Whether bonds hedge stocks depends on which shock dominates. In a demand-shock world, bad news lowers growth and inflation together — bonds rally when stocks fall, and 60/40 self-hedges. In a supply-shock world, inflation rises while growth falls — stocks and bond prices fall together (2022 the modern specimen), and the bond allocation becomes a second inflation bet instead of insurance.
The SF Fed’s 08/2026 Economic Letter says the correlation “has flipped for the first time in decades,” reading the flip (via stock–bond–oil correlations) as the risk source shifting to supply shocks. Their Figure 1 shows the regime turning around 2022–23 and staying flipped for three years, outside the 95% confidence band — a regime, not a blip. Consequence: the hedge everyone’s portfolio math assumes is priced for the old regime; hedging demand migrates to what works against supply shocks — Hard assets, the “expensive to hold” calculus of Real rates and gold.
Worth noting the source: a Fed research shop with nothing to sell — rare among this KB’s specimens. The transmission channel that keeps the flip alive is oil (Oil is the inflation transmission).
BCA’s version of the same flip (Berezin, 09/2026, https://x.com/PeterBerezinBCA/status/2094788579612413969): the rolling 5-year correlation between daily US equity returns and the change in the 10y yield — positive ~0.2–0.5 from 2001 to 2024 (stocks down, yields down, bonds hedged) — dipped below zero in 2026 for the first time since the late 1990s. His chain: no hedge value → less demand for duration → higher yields, and the lower panel shows the 10y term premium climbing from −1.5% (2020) to ~+0.8%. Two reading notes. The sign convention is the inverse of the SF Fed’s (returns vs yields, not vs bond returns), so “turned negative” means the same regime. And a 5-year window is a lagging detector: the SF Fed dated the flip to 2022–23, BCA’s series only crosses zero three years later because 2021’s data had to fall out of the window — the window length is the difference between “just happened” and “has been true for years.”
Corollary for the funding side: “the biggest and best Treasury buyback program would be a stock market crash” (MakeGoldGreat, 08/2026) — the flight-to-quality bid is the state’s free demand for its own paper, and it exists only in the demand-shock regime. In the flipped regime a crash can sell bonds too, which is why the Treasury is doing the buying itself (Treasury buybacks are not debt reduction).