Caps, Floors, and Swaptions

A swap is a handshake: you agree to swap a floating rate for a fixed one, no take-backs. It locks your rate dead — wonderful when rates rise, infuriating when they fall and you’re stuck paying above market. But what if you only want the good half of that deal? Protection when rates climb, freedom to enjoy it when they drop? That’s not a swap — that’s an option on rates. This lesson covers the three workhorses of the optional rate world: caps (ceilings on floating cost), floors (guaranteed minimum yield), and swaptions (the right to enter a whole swap later). You already know calls, puts, put–call parity, and that volatility drives premium — we’re just pointing all of that at interest rates instead of stock prices.

The optional cousins: rights, not obligations

Analogy. A swap is like signing a fixed-rent lease — your payment is nailed down whether the neighbourhood booms or busts. An option on rates is like travel insurance: you pay a small premium up front, you only collect if the bad thing happens, and if the trip goes fine you simply don’t claim. You’re never forced to use it.

That single word — optional — is the whole lesson. A swap is an obligation to exchange cash flows. A cap, floor, or swaption is a right you bought with an up-front premium, and you exercise it only when it pays. The asymmetry is the point: limited downside (your premium), open-ended protection.

Who wants what?

Interest-rate caps: a strip of caplets

The core idea. A cap is not one option — it’s a strip of caplets, one for each reset period of the underlying floating loan. Each caplet is a call option on the floating reference rate (think SOFR or the like) for that single period. When the index sets above the strike $K$, the caplet pays the difference; when it sets below, the caplet expires worthless and you simply pay the (lower) market rate.

The payoff of one caplet, settled for an accrual period of length $\tau$ (the year-fraction, e.g. $0.25$ for a quarter) on notional $N$:

Stack the caplet payoff on top of your floating interest bill and your net cost is capped at $K$. Above the strike the caplet refunds every basis point over the ceiling.

Worked example — a single caplet. Notional $N$ of $10,000,000, strike $K = 4\%$, accrual $\tau = 0.25$ (a 3-month period). The reference rate sets at $r_{\text{ref}} = 5.2\%$.

• In-the-money amount: $r_{\text{ref}} - K = 5.2\% - 4\% = 1.2\% = 0.012$.
• Payoff $= N \cdot (r_{\text{ref}} - K) \cdot \tau =$ $10,000,000 $\times\, 0.012 \times 0.25 =$ $30,000.

That $30,000 lands at the end of the period and exactly offsets the extra interest you owe above 4% — your effective cost for the quarter is capped at the strike. If instead the rate had set at $3\%$, the caplet pays $\max(3\% - 4\%, 0) = 0$, and you simply enjoy the cheap 3% loan.

Interest-rate floors: a strip of floorlets

A floor is the mirror image: a strip of floorlets, each a put option on the reference rate. A floating-rate investor — someone receiving the floating coupon — fears rates falling, so they buy a floor to guarantee a minimum yield. Each floorlet pays:

When the rate sets below $K$, the floorlet tops your income back up to the strike; when it sets above, the floorlet lapses and you keep the higher market coupon.

Worked example — a single floorlet. Notional $N$ of $20,000,000, strike $K = 4\%$, accrual $\tau = 0.25$, and the rate sets at $r_{\text{ref}} = 2.5\%$.

• In-the-money amount: $K - r_{\text{ref}} = 4\% - 2.5\% = 1.5\% = 0.015$.
• Payoff $=$ $20,000,000 $\times\, 0.015 \times 0.25 =$ $75,000.

The floorlet pays $75,000, lifting your effective yield for the quarter back up to 4% even though the market only offered 2.5%.

Cap–floor parity and collars

Here’s where your options training pays a dividend. Cap–floor parity is the rate-world twin of put–call parity:

Buy a cap and sell a floor at the same strike $K$, and the two strips collapse into a payer swap that pays fixed at $K$. Why? At every reset, either the cap pays you $\max(r - K, 0)$ or the floor you sold costs you $\max(K - r, 0)$ — and $\max(r-K,0) - \max(K-r,0) = r - K$ for every $r$. That’s exactly the floating-minus-fixed cash flow of a payer swap. A neat corollary: when the strike equals the par swap rate, the cap and the floor have equal value (the payer swap is worth zero at par), so the package costs nothing net.

