How Exactly the SEC Detects Illegal Insider Trading?
From using artificial intelligence techniques to accesing your Google search history
In this post I’m going to show you how the Securities and Exchange Commission (a.k.a the SEC) detects illegal insider trading activity. If you don’t know what insider trading is or you want to refresh the basics, I suggest you read this post first.
The topic is particularly interesting since this is an extraordinarily difficult crime to prove. The act of purchasing and selling shares is perfectly legal, it is only what is in the mind of the trader that can make this legal activity a prohibited act of insider trading.
Moreover, most of the insider trading cases involve no direct or physical evidence (except if there’s a confession by the perpetrator) so the whole investigation is a matter of putting together pieces of a puzzle (trading activity, phone calls, emails, physical meetings…etc.).
To illustrate how difficult this is, consider this recent case of shadow trading (using material information of your own company to trade in a different company): the SEC investigated a pharmaceutical executive who purchased options of a competitor stock after learning that his own company was acquired by Pfizer.
Note that when an acquisition is announced to the market, the stock price of the firm being acquired goes up (almost reaching the acquisition price) but the shares of other potential targets usually go up too.
The tricky thing here is that technically the corporate executive did not have any confidential information of the competitor, which is the subject of the material event, and either wasn’t trading shares of his own company, which is the action that the current insider trading laws are design for. On top of that, you have to prove the fraudulent “intent” of the trade, which is like trying to prove a thought.
Now that we agree that insider trading cases are pretty complex, we can define two general categories of sources of cases for the SEC:
- Informants
- Market Surveillance
Let’s see both in detail!
Informants
Whistleblowing
One of the first sources of information is the whistleblowing program: the SEC may award a bounty to any person who provides information leading to a civil penalty in a case of insider trading and/or tipping.
The total amount of the bounty cannot exceed 10% of the penalty, and for the bounty to be issued this information must lead to the successful enforcement of covered judicial or administrative actions.
For instance, in 2010 the SEC awarded $1 million to a couple who provided information and key documents regarding an investigation of an insider trading case involving a ex-Microsoft employee and the CEO of a hedge fund. Apparently the ex-employee, who was hired by the hedge fund, was being tipped with material information by a former colleague at Microsoft.
The Microsoft ex-employee happened to be the ex-husband of one member of the awarded couple, so remember to end your relationship on good terms just in case.
(Angry) Market Participants
A common and profitable way to implement your illegal insider trade is to use short-dated out-of-the-money (OTM) options.
I don’t want to get too technical here, for simplicity let’s say that these are financial derivatives which can multiply your investment by a lot in a short timeframe if you are very lucky or, like in this case, you know for a fact what is going to happen.
Now imagine that you are on the other side of the trade (because you believe that the market is fair and market participants are ethical people, lol) and lose a ridiculous amount of money because an almost impossible event has happened.
The result is that some of these angry (and now poor) traders will contact the SEC showing them this suspicious transaction to see whether they have been played.
(Angry) (ex-) Employees or Competitors
This one does not need much explanation: if any of these categories somehow gets information of the wrongdoings of any company officers, they can use it in exchange for a reward (e.g. an angry ex-employee joining the whistleblowing program), for pure revenge (e.g. you hate your boss) or just to hurt a competitor’s business or stock price.
Market Surveillance
Bluesheet data and ARTEMIS
The first thing to note is that the market surveillance activity is not carried out by a single divison but is usually the result of a team effort of several groups within the SEC, such as the DERA (Division of Economic and Risk Analysis), the OMI (Office of Market Intelligence, within the Enforcement Division), the CRQA (Center for Risk and Quantitative Analysis) or the MAU (Market Abuse Unit, which is probably the most relevant for insider trading cases), to name a few. Leads on new insider trading cases can come from any of those divisons too.
The bulk of the information that these units analyse comes from the so-called “bluesheet data” (basically trading and customer transaction data) that the broker-dealers