Deriving the payoff, by cases (there's only one case)
Take the long side first. At maturity T the long is
obligated to buy the asset for K, and that asset is worth
S_T on the open market. If they wanted to, the long could
immediately sell what they just bought at the market price and pocket the difference:
\text{long payoff} = S_T - K.
Compare that with a call option's payoff, (S_T - K)^+: the call
needed a case split, because the holder could always walk away when
S_T < K. The forward's long has no such escape — the formula is
S_T - K in every case, full stop, with no
\max(\cdot, 0) clamping it at zero. When
S_T < K that payoff is simply negative: the long loses money, for
real, exactly as if they'd bought high and sold low.
The short side is obligated to sell at K an asset
worth S_T on the market — the mirror image, and (since it's a
zero-sum bilateral contract) the exact negative of the long's payoff:
\text{short payoff} = K - S_T = -(S_T - K).
For a contract with delivery price K and maturity
T, at maturity the payoff to each side is a linear
function of the terminal price S_T alone — no kink, no floor at
zero:
- Long (agreed to buy): S_T - K, slope
+1, unbounded gain and unbounded loss.
- Short (agreed to sell): K - S_T, slope
-1, the exact opposite of the long's payoff.
This is what "symmetric risk" means for a forward: the long gains a dollar for every dollar
S_T rises above K, and loses a dollar for
every dollar it falls below — in both directions, without limit. A long option
position, by contrast, has a floor: its worst case is losing the premium it paid, full stop.
That missing floor is the single biggest structural difference between a forward/futures
position and an option position, and it's why margin (Lesson 4) exists at all — an exchange
needs collateral against a loss that, in principle, has no bound.
Worked example: pricing the jet-fuel hedge exactly
Return to the airline from Lesson 1: long a six-month forward on
N = 2{,}000{,}000 gallons of jet fuel at
K = \$2.68. The per-gallon payoff formula is
S_T - K; multiply by the notional
N to get the total dollar payoff:
\text{payoff} = N(S_T - K).
With S_T = \$3.10:
2{,}000{,}000 \times (3.10 - 2.68) = 2{,}000{,}000 \times 0.42 = \$840{,}000
— exactly the saving computed by hand in Lesson 1, now read straight off a single linear
formula. With S_T = \$2.20:
2{,}000{,}000 \times (2.20 - 2.68) = 2{,}000{,}000 \times (-0.48) = -\$960{,}000
— a loss, because the formula doesn't clamp at zero the way an option's would.
Put-call parity — coming properly in
Mathematics of Finance
— says that at the same strike K and maturity
T,
(S_T - K)^+ - (K - S_T)^+ = S_T - K.
Read the left side as "own a call, write a put, same strike and maturity." When
S_T > K your call is worth S_T - K and
the put you wrote is worthless — net S_T - K. When
S_T < K your call is worthless but the put you wrote costs you
K - S_T — net -(K - S_T) = S_T - K
again. Either way, the combination reproduces the long forward payoff exactly. A forward
isn't a new kind of risk at all — it's the specific bundle of options that cancels out all
the optionality and leaves pure, symmetric exposure to S_T.
Because every option payoff you've seen so far has a floor at zero, it's an easy habit to
assume all derivative payoffs do. They don't. A long forward's
payoff S_T - K can be any real number, arbitrarily negative if
S_T collapses far enough below K — there
is no built-in stop-loss. This is precisely why a wrong-way forward or futures position can
wipe out far more than an initial stake, and why exchanges force futures traders to post and
continuously top up margin (Lesson 4) — collateral against a downside that a plain option
buyer simply doesn't have.