Why Swap Errors Often Come From Signs and Timing
A swap valuation can use correct interest rates and discount factors yet still produce the wrong answer if a cash flow is assigned to the wrong party, a rate fixing is treated as a forecast, or a notional amount is inserted as though it were exchanged. These errors are particularly consequential in hedge calculations: reversing the swap direction can turn an intended offset into an added exposure.
The safest method is to set a valuation perspective, list the remaining contractual cash flows by date, and discount each leg using the appropriate discount factors. The discussion below assumes a plain-vanilla interest-rate swap with no principal exchange unless the contract explicitly specifies one.
Let \(N\) be the notional principal, \(K\) the annual fixed rate, \(\alpha_j\) the accrual fraction for payment period \(j\), \(D_j\) the discount factor to payment date \(t_j\), and \(L_j\) the annualized floating rate applying to that period. If all remaining floating rates are known or projected consistently, the value to the fixed-rate receiver is
This expression assumes that the cash flows being valued are the remaining periodic interest payments, with no principal exchanged. Each leg’s present value is calculated separately; the difference is the value to one side of the swap. If the contract does exchange principal, the principal cash flows must also be included at their specified dates.
Deriving the Floating-Leg Shortcut
Under a single-curve framework, suppose each floating rate is the forward rate implied by the same discount factors used for valuation. For a period from \(t_{j-1}\) to \(t_j\), that simple forward rate satisfies
Multiplying by the accrual fraction and discount factor shows why the projected floating coupons telescope:
Therefore, if the swap is valued at a reset date, the next coupon is reset to the current market rate, and all later coupons are projected from this same curve, the present value of the floating coupons is
The last equality follows because each intermediate discount factor appears once with a positive sign and once with a negative sign. This shortcut does not apply unchanged when the next floating coupon has already been fixed at an earlier reset date. In that case, use the known rate for that coupon and project only the later ones. Nor should the shortcut be assumed in a multi-curve valuation where floating rates are projected from a curve different from the discount curve.
Worked Example: Correcting the Payer-Receiver Sign
A hypothetical two-year annual-pay swap has notional \(N=\$1{,}000{,}000\) and fixed rate \(K=5\%\). It is valued just after a reset date, so the first floating coupon is set to the current market rate. The discount factors are \(D_1=0.96\) and \(D_2=0.90\), with \(D_0=1\). Annual accrual fractions are \(\alpha_1=\alpha_2=1\). Find the value to each side.
Method. Calculate the present value of each leg, then subtract in the stated party’s direction. This avoids treating “payer” as the party who receives the quoted fixed rate.
Step 1: Find the floating rates implied by the curve.
The first floating coupon is \(\$1{,}000{,}000(0.0416667)(1)\), or approximately \(\$41{,}666.67\). Its present value is \(\$41{,}666.67(0.96)=\$40{,}000\). The second coupon is approximately \(\$66{,}666.67\), with present value \(\$66{,}666.67(0.90)=\$60{,}000\). Thus the floating-leg present value is \(\$100{,}000\).
Step 2: Find the fixed-leg present value. Each annual fixed coupon is \(\$1{,}000{,}000(0.05)(1)=\$50{,}000\). Discounting gives
Step 3: Apply the perspective. The fixed-rate receiver receives \(\$93{,}000\) in fixed-leg value and pays \(\$100{,}000\) in floating-leg value, so \(V_{\text{receiver}}=\$93{,}000-\$100{,}000=-\$7{,}000\). The fixed-rate payer’s value is \(+\$7{,}000\).
Check. The floating-leg shortcut gives \(\$1{,}000{,}000(1-0.90)=\$100{,}000\), agreeing with the coupon calculation. A common mistake is to report \(+\$7{,}000\) for the receiver by subtracting the legs in the payer’s order.
Reset Timing: A Fixing Is Not a Forecast
A floating rate is usually set at the start of its accrual period and paid at the end. Once a rate has been fixed, a later change in the current forward curve does not replace that contractual fixing. A valuation should therefore distinguish three items: the period’s accrual fraction, the rate already fixed for a coupon, and any rates still to be projected.
Worked Example: Using a Known Fixing for the Next Coupon
A hypothetical swap has notional \(N=\$2{,}000{,}000\), fixed rate \(K=5\%\), and two remaining half-year periods. The discount factors to the payment dates are \(D_1=0.98\) and \(D_2=0.95\). The first floating rate was already fixed at \(6\%\) at the previous reset. The second period’s projected annualized rate is \(4\%\). Find the current value to the fixed-rate receiver. Assume the quoted floating rates apply over their respective half-year accrual periods.
Method. Value the already-fixed coupon using \(6\%\), and the later projected coupon using \(4\%\). Do not substitute the second period’s forward rate into the first period.
Step 1: Calculate the floating coupons. For the first period, the coupon is \(\$2{,}000{,}000(0.5)(0.06)=\$60{,}000\), with present value \(\$60{,}000(0.98)=\$58{,}800\). For the second period, the coupon is \(\$2{,}000{,}000(0.5)(0.04)=\$40{,}000\), with present value \(\$40{,}000(0.95)=\$38{,}000\). Therefore, the floating-leg present value is \(\$96{,}800\).
Step 2: Calculate the fixed coupons. Each is \(\$2{,}000{,}000(0.5)(0.05)=\$50{,}000\). Thus,
Step 3: Subtract from the receiver’s perspective. The fixed-rate receiver’s value is \(\$96{,}500-\$96{,}800=-\$300\). The payer’s value is \(+\$300\).
Check and trap. If \(4\%\) were incorrectly used for the already-fixed first coupon, its present value would be \(\$2{,}000{,}000(0.5)(0.04)(0.98)=\$39{,}200\), understating that floating coupon’s present value by \(\$19{,}600\). The resulting receiver value would incorrectly be \(+\$19{,}300\). The fixing date, not the valuation date alone, determines which rate applies.
