A friend texted me while he was floating a mortgage rate into a Fed meeting week, asking whether there was any way to insure the lock. There is, sort of. Kalshi lists contracts on what the Fed does at each FOMC meeting, they cost cents each, and a small account can build a genuine hedge out of them. The catch is that sizing it wrong means you either overpay for protection or protect the wrong six weeks, so here is the arithmetic I walked him through.
What the contract pays, and what you are actually exposed to
Mechanics first. A Fed meeting contract is a binary on the target rate decision at one specific meeting, split into brackets like cut 25, hold, or hike 25. Each contract pays a dollar if its bracket happens, zero if not, and the price is the market's implied probability. If the hold bracket trades at 30 cents, buying it costs 30 cents and nets you 70 cents if the Fed holds. That asymmetry is what makes these usable as insurance.
First figure out which of two kinds of rate exposure you have, because they size completely differently.
Carry exposure is cash-flow exposure. A HELOC or a floating business loan tied to prime moves nearly one for one with the fed funds rate. If you were counting on a cut and the Fed holds, you pay the full 25 basis points of extra interest for as long as the hold persists.
Mark-to-market exposure is expectations exposure. A bond fund, or a mortgage rate you have not locked yet, prices off what the market believes about the whole path of rates. Part of every outcome is already priced in, so only the surprise portion moves you. If a hold was trading at 30 percent and it happens, near-term yields reprice by roughly the unpriced 70 percent of the bracket, call it 17 or 18 basis points rather than the full 25.
Sizing from dollars per basis point
The workflow is five steps, and the first one has nothing to do with any exchange.
- Compute your dollar sensitivity per basis point. For floating debt it is balance times 0.0001. For a bond position it is value times duration times 0.0001.
- Pick the bracket that hurts you and an honest horizon for how long the damage lasts.
- Multiply those into a single dollar loss for that scenario.
- Divide the loss by profit per contract, one dollar minus the price, to get the contract count.
- Compare the total premium to the loss, and skip the hedge if it is more than a modest fraction of it.
Worked example one, floating debt. Say you carry a $200,000 HELOC and the market is split on whether the next meeting brings a cut. A 25 basis point difference on that balance is roughly $500 a year of interest. A hold usually means the cut arrives a meeting or two later, so pick a horizon like six months, putting the damage around $250. With the hold bracket at 30 cents, each contract nets 70 cents in that scenario, so you need around 360 contracts at a cost of roughly $108. If the hold happens you collect about $360, covering the extra interest with the premium netted out. If the cut happens you are out $108 and your loan just got cheaper, which is insurance you were happy to waste.
Worked example two, a duration-heavy portfolio. A $150,000 bond fund with a duration around six moves about $90 per basis point. Here the surprise math does something convenient. Your loss on a hold is roughly $90 times 25 basis points times the unpriced probability, and the profit per contract is also the unpriced probability, so the probability cancels out of the sizing. The contract count is simply dollars per basis point times bracket width, 90 times 25, or about 2,250 contracts, roughly $675 of premium at 30 cents. You do not have to guess how much is priced in, the market price handles that, and the premium scales with the probability of the bad outcome, the way insurance should.
That $675 is real money on a $150,000 portfolio, and there are typically eight meetings a year, so this is event insurance for moments that matter, the week before you lock, the month you rebalance, a meeting where you think the market is badly split. Running it at every meeting will bleed premium.
The mortgage lock is the messiest case. Thirty-year mortgage rates track the ten-year Treasury and mortgage bond spreads, which price the whole expected path plus a term premium, so a single meeting passes through only partially. A rough rule: 25 basis points on a thirty-year mortgage changes the payment by roughly $15 a month per $100,000 borrowed, about $60 a month on a $400,000 loan. If a surprise hold bleeds maybe 10 to 15 basis points into mortgage quotes and you would carry the loan around five years before refinancing or moving, the ballpark damage is somewhere near $1,500. Size against that rather than the full bracket, or you will pay double.
The basis risk between one meeting and the path
The part that catches people is the mismatch between what the contract settles on and what your exposure prices. The contract settles on a single decision, while your mortgage and your bond fund price the cumulative path. The Fed can hold and long rates can still fall because the statement reads dovish, or cut and long rates can rise because the market treats the cut as inflationary. Both have happened historically. When path and meeting diverge, the contract pays while your rate never moved, or expires worthless while your rate moves against you anyway. You cannot eliminate this with meeting contracts, only be honest that it exists and size accordingly.
Two habits shrink it. First, only hedge meetings where the market is genuinely split. If the bracket you fear trades at 90 cents, there is almost no surprise left to hurt you and almost no payout left to collect. The tradeoff is that exchange fees on these venues scale with price times one minus price, largest near 50 cents, exactly where the split meetings live. At a few thousand contracts the fees stop being a rounding error, so put them in the arithmetic.
Second, match the instrument to the exposure. The venues also list cumulative contracts on where the target rate will sit by some future date, and those map far better onto a mortgage lock or a duration position because they price the path instead of one decision. My split is meeting contracts for carry, path contracts for anything that trades on expectations.
The failure modes I see most are sizing a carry hedge to a full year when the exposure realistically lasts one or two meetings, buying the 90 cent bracket because it feels safe, and being surprised that the good scenario costs you the premium. That last one is just what insurance costs, but it should stay a small fraction of the loss you are protecting. I keep these implied probabilities in my Blockcircle watchlist next to the rest of my rate exposure, because the drift between meetings tells you when the market has stopped being split and the hedge window has quietly closed.
Do the dollars-per-basis-point number on your own balance sheet before you look at a single contract price. If it comes out small, keep the premium and take the meeting as it comes.