Back to the oil flash crash: the role of options studied

In the aftermath of the sharp selloff in NYMEX crude futures nearly two weeks ago, analysts and pundits remain conflicted about how and why the front-month contract fell more than $3/b in just one minute. As we wrote earlier, analysts and pundits just don’t seem to know what happened.

Unfortunately, NYMEX crude options data available to us at Platts–which often tell a story beneath the surface–does not provide much of an answer either. Sometimes any one, or a combination of, volume, open interest, implied volatility and put/call ratios offer a small glimpse into the minds of traders both before and after big moves in the benchmark price.

On September 17, oil markets reacting to a Federal Reserve announcement of the QE3 stimulus program infused an energetic but short-lived bull run, followed by the bizarre, sharp and unexplained crash the following Monday. All this occurred against a backdrop of much chatter regarding a potential release from the US Strategic Petroleum Reserve.

Situations like these often present analysts with an opportunity to go looking for evidence in the options pits. Or so one might think. Even Price Futures Group analyst Phil Flynn chalked up the much of the crash to the expiry of the October NYMEX crude options contract, which expired after trading on the day of the crash.

But this is simply not reflected in the data. Open interest in NYMEX October and November crude options shows puts and calls were steady both before the FOMC announcement and after the crash. This suggests that large moves into and out of options did not occur in the lead-up to or during the direct aftermath of large-scale events that took futures unexpectedly by storm.

Open interest for NYMEX October crude puts in the $80-$96/b range was flat between September 7 and September 17, when the contract expired. Open interest in November puts at the same strikes were similarly flat in the initial days after the roll.

The only exception among heavily-traded strike prices was the November $80 put. In a sharply falling market, you’d expect open interest in forward months to rise. Open interest in that contract had reached 13,325 lots on September 17, up from 6,639 lots a week prior.

Conversely, open interest in the November $90 put stayed comparatively flat over that same period, at 12,760 lots on September 17 from 9,807 lots a week prior.

Volume data on the day of the crash offers little more insight. The Monday of the crash, the largest movers were the December $85 put, at 4,476 lots and the January $120 call, at 5,945 lots, according to Parity Energy data.

Tuesday’s volume was a little more in line with what would be expected, with 9,597 lots trading for the December $95 put, which is a fairly large number. But again, this is the day after the crash and reflects a continued bearish outlook on crude futures, rather than answers for why Monday happened in the first place.

So it’s clear that there was no large-scale selloff in the options market, nor was there a large scale buy-in to hedge against crude falling further.

Implied volatility can also be instructive. Often before or after a major price swing, implied volatility data in crude options moves up or down. These moves suggest a possible tradable range for the underlying commodity. When implied volatility is high, the range is wide, and difficult to predict. When it is low, the opposite tends to hold.

Implied volatility drops as the underlying commodity seems to pick a direction and sticks with it. This type of scenario lessens the allure of an option. Seen by many as insurance, options are a hedge against direct exposure to futures. When the underlying contract is showing a clear tendency up or down, direct exposure to plain vanilla futures becomes less risky.

The Monday of the crash, “at-the-money” implied volatility for the three front-most contracts was slightly elevated before tapering off, according to Kip Perkins of Parity Energy. But after the floor dropped out from below the underlying crude contract, only then did implied volatility jump. And by the end of the day, with futures having rallied off the September 17 drop of almost $4/b, the put/call ratio tilted in favor of calls, suggesting a correction to the upside was possible during the next day’s trading. But that didn’t end up happening, as the market trended down further.

Although put/call ratios measure the market’s general tendency one way or the other, the measure is often misleading. A large number of put buyers can signal that a market bottom is not far away, according to conventional wisdom, just as too many call buyers can show that the market could be close to its top. But the options data around the crash don’t signal that this occurred, damaging the “it was options” argument.

And we remain scratching our head.

Creative Commons License

To subscribe or visit go to:  http://www.platts.com