For two decades, the American investor essentially got a free insurance policy. When equities fell, Treasuries rallied, and the loss on one side of the portfolio was partly covered by the gain on the other. That relationship became so reliable that an entire industry built products on it, and an entire generation of allocators started treating it as a given.
However, it stopped working around 2020, and it hasn’t really worked since.
UBS now puts the two-month rolling correlation between the S&P 500 and the 10-year Treasury yield at -0.69, the lowest reading since 1996.
That means stocks and bonds are moving together to a degree not seen in thirty years, and the asset that exists to offset an equity loss has now become a source of one.
What’s a safe haven now if bonds aren’t?
It’s easy to say that the reason why bonds and equities have converged is that investors lost faith in US government debt. However, as always, the answer is much more complicated than that. The data tells us that investors still want the safety they got from bonds, but now they want it without the duration.
Duration is the sensitivity of a bond’s price to a change in interest rates. A 30-year Treasury protects the holder from default in nominal terms and exposes them completely to inflation and to the path of policy rates. Even though those are two different risks, the distinction didn’t really matter after the financial crisis of 2008, because inflation was mostly dormant.
Once we start seeing inflation go up, the hedge breaks apart. The correlation between stocks and bonds doesn’t depend that much on the actual level of inflation, but on its volatility. It also depends on what drives the markets: news about growth or news about inflation.
When growth dominates, equities and bonds respond in opposite directions, because weaker growth hurts stocks and helps bonds. When inflation dominates, they move in the same direction because higher inflation hurts both of them equally. Research at AQR found that this explains roughly 70% of the long-term variation in the US stock-bond correlation, with similar results internationally.
Since 2022, inflation has been the dominant input, and it has remained dominant longer than we’ve ever seen. Even a cooling inflation print, like the June report that pulled headline CPI to 3.5% and left long yields drifting back toward 5% at the 30-year, didn’t change anything, because the volatility of inflation is the problem rather than any single reading of it.
The 30-year Treasury yield crossed 5% for the first time since 2007, has spent much of 2026 above that line, and sits near 5.1% as of July 16. A $25 billion auction of new 30-year bonds cleared above 5% earlier in the year, the first time investors were paid that much on the long bond in eighteen years.
US deficits are projected to widen from roughly 5.8% of GDP in 2026 toward 6.7% by 2036, with net interest payments growing as a share of the economy every year in between. OECD governments collectively need to raise something in the region of $18 trillion this year.
Foreign demand is thinning just as the supply is thickening. Japanese investors sold $29.6 billion of US government, agency and local authority debt in the first quarter, the largest net sale since 2022, as domestic yields finally became worth owning. Japan’s 10-year climbed above levels last seen in 1997, and Germany’s 10-year Bund reached 15-year highs. The global bid that suppressed long-end borrowing costs for two decades is being withdrawn in several places at once, and term premium is the price of that withdrawal.
All of this tells us that investors are buying dollars, bills, and short-dated paper, which are liquid and carry almost no duration. They’re selling the long end, which carries all of it. That’s a 180-degree rotation of the haven trade, and it explains how the dollar can stay firm in a week when the 30-year is being sold.
Where does this leave Bitcoin?
Bitcoin is now as sensitive to macro conditions as the dollar and gold are.
BTC performs when real yields fall, when the dollar weakens, when financial conditions loosen, and when investors go looking for alternatives to conventional assets. A Treasury rally delivers the first three at once, which is why a falling bond market removes three pillars of support at a time. The recovery that carried Bitcoin back above $64,000 this week came when a soft inflation report pulled front-end yields lower.
Societe Generale’s research identifies roughly 4.5% on the 10-year as the level where the relationship between yields and equities turns hostile. Below it, rising yields and rising stocks can coexist. Above it, further increases drag equities down through the discount-rate channel.
Goldman Sachs reached a similar conclusion from another angle, warning that the rise in yields has compressed the equity risk premium to the point where investors are barely compensated for owning stocks relative to risk-free assets. The 10-year has spent most of 2026 above that threshold, easing only to around 4.55% after this week’s cooler data.
Bitcoin sits further out on the same curve than equities, which means it absorbs both pressures at once. Higher risk-free yields raise the opportunity cost of holding an asset that pays no coupon. Falling equities reduce the appetite for risk that would fund a stock position.
Neither of those is a crypto-specific problem, so neither can be solved by crypto-specific news, which is why regulatory progress in Washington has repeatedly failed to hold a bid this year.
But despite their correlation, this isn’t a fight between Bitcoin and Treasuries. In an inflationary risk-off regime, they compete for nothing. They’re on the same side of a single position that sells duration and volatility, raising cash. Gold, long bonds, and Bitcoin can all fall in the same week while the dollar stays strong, telling us just how much interest-rate and volatility exposure anyone currently wants to own.
The fiscal conditions producing 5% long yields, deficits, interest burden, and the fading foreign bid are the same conditions that make a fixed-supply asset outside the sovereign credit system attractive to institutional holders.
Some of that capital is already visible in the $15 billion of tokenized US Treasuries now held on-chain, which is a crypto-native bet on yield rather than on scarcity. The problem for Bitcoin is that the conditions strengthening its long-term case hurt it in the short run.
Treasuries can reclaim the role they held from 2000 to 2019. It would require inflation volatility to subside, growth risk to become the dominant input again, and the Fed to have room to ease into weakness.
We saw that combination of factors after every previous inflation shock, and so far nothing rules out that it’ll come after this one. A single soft inflation month is not that combination, though it’s the kind of data point that would eventually build toward it.
Until it does, Bitcoin trades in a market where the deepest asset class in the world no longer absorbs a shock on anyone’s behalf. That removes a floor beneath every risk asset, and it removes it fastest beneath the assets that pay nothing to wait.
