• 08 Jul 2013 at 11:40 AM

SEC Not Really Big Believers In Luck

A thing that the SEC sometimes does now is

  • notice when an anonymous foreigner buys a ton of otherwise thinly traded short-dated out-of-the-money call options just before a company announces big merger news that pushes the stock way up and makes those options suddenly hugely valuable, and
  • go to court to take away the anonymous foreigner’s possibly ill gotten gains.

What do you think of that? Like, on the one hand, my aesthetic sensibilities are offended, and my sense of fair play: really they ought to have some evidence of insider, as opposed to just lucky, trading. On the other hand it does seem like good police work, and if the anonymous foreigners want their money back they can always show up and dispute the SEC’s charges.1

Certainly the latest case, where the SEC froze two trading accounts of shadowy offshore figures who went and bought a lot of call options on Onyx Pharmaceuticals two weeks ago, just before Onyx announced last Sunday that it had received and rejected an acquisition proposal from Amgen, is pretty suspicious. From the SEC’s complaint:

Defendants in this action are either foreign traders or traders trading through foreign accounts whose timely purchases of Onyx calls generated profits of over $4.8 million. On information and belief, the Defendants are either located or trading through accounts located in the Canary Islands and Beirut. …

On June 26, 2013, on the day that Onyx’s board was considering the Amgen offer, one or more of the Defendants purchased 80 Onyx call options with a strike price of $80 (the “July 80 call options”) and 175 call options with a strike price of $85 (the “July 85 call options”), which were out of the money at the time based on Onyx’s closing price of $84.17 that day. These purchases deviated from the historical trading for these series of calls. The July 80 calls began trading on June 20, 2013, and the average daily volume was 16 contracts per day. The 80 July 80 calls purchased on June 26 represented an approximately 400 % increase over the average daily volume. The July 85 calls began trading on began trading on June 11, 2013, and the average daily volume was 6 contracts per day. The 175 July 85 calls purchased on June 26 represented an approximately 2,817% increase over the average daily volume.

So I at least would be intrigued to hear the non-fishy explanation. On the other hand: I’d be intrigued to hear the fishy explanation too! (“Onyx director got off the board call, called his buddy in Beirut …”) The SEC doesn’t have that but right now the tie goes to them.

Recently we’ve discussed evidence that the SEC’s well publicized focus on insider trading may or may not have actually reduced insider trading. That’s a hard question to answer because you don’t have to, like, publicly file a form every time you insider trade,2 so a rough proxy for it is to look at questions like:

  • did the target’s stock go up just before a merger was announced, and
  • was volume in the stock heavier than usual?

If more people are trading more shares at higher prices, it stands to reason that some of them know that something’s up. Maybe not! But probably.

That’s the academic/pseudo-academic view and it now seems to be the SEC’s view as well, part of a general embrace of efficient-markets thinking by the SEC that we’ve talked about before. There’s a certain aesthetic loss in that view: if you outperform the market, whether in one big lucky trade or in a pattern over time, the SEC will think it’s suspicious, and will investigate you.3 That’s a sort of humorless view of the trading world, which saddens me, but on the other hand it’s probably true.

Also I guess it’s good for the stats? One model you could imagine is something like:

  • Insider trading is illegally trading on insider information.
  • Insider trading is measured by existence of abnormal trading before merger announcements.
  • Investigations into insider trading are likely to deter insider trading.
  • Investigations into abnormal trading before merger announcements are likely to deter abnormal trading before merger announcements.
  • If you can have only one, which would you rather?

But by that standard, there’s a ways to go. The SEC’s complaint lists a bunch of other fishy trades in Onyx by the two frozen acccounts, some fishier than others. For instance, “The 50 July 90 calls purchased on June 28, 2013 represented an approximately 138% increase over the average daily volume.” Here, meanwhile, is trading in Onyx’s common stock. On Friday, June 28, the last trading day before the merger proposal was announced, the stock traded 1.9mm shares and closed at $86.82, up 1.9%, versus the S&P being down 0.4%. Those 1.9mm shares represented an 84% increase over the average daily volume in June, and a 114% increase over the average volume over the previous 10 trading days. An extra million-ish shares of Onyx traded the day before the merger was announced;4 the people buying those shares – which opened at $132.63 the next trading day – made an instant $40 million and change, ten times what the SEC’s shady options-trading targets made. Some people might find that pretty fishy.

SEC Freezes Assets Of Insider Traders in Onyx Pharmaceuticals [SEC, and complaint]

1. Also it’s a fun test of your probability theories and your belief in the law. Civil insider trading tends to be a preponderance-of-the-evidence type thing, meaning in theory that if it’s more likely than not that it was insider trading, the SEC should win. So good questions include:

  • (1) Is the bare fact that someone bought a lot of thinly traded call options a few days before a big positive announcement, with no other information, enough to make it more likely than not that he was insider trading?
  • (2) Does your answer change if you have two facts, viz. (a) the options and (b) the fact that he shows up in court and says “hi, those were my trades, I promise I was not insider trading, give me my money back”?
  • (3) What about (a) the options and (b) not showing up in court?

I go with probably insider trading on (1) and (3), probably not on (2), but I could see other interpretations.

2. Well Form 4 but you know what I mean.

3. This suggests that smart criminals should be fraudulently underperforming the market, Producers-style.

4. These guys bought a total of 1,119 contracts (111,900 shares) over that week, 320 of them on Friday, so dealers hedging their delta was only a very small part of that.

8 comments (hidden to protect delicate sensibilities)
Show all comments ↓

Comments (8)

  1. Posted by WHERE'S BESS? | July 8, 2013 at 11:53 AM

    Matt, did you make Bess drink too many margs over the long weekend?

  2. Posted by Guest | July 8, 2013 at 12:27 PM

    Matt you are a jerk. For the last time, I am not your grenade.

    -Shazar

  3. Posted by quant me maybe... | July 8, 2013 at 1:57 PM

    My takeaway:
    1) Buy random super cheap out-of-the-money calls for the next year.
    2) Come into some inside info next year (Note to self: remember to get to know the guy that knows the guy whose coke dealer has the good info).
    3) Purchase out of the money calls in target.
    4) Profit!

    Or

    Just do whatever Steve Cohen did to get away with it which I assume was:
    1) Control the setting — face to face talks only no e-mail, no IM, no phones and no parking lots.
    2) Control the conversation — I ask questions, my guy just answers and he does it in the office.
    3) Hire a bunch of guys who know someone and ignore their expenses in maintaining the mark.
    4) Buy a pickled shark because why else would anyone buy a pickled shark with a taxi driver's air freshener for a hat because that obviously did something to the SEC.

  4. Posted by anon88 | July 8, 2013 at 3:18 PM

    Roch Rogue was scheduled to be sentenced today……any info???

  5. Posted by Fluent Capital | July 8, 2013 at 6:15 PM

    I would like to say, "should have stopped after your first try", but even the first was terrible.

  6. Posted by Bayesian Quant | July 9, 2013 at 9:16 AM

    Here are some suggested analyses for someone who likes Bayesian statistics:

    What were the historical profits of all "suspiciously looking trades" (for example, purchases of out-of-the-money call options by foreigners that are "large" relative to the usual trading volume)? If the average profit is less than zero, that would suggest that this group of trades is largely not driven by insider information. If the average profit is greater than zero (check statistical significance here), some of them may be driven by insider information, some of them may not. It may be that additional evidence is necessary in order to increase the ratio of true to false positives. What's the prior probability that a "large" trade was done based on insider information? What's the posterior probability after considering additional factors/evidence? What ratio of true to false positives is desirable from a legal standpoint or a policy standpoint? 9:1? 1:1? 1:9?

    There are hundreds, maybe thousands of stocks with thinly traded options. It's possible that there are trades that appear "large" quite frequently. Is there any observable correlation between the frequency of "large" options trades and corporate events that cause big stock moves for "small" stocks with "thinly traded" options volume?

    Or some other analysis like that.

    Maybe there are some academics reading this (financial economists/lawyers or some other field?) who would be interested in pursuing such questions?

    The SEC seems to have (at least some) relevant data. They could publish the data (maybe in an anonymous form) to make them accessible to people interested in researching such questions.

  7. Posted by guest | July 9, 2013 at 12:47 PM

    This is pretty weak. Buying 80 or 175 contracts is barely out of bush league retail trader size. So the fact that it's "more" than previous trading volume is almost irrelevant; if you trade that option, you can't help but be over the ADV.

    Saying an option is thinly traded is almost redundant. Outside of SPX, GLD, RUT, NDX etc. and some couple dozen stocks/etfs, trading is nearly non-existent. Most single name options trade exactly 0 on a given day. Total open interest is often 0, or maybe a couple hundred contracts.

    These were small orders for at-the-money calls. Calling an 85 strike when the common is trading at 84.17 "out-of-the-money" is just wrong.

    Certainly trading out of Beruit seems weird, but this moves the presumption of guilt too far.

  8. Posted by Quant me maybe ... | July 9, 2013 at 8:24 PM

    On a serious note, I have to agree. We often encounter situations where we wish to hedge but can't because there is insufficient interest — totally blows the strategy. I look at a lot of equities and think that it's nothing but a test mechanism for a bunch of robots playing with a few creepy old guys day trading in their basements using the refresh button on google to get 'real time' data.