A collar uses the same building blocks but for cheapness, not replication: long cap + short floor at different strikes. You buy the cap for protection and fund it by selling a floor, accepting a floor on your savings in exchange for a smaller (or zero) up-front premium. You give up some downside benefit to cheapen the upside protection.

Swaptions: the option to enter a swap

A swaption is an option on a swap: the right — but not the obligation — to enter a swap at a preset fixed rate on a future date. Unlike a cap (many small period-by-period decisions), a swaption is one all-or-nothing decision: on the exercise date you either step into the whole underlying swap or you walk away.

• Payer swaption = the right to pay fixed (receive floating). It gains if rates rise — your locked-in fixed rate is now below market, a bargain. This is the “I’m going to borrow later” hedge.
• Receiver swaption = the right to receive fixed (pay floating). It gains if rates fall — receiving an above-market fixed rate is gold. This is the “I’ll have money to invest later” or “protect a bond” hedge.

Uses worth knowing:

1. Hedging future borrowing. Lock the right to pay fixed before you’ve even drawn the loan (payer swaption).
2. Monetizing callable bonds. When a company issues a callable bond, it keeps the right to redeem early if rates fall — that embedded option is economically a receiver swaption the issuer is long (and the bondholder is short). Strip it out and you can value or hedge the call feature directly.
3. Expressing rate-vol views. Because swaption value rises with implied volatility, traders buy/sell swaptions to bet purely on how turbulent rates will be, not just their direction.

Volatility is the price

These are options, so the lesson from Black–Scholes carries straight over: volatility is the price. Rate options are quoted and priced off interest-rate volatility using Black-76 (a forward-rate flavour of the Black–Scholes machinery) — each caplet/floorlet is priced as a Black-76 option on its forward rate, and a swaption as a Black-76 option on the forward swap rate. We won’t derive it; just hold the intuition.

Because vol is the input, the market quotes a whole cap/swaption volatility surface — implied vols varying by expiry, by underlying tenor, and by strike (the smile/skew). A swaption desk trades that surface the way an equity desk trades the equity vol surface: the “price” of a rate option is really a volatility number plugged into Black-76.

When to use which

Reach for a cap when you carry floating-rate debt and want a ceiling on cost while keeping the savings if rates fall. Reach for a floor when you hold floating-rate assets and want a guaranteed minimum yield. Use a collar when the up-front premium of a cap is too rich and you’re willing to surrender the deepest rate savings to fund it. Reach for a swaption when the decision is binary and forward-dated — you’ll commit to a whole swap later (a future loan, a bond call, a planned investment) and want the right, not the obligation, to do it at today’s rate. And whenever you price any of them, remember you’re really trading volatility through Black-76, not just a directional view.

Putting it together

Swaps lock a rate both ways; caps, floors, and swaptions hand you the right without the obligation, paid for with an up-front premium. A cap is a strip of caplets — each a call on the rate paying $N\max(r-K,0)\tau$ — that bounds a borrower’s cost. A floor is a strip of floorlets — puts paying $N\max(K-r,0)\tau$ — that guarantees an investor’s yield. Cap–floor parity says long cap minus short floor at one strike is a payer swap (the rate twin of put–call parity), and a collar mixes a long cap with a short floor to cheapen protection. A swaption is a single decision to enter a whole swap: payer swaptions gain when rates rise, receiver swaptions when they fall, and they’re the engine behind callable-bond optionality. All of it is priced off interest-rate volatility via Black-76 — vol is the price. Mind the two traps: a cap is many options, and a swaption is not a cap.

Next — hedging a rate book with swaps — we put the whole kit to work, neutralizing the directional and curve risk of an entire portfolio with a blend of swaps and these optional overlays.