, financial conditions, and sometimes volatility. In their analysis, equities tend to fare best in “Goldilocks” conditions, while commodities are favored in overheating regimes.

Business cycle, inflation, and rates

• PMIs/ISM: Willer favors PMIs partly because they have longer histories that include recessions. The 2022 ISM decline was severe but stayed above 45, which he described as a level associated with a clearer recession signal. The S&P weakened in 2022 nonetheless. The S&P has historically bottomed one to two months before the ISM and has tended to rise after the ISM bottoms.
• Other indicators: Their composite using gasoline prices, mortgage rates, and the S&P is intended to help anticipate ISM behavior. Willer considers oil important to inflation: it is difficult to have a significant inflation problem without oil participating. He views nonfarm payrolls (NFP) as noisy; at the start of a recession, NFP can still be printing around +140,000 while PMIs are already below 50.
• Yield-curve inversion: He interprets inversion primarily as a signal of possible Fed cuts, not necessarily a recession, and says it is not by itself a reason to exit equities. Inversions can precede cuts by a long time, particularly when inflation is high.
• Rates and hard landings: Rates may be less useful in short soft-landing cycles with only a few cuts and can struggle during inflationary shocks. Willer expects Treasuries to work better in a hard landing, despite episodic fiscal concerns.
• Fiscal risks and term premium: Fiscal-driven bear steepening is a risk that may recur, especially when the Fed is not dovish, but he characterizes it as sporadic. He cites the 2023 sell-off as unusually large for that point in the Fed cycle: roughly three times the largest prior bear steepening he referenced. It contributed to a roughly 10% S&P sell-off, but the move faded after the Fed signaled it was done hiking and the Treasury reassured markets about coupon issuance. Term premium matters for equities: a rise driven by term premium is more likely to cause an equity pullback. It was near the upper end of its recent range in the discussion; a breakout could cause “indigestion,” though he did not call that the end of the bull market.

Equities, valuation, and bubbles

• Willer is skeptical of valuation as a one-year market-timing signal. He says momentum has generally outperformed valuation-based approaches in equities, while valuation is more informative over a roughly 10-year horizon.
• Their bubble definition, based on a GMO methodology, is when an asset rises more than two standard deviations above its long-term trend in real terms. The signal triggered around the middle of the prior year and, in their historical analysis, usually preceded further gains. 1929 was the cited case where the market fell directly after reaching the threshold.
• The methodology does not predict when a bubble will burst. Investors should be prepared to give back a substantial share of bubble-period gains—sometimes 50–60%, all of them, or more.
• They also examined Shiller CAPE as a valuation measure and found it broadly consistent with the bubble signal. Willer cautions that forward earnings estimates can themselves be distorted during supply-constrained booms.
• Concentration in a few leading companies can be part of a bubble. Their potential warning sign is a breakdown in market leadership. The discussion references the S&P, Nasdaq, and “Magnificent Seven” companies.

Credit and risk management

• Credit spreads have historically troughed about 12 months before recessions on average, compared with the S&P peaking roughly six months beforehand. Willer attributes some of credit’s earlier warning to credit investors’ greater focus on downside risk and to credit indices’ larger representation of weaker issuers.
• He describes a risk-management test: when the VIX rises by two standard deviations, exiting the S&P for a month on that signal alone did little to improve results. Requiring high-yield credit spreads to have widened over the prior six months as well materially improved the information ratio in their study.

Dollar and foreign exchange

• Their dollar model uses four relative measures: interest-rate performance, equity performance, terms of trade, and macro momentum (measured with Citi data-change indices). When all four point in the same direction, the dollar’s Sharpe ratio is higher; the signal weakens when fewer indicators align.
• Willer questions the dollar “smile” as an empirically robust framework. He prefers relative U.S. equity performance to a simple VIX-based risk-aversion measure because it captures whether U.S. equities are outperforming or underperforming.
• If an AI-driven equity bubble were to burst and all four indicators turned against the dollar, the model could point to a large short-dollar opportunity. But he notes that after the Nasdaq peaked in March 2000, the dollar did not fall sharply until 2002, so such a trade may require patience.

Geopolitical shocks and oil

• Willer generally favors fading geopolitical shocks in equities when they do not materially damage earnings. He says markets tend to recover quickly from geopolitical sell-offs, while emphasizing that investors still need to assess the transmission to earnings—often through oil or tariffs.
• In the episode discussed, forward earnings per share had risen since March 1 in nearly every country, with Hong Kong cited as an exception. He says the S&P has typically been down only briefly after major geopolitical shocks, including the Iraq wars, Russia–Ukraine, an Iran-related episode, and the Arab Spring.
• His regime analysis suggested oil would need to rise above roughly $150 to push conditions into stagflation. If oil caused earnings estimates to fall, the equity-market response could change.
• Rates may be more sensitive than equities to oil-linked geopolitical shocks because bond markets are more tied to expected central-bank behavior. For bonds, Willer says a clear oil-price peak is important before markets can relax.
• He cites volatility risk premium as one tool for buying S&P dips, while noting it did not trigger during the particular shallow sell-off discussed.

Systematic models and human judgment

Willer sees backtests and models as useful but not sufficient. Investors must judge when a model has broken or should be switched off, and recognize events that do not fit historical rules. He recommends limiting overfitting while incorporating narrative and judgment, especially for unusual developments such as major policy shifts or exceptionally large investment cycles.

Assets, instruments, and sectors mentioned

• Equities: S&P, Nasdaq, U.S. and global equity markets, cyclical and defensive stocks, and concentrated mega-cap/“Magnificent Seven” leadership.
• Rates and fixed income: U.S. Treasuries, front-end and back-end yields, yield-curve inversion and steepening, term premium, swap spreads, duration, and QE.
• Credit: Credit spreads and high-yield credit.
• FX: U.S. dollar and relative currency performance.
• Commodities: Oil and gasoline; gold is mentioned as an asset class in the regime discussion.
• Macro indicators: PMIs/ISM, NFP, mortgage rates, VIX, and Citi macro-momentum indices.
• Sectors/themes: AI, Korean chipmakers, and technology-related capex.

Disclosures and sources

No explicit “not financial advice” disclaimer appears in the subtitles. Bilal Hafeez recommends Willer’s book, Trading Global Macro Markets, coauthored with Alex Saunders. Willer describes it as a dense, professionally oriented book rather than light reading.

Presenters and sources: Bilal Hafeez (MacroHive host/interviewer); Dr. Dirk Willer (Citi Research guest). The discussion draws on Willer and Alex Saunders’s book and on Citi Research analysis.