The 2005 Convertible Arb Unwind

A convertible bond pays a coupon like ordinary debt but can be converted into shares of the issuing company, so it embeds a call option on the stock. Convertible arbitrage buys the bond and shorts the right amount of stock to cancel out the directional risk — delta-hedging — leaving a position that profits mainly from the bond's embedded volatility being underpriced. It's a market-neutral strategy on paper. In practice, by 2005, hundreds of funds were running nearly identical versions of it.

In May 2005, Standard & Poor's and Moody's cut General Motors and Ford's credit ratings to below investment grade within days of each other. Both companies had large convertible bonds outstanding, widely held by convertible arb funds. As credit spreads on GM and Ford paper blew out, the bonds' prices fell sharply — more than the delta hedge alone would predict, because credit risk and equity-option value moved together in a way many models hadn't captured cleanly.

Why the pain spread beyond GM and Ford

The losses didn't stay contained to funds actually holding GM and Ford paper. Convertible arb desks share a structural feature: they're often leveraged, and when losses hit, funds sell their most liquid convertible positions first to raise cash and meet margin calls — regardless of whether those positions had anything to do with GM or Ford. That selling pushed down the entire convertible bond market, hurting funds with no auto exposure at all, purely because they held convertibles that suddenly became the easiest thing to sell in a hurry.

Worked example

A fund holds a convertible bond priced at 98 (i.e. $980 per $1,000 face value), delta-hedged by shorting stock so that a 1% move in the underlying equity changes the bond's value by roughly the same amount the short position offsets. When the GM downgrade hits, the bond's price falls to 90 over two weeks — an 8-point drop — even though the underlying stock only moved a few percent, because the widening credit spread devalued the bond's fixed-income component in a way the delta hedge never covered. On $50 million face value, that's a $4 million loss the hedge did nothing to prevent, purely from credit spread widening. Funds facing margin calls from losses like this then sold other convertible positions — say, an unrelated retailer's convertible bond — pushing that price down too, spreading the damage well past the automakers.

What this means in practice

The episode pushed convertible arb funds to explicitly model and hedge credit spread risk, not just equity delta and volatility, and to hold more liquidity buffers against the risk of a shared, crowded unwind. It's now a standard cautionary case for why "market-neutral" describes exposure to one risk factor, never exposure to all of them.