A friend walked me through a delta-neutral farming loop a while back, one of those setups where you stack a lending yield on top of a staking yield and land on a projected return in the high single digits. My first question was what a three-month Treasury bill paid that same day. He had no idea. Most people running money in crypto have no idea, which is odd, because that one number is the benchmark every trade they make is quietly competing against.
A T-bill is short-dated debt of the US government, and its yield is what you earn for taking about as little risk as modern finance offers. Hold it to maturity and there are no drawdowns, no exchange that can freeze withdrawals, no contract that can get drained. When people say the risk-free rate, this is roughly the number they mean. The entire case for owning anything else rests on one comparison. If a position does not offer meaningfully more than the bill, after accounting for all the extra ways it can go wrong, the rational move is to own the bill.
Every risky dollar has a boring alternative
The mechanism is opportunity cost, and it runs in both directions.
When bills yield close to nothing, cash is dead weight. It earns nothing while inflation chews on it, so capital goes hunting. It reaches for dividend stocks first, then credit, then growth equity, then venture, and eventually it finds the far end of the risk curve, which is where crypto lives. There were long stretches after 2008, and again in 2020, when the hurdle was effectively zero. A token with a story and some momentum was competing against an asset that paid nothing, and almost everything cleared that bar because there was no bar.
When bills pay something real, call it roughly four or five percent, the chain runs in reverse. A pension fund, a corporate treasury, or a regular person with a brokerage account can get a solid nominal return for doing nothing. Every speculative position now has to argue with that. The marginal dollar that might have chased a mid-cap altcoin can sit in a money market fund instead, and plenty of it does. This is a large part of why risk assets tend to struggle when short rates rise quickly. Capital is being paid, for the first time in a while, to wait.
Why crypto feels it more than stocks do
A stock has earnings, and you can at least argue about what those earnings are worth when the discount rate moves. Most crypto assets have no cash flows at all. Their value is either a claim on adoption years down the road or a bet on future inflows, and both of those get discounted hardest when the rate you are discounting with goes up. In bond language, crypto trades like the longest-duration asset in the room. Small moves in the risk-free rate swing the present value of a distant, uncertain payoff far more than they swing the value of a company earning money today.
The plumbing reinforces this. Perp funding rates, margin costs, and the cash-and-carry basis trade all key off short-term rates. If the annualized basis on a futures curve pays less than a bill, arbitrage capital has no reason to be there, and it leaves. If on-chain lending markets pay less than bills, stablecoin balances stop growing, because a stablecoin already carries issuer risk and depeg risk that a bill does not. None of this requires anyone to hold an opinion about price. The flows are close to mechanical.
I will flag the caveat before someone else does. This relationship is a tide, and it works over quarters. Crypto has rallied through high-rate stretches and dumped through easy ones plenty of times, because halvings, liquidity cycles, and pure narrative can overwhelm the rate effect for months at a stretch. I use the bill yield as background pressure on my sizing and my patience, and I gave up on using it to time anything shorter than a quarter.
The habit that does the work
Here is the practical version, and it takes about two minutes per trade.
- Look up the current three-month bill yield. It is free, it is published everywhere, and it takes ten seconds.
- Write down your honest expected annualized return for the trade, net of fees, funding, slippage, and tax drag if it applies to you.
- Subtract the bill yield from that number. Whatever is left is what you are actually being paid to carry the risk.
- Ask whether that leftover compensates you for the specific ways this position can die. Exchange failure, contract exploit, depeg, liquidity vanishing on the exit, your own fat finger at 3am.
- Size accordingly. Trades that clear the hurdle by a wide margin can get real size. Trades that barely clear it get token size or a pass.
My rule of thumb for anything with genuine tail risk is that I want a multiple of the bill yield, and for the messier stuff a large multiple. A farm quoting a couple of points above bills is offering me almost nothing to hold smart contract risk, and the honest answer is usually to skip it. Bills pay the same whether I check them or not, so I might as well check.
The failure mode I see most often is comparing against zero. Nine percent from a lending protocol sounds generous in a vacuum. Against a five percent bill it is four points of spread for taking risks that have wiped out entire platforms, and once you frame it that way a lot of popular trades stop making sense. The second failure mode is comparing gross numbers. The headline APY on a strategy and the return that survives fees, funding flips, and slippage are often not close, while the bill yield you actually receive is the bill yield.
None of this is an argument for avoiding risk. Some trades clear a five percent hurdle with room to spare, and when cash pays nothing the hurdle drops and being far out on the risk curve is a defensible place to be. The point is to know the number you are competing against before you size the position. It is one lookup, it changes slowly, and it is the cheapest piece of discipline available to anyone trading anything.