The trade that broke the market's back

May 11, 2010 in Finance

In a thoroughly ridiculous analysis, the WSJ is trying to pin Thursday's crash on a single options trade. Conveniently, the trade was made by Universa, the fund advised by black swan devotee Nassim Taleb. The body of the article uses phrases like "contributed" and "along with likely dozens of other trades", but the title's unanswered question ("Did a Big Bet Help Trigger 'Black Swan' Stock Swoon?") reveals the author's misdirection. Spoiler alert: a $7.5mm dollar trade did not impact markets, especially not on a day which saw as much volume as May 6. Even if Barclays was delta-hedging their exposure, it wouldn't cause markets to seize. Moreover, the funded trade was just 10 basis points of Universa's capital - so if you buy what the WSJ is selling, you should be very concerned about the market impact of their "real" positions.

Financial journalism doesn't need stories; it needs understanding. Once a grasp of what is realistic and what is absurd is had, the stories - far more interesting than this nonsense - will follow.

And can we put an end to every newspaper's characterization of options in terms of their payoff at some arbitrary level? In this article:

Through the trading desks at Barclays, Universa bought 50,000 options contracts, according to people familiar with the matter. The contracts would pay off about $4 billion should Standard & Poor's 500-stock index fall to 800 in June. It was at 1145 points at the time of the trade.

What a waste of time. This is like describing a stock position as "paying off $1mm if the S&P 500 reaches 1200!" - it only provides information about that single, unlikely point. Furthermore, it makes it hard to back into the real position. A standard option multiplier is 100, so 50,000 options is like owning 5mm "shares" of the SPX. Thus, at maturity the options will pay $5mm for every point the SPX lies under the strike level (isn't that a much better way of explaining the bet!)

However, $5mm/point would require a strike price of 1,600 in order to pay off $4 billion at 800, and in that case the options would cost at least $2.3B at 1145, not $7.5mm. So either the options have a non-standard multiplier or are non-standard themselves.

Please, journalists, learn to describe options properly. The payoff per point above or below the strike is an excellent starting point. And while we're at it, no more articles claiming a $7.5mm transaction was the trade that broke the market's back.

Leave a Comment

Previous post:

Next post: