"Bonds are attractive for the first time in decades." Bank of America's Savita Subramanian said that when the 10-year Treasury yield pushed through 5%, and the crypto market barely blinked. It was busy. There was a token launch, an airdrop, a JPEG, a points program, a leaderboard. There is always something.
That is the tell. The single most important input to every crypto valuation model on earth just moved, and the market filed it under background noise. Meanwhile the same market is busy pricing a supercycle.
Here is the mechanical fact underneath the headline. The 10-year yield crossed 5%. Subramanian's argument is that this sits near a ceiling — that yields struggle to hold above 6% to 7% because policymakers do not want them there and demographics keep the neutral rate suppressed. She is making a mean-reversion call on the risk-free rate, and telling allocators that the bond leg of their portfolio finally competes with the stock leg.
Strip the bank branding away and one number remains: 5% risk-free. That number is the gravity well of global finance. Crypto lives inside it whether it admits it or not.
The risk-free rate is not a yield. It is a discount rate, and every asset on earth is priced as a spread off it. When the 10-year sat near 0.5%, a DeFi pool paying 8% looked like a gift. When the 10-year pays 5%, that same pool is paying a 3% spread for taking smart-contract risk, liquidity risk, oracle risk, and governance risk. The premium inverted. The only question that matters is never "what does this asset yield?" It is "what does this asset yield net of the thing that cannot default?"
I spent the 2020 DeFi Summer making that argument to a timeline that did not want to hear it. I pulled the Compound and Aave yield curves and laid them next to the Fed's balance sheet, and the "double-digit APY" resolved into what it always was: fiat debasement arbitrage wearing a lab coat. The yield was not generated. It was redistributed — from token inflation to depositors, from late buyers to early ones. When the Fed stopped expanding the liquidity base in 2022, the yields collapsed. Not because the code broke. Because the subsidy ran out. Liquidity mining APY is the project paying you to be a number in a dashboard. Cut the subsidy and the number walks.
Now place Subramanian's call inside that map. The macro backdrop is heavy Treasury issuance, a Fed running down its balance sheet, widening term premiums, and deficits that show no sign of closing. That combination is what pushes the 10-year toward 5% and beyond. None of it is a crypto story. All of it is a dollar-liquidity story. Crypto simply happens to be the highest-beta expression of dollar liquidity on the planet.
The map in one line: if the risk-free rate is the sea level, crypto is the boat. You do not argue with the tide. You read it, and you position for where it is going.
Follow the stablecoin float. At a 5% risk-free rate, the reserve income on a hundred-billion-dollar float is a multi-billion-dollar annuity, and that annuity — not transfer volume, not payment rails — is the business. Tether's reserve income has long dwarfed its fee revenue; at 5% the gap becomes a chasm. The peg is not the innovation. The float is. A stablecoin is a money-market fund that pretends to be a payment network, and at 5% the fund finally pays.
This reframes the entire regulatory conversation. The licensing regimes moving through Hong Kong, Washington, and Brussels are not debates about innovation. They are fights over who gets to hold the float and collect the risk-free spread. Regulation is never about the technology. It is about the rent. The jurisdiction that wins the stablecoin licensing race does not win because it loves decentralization. It wins because it captures the float and the Treasury income attached to it.
When I audited contract flows in Cape Town in 2017, tracing liquidity by hand for an exchange most people have forgotten, I learned to stop reading the pitch and start reading the ledger. The ledger on stablecoins says: this is a Treasury fund with a marketing department. That is not a criticism. It is the most honest product in crypto.
Now stress-test the other flagship claim. Gold pays no yield. Bitcoin pays no yield. When cash pays 5%, the carry cost of holding a non-yielding asset stops being theoretical and becomes a line item. Gold's century of data is unambiguous: it struggles when real rates rise. Bitcoin's correlation to the Nasdaq means it gets hit through two channels at once — the discount-rate channel that reprices every long-duration asset, and the risk-asset channel that drains speculative capital once the safe alternative finally pays.
The "digital gold" thesis needs a low-real-rate world to breathe. We do not live there. That does not make Bitcoin worthless. It makes Bitcoin a duration asset — a very long-duration bet on a future monetary regime — and long-duration assets are precisely what a 5% risk-free rate punishes first.
Now the part that actually matters for builders. The value of a DeFi protocol is, at bottom, a spread: the yield it offers minus the risk-free rate, adjusted for risk. At 0% risk-free, a 10% pool carries a 10-point spread and the TVL chart goes vertical. At 5% risk-free, the same pool carries a 5-point spread, and the marginal depositor — the one with an alternative — leaves for T-bills. TVL is not a function of code quality. It is a function of the spread. Hype is just liquidity with a distorted memory.
I learned that the hard way, and then I learned it again in 2022. When Terra/Luna unwound, the autopsy focused on the algorithmic peg. The real failure was upstream: an entire yield structure built on the assumption that dollar liquidity would keep expanding. It stopped. The peg did not break the system. The tide did.
Every crypto valuation, however hand-waved, is a discounted cash flow at heart. You project a terminal value and you discount it back at a rate anchored to the risk-free. Raise the anchor from 1% to 5% and terminal values fall by a third or more, mechanically, with no change in the underlying business. The "supercycle" was never really a technology thesis. It was a rates thesis in disguise. The supercycle was a low-rate artifact, and low rates are gone.
Subramanian's model makes the equity side explicit: it projects the S&P 500's annualized return over the next decade below the yield on the 10-year. That is the equity risk premium going negative — investors taking equity risk for less than the risk-free rate. When that happens, the marginal institutional dollar does not need a crisis to leave equities. It needs a spreadsheet. And crypto, priced at the far end of the risk curve, is the first thing sold when the spreadsheet says the safe asset wins.
The mechanism is not mysterious. When a government runs large deficits, it issues more bonds. When the central bank is simultaneously shrinking its balance sheet, it buys fewer. More supply, less demand, higher term premium, higher long yields. The 5% ten-year is the price of that imbalance. Any thesis that assumes it reverses has to explain what closes the deficit or who buys the duration. Subramanian offers demographics and policymaker intent. Those are real forces. They are also slow forces, and markets are not.
When everyone can earn 5% doing nothing, the burden of proof shifts onto everything that asks you to do something. That is the deepest change. It is not that crypto yields fell. It is that the baseline rose, and the baseline is what every risk asset is measured against. A world where cash pays nothing forces capital out the risk curve. A world where cash pays 5% pulls it back. Crypto's entire growth model depended on the first world.
Here is the feedback loop the perma-bulls keep missing. A 5% risk-free rate is a brake. It raises mortgage costs, corporate borrowing costs, and equity discount rates. High rates slow the economy. A slowing economy eventually forces the Fed to cut. And rate cuts are when crypto gets its liquidity back. So the pain is not permanent — it is a phase. But it is a phase you have to survive, and most of the market is positioned as if it does not exist.
This is where my 2026 work on AI and decentralized compute runs into the same wall. I led a team building verifiable AI training datasets on decentralized compute networks, and the pitch to institutions was always the same: the compute is cheaper, the data is verifiable, the network is trustless. Every one of those claims is true. And every one of them got discounted harder when the risk-free rate doubled, because the payoff sits years out. Distraction is the tax we pay for novelty — and in a bull market, the tax gets deferred, never cancelled.
The AI-crypto convergence is real. It is also long-duration, capital-hungry, and priced off the same curve as everything else. A 5% risk-free rate does not care how good your model is. It cares when the cash arrives.
The consensus crypto read on Subramanian's call is simple: bonds are back, so risk assets are out. Lazy. Here is the steel-manned opposite. If the 10-year has a ceiling, then the risk-free rate has a top, and a top in the risk-free rate is a floor for everything priced off it. Betting on bonds and betting on crypto become the same mean-reversion trade, just with different durations.
I do not buy the decoupling story now circulating — the idea that crypto trades on its own fundamentals, immune to the rate cycle. I have watched that story surface in every bull market since 2017 and evaporate in every drawdown. Correlation is a fair-weather friend. It reads near zero when prices rise and snaps to one when liquidity drains. Crypto is not decoupling. It is leveraging.
But here is the part the bears miss. If Subramanian is right — if 6% to 7% is a ceiling policed by policymakers and demographics — then the downside is bounded. The question stops being "can crypto survive high rates" and becomes "how long until the ceiling holds and the liquidity returns." That is a timing question, not a solvency question. The trap is believing you can call the turn. You cannot. You can only position for the spread between where the rate is and where it can go.
Liquidity does not leave. It just changes address.
Position for the spread, not the story. Watch the 10-year, not the timeline. When the risk-free rate rolls over — and mean reversion says it eventually will — the liquidity that left will return, and it will return to the assets with the deepest books, not the loudest narratives. The bond market just told you where the cycle sits. Most of crypto is still reading the wrong chart, and the ones who survive this phase will be the ones who read the rate before they read the roadmap.
