Credit Index Options

A CDS index like CDX or iTraxx already lets you buy or sell protection on a basket of 100+ corporate credits in one trade at a single quoted spread. A credit index option goes one layer further: the right, but not the obligation, to enter that index trade later at a spread fixed today. If credit spreads blow out before the option's expiry, a payer option (the right to buy protection, i.e. pay the fixed spread) becomes valuable, since it locks in the old, tighter spread while the market has moved to a wider one.

It's structurally the same idea as an interest rate swaption — the right to enter a swap at a pre-agreed rate — except the underlying is a credit spread, and "in the money" means spreads moved against the seller of protection rather than rates moving against the fixed-rate payer.

The payoff and the knockout feature

$$\text{Payer option payoff} = \max\big(S_T - K,\ 0\big) \times \text{Risky Annuity} \times N,$$

$S_T$ is the index spread observed at expiry, $K$ is the strike spread fixed at trade inception, and the "risky annuity" converts a spread difference into a present-value dollar amount by weighting it by the surviving portion of the index (it accounts for names that have already defaulted and dropped out). $N$ is the notional. A payer option pays off when the market spread $S_T$ ends up above the strike $K$ — protection buyers were right that credit got worse. A receiver option pays the mirror image, off when spreads tighten below the strike.

Credit index options carry one feature bond and equity options don't: a knockout clause. If a name in the index defaults before expiry, the option is typically adjusted or, in older conventions, can knock out entirely, because the underlying index itself just changed composition — protection on a defaulted name has already paid out through the index's own default settlement mechanism, separate from the option.

Worked example 1 — a payer option pays off

An investor buys a 3-month payer option on a CDX index, struck at $K = 80$ basis points, paying an upfront premium of 25bp of notional on a $10 million position. At expiry, the index spread has widened to $S_T = 110$bp, with a risky annuity of roughly 4.5. The payoff: (110 - 80) \times 0.0001 \times 4.5 \times \10{,}000{,}000 = $135{,}000. Net of the \25,000 premium paid, the investor nets $110,000 — a long-volatility, long-spread-widening payoff, with maximum loss capped at the premium if spreads had instead tightened or stayed flat.

Worked example 2 — implied volatility and strike selection

Two payer options on the same index and expiry, one struck at $K = 80$bp (near the current 78bp index spread) and one struck at $K = 100$bp (further out of the money), both trade. The at-the-money option implies 55% annualized spread volatility; the out-of-the-money implies 68%. Credit index options show a volatility skew that's upward sloping for payer strikes, since a sudden credit event pushes spreads up sharply (a jump) while tightening is a slower, calmer process — the market prices the fast, violent direction richer, the same asymmetry seen in VIX call skew for the same underlying reason: bad news moves faster than good news.

Read "call" here as the payer option and the strike in credit-spread basis points rather than dollars: the payoff line still kinks upward past the strike exactly the way a call option's does, except what's rising past the kink is the index's credit spread, not a stock price.

What this means in practice

Credit index options let a desk express a spread-widening or spread-tightening view with a defined maximum loss, unlike an outright index position, whose loss is theoretically as large as the full notional if the whole basket deteriorated. They're also the standard tool for overlaying convexity on a credit book — a fund can hold a core CDS index position and buy payer options as tail protection against a sharp spread shock, paying a known premium rather than the ongoing carry cost of an outright short-credit hedge.

Practice

1. If the index spread at expiry finishes below the payer option's strike, what does the payer option holder receive, and what was their maximum possible loss on the trade?
2. Why does a defaulting constituent name require special handling in a credit index option, when an ordinary equity index option doesn't need anything similar if one stock in the index goes to zero?