Notional Principal and Actual Cash Flows
In a standard single-currency interest-rate swap, the notional principal is the amount used to calculate interest. It is not itself paid from one party to the other. Consequently, adding \(N D_j\) as a cash flow at maturity in an ordinary swap changes the contract being valued.
This convention is different from a contract that explicitly exchanges principal, such as many currency swaps. If principal is exchanged, write the payment in the cash-flow schedule with its correct payer, receiver, and date, and discount it along with the interest payments. The label “notional” alone does not establish whether principal changes hands.
A useful check is to ask whether the contract says the principal is exchanged. If not, the swap’s value is the difference between the interest legs, not the value of a notional principal repayment added to one side. Also check that a payment at the valuation date has not already been settled and should therefore be excluded from remaining cash flows.
Hedge Direction: Match the Sign of the Exposure
A fixed-rate receiver generally has positive exposure to fixed-rate bond values: when market rates rise, the present value of fixed receipts tends to fall. A fixed-rate payer has the opposite direction of exposure, all else equal. Thus, a position that loses value when rates rise is commonly hedged with a fixed-rate payer swap, not a receiver swap.
The hedge should be selected from the signed sensitivity of each position under the same rate change and valuation assumptions. A present-value sensitivity, or PV01, is the approximate change in value for a one-basis-point change in rates. The sign matters: a position with negative PV01 loses value for an upward one-basis-point move, while a position with positive PV01 gains value for that move.
Worked Example: Choosing the Hedge Direction and Notional
A hypothetical fixed-income portfolio has a signed PV01 of \(-\$1{,}800\) per basis point for the position being hedged. Under the same valuation assumptions, a fixed-rate payer swap with notional \(\$1{,}000{,}000\) has signed PV01 \(+\$1{,}200\) per basis point. Assume swap PV01 scales linearly with notional over the hedge sizes considered. Find the swap direction and notional that neutralize this first-order exposure.
Method. Add signed PV01s and choose the swap amount that makes the total zero. This directly avoids confusing the party that pays fixed with the party that benefits from a rate increase.
Step 1: Choose direction. The portfolio has negative PV01, so it loses value when rates rise by one basis point. The given fixed-rate payer swap has positive PV01, so it gains value for that move. It is the direction that can offset the portfolio.
Step 2: Scale the swap. Let \(x\) be the swap notional in millions of dollars. Its PV01 is \(+\$1{,}200x\) per basis point. Set the combined sensitivity to zero:
The required hedge is a fixed-rate payer swap with notional \(\$1{,}500{,}000\). Its PV01 is \(1.5(\$1{,}200)=+\$1{,}800\) per basis point, so the combined PV01 is \(-\$1{,}800+\$1{,}800=\$0\).
Check and trap. A fixed-rate receiver swap of the same notional would have the opposite sensitivity, \(-\$1{,}800\) per basis point under the stated linear-scaling assumption. Adding it to the portfolio would produce \(-\$3{,}600\) per basis point, doubling rather than hedging the first-order exposure. A zero PV01 does not eliminate curve-shape, basis, or cash-flow mismatch risk.
A Reliable Error-Checking Sequence
Before accepting a swap value or hedge recommendation, make the following checks in order. They are not substitutes for valuation, but they expose many sign and timing errors quickly.
Write “fixed-rate receiver” or “fixed-rate payer” before subtracting the legs. For the receiver, use fixed receipts minus floating payments.
Record each payment date, accrual fraction, fixed rate, floating fixing or projection, and discount factor. Exclude payments that have already settled.
Do not add a notional exchange to a standard interest-rate swap unless the contract explicitly calls for that payment.
Calculate or obtain signed sensitivities under the same rate shock, then select a position whose sensitivity offsets the exposure.
These checks also help identify when a shortcut is inappropriate. A floating-leg telescoping formula depends on its curve and reset assumptions; it cannot repair a mistaken fixing, omitted contractual principal payment, or inconsistent discounting basis.
Check Your Understanding
For each problem, keep the stated valuation perspective and timing explicit.
- A one-period swap has \(N=\$500{,}000\), \(\alpha=0.5\), \(D_1=0.97\), fixed rate \(5.2\%\), and a floating rate already fixed at \(4.8\%\). Find the value to the fixed-rate receiver. (Answer: \(+\$970.00\))
- A two-period annual-pay swap is valued just after reset, with \(N=\$1{,}000{,}000\), \(K=6\%\), \(D_1=0.95\), and \(D_2=0.88\). The floating rates are implied by those discount factors under a single-curve model. Find the value to the fixed-rate receiver. (Answer: \(-\$10{,}200.00\))
- A swap has \(N=\$3{,}000{,}000\), two remaining quarterly periods, fixed rate \(5\%\), and discount factors \(D_1=0.99\), \(D_2=0.975\). The first floating rate has already been fixed at \(4\%\); the second period’s projected rate is \(6\%\). Find the fixed-rate receiver’s value. (Answer: \(+\$112.50\))
- A portfolio has signed PV01 \(-\$2{,}500\) per basis point. A fixed-rate payer swap with notional \(\$1{,}000{,}000\) has PV01 \(+\$1{,}000\) per basis point, and sensitivity scales linearly. What direction and notional give a zero combined PV01? (Answer: Fixed-rate payer, \(\$2{,}500{,}000\))
- A single-currency fixed-for-floating swap uses \(\$4{,}000{,}000\) as the notional to calculate coupons. The confirmation specifies periodic interest payments but no exchange of principal. Should a \(\$4{,}000{,}000\) maturity cash flow be included in the swap valuation? (Answer: No; include only the specified remaining interest cash flows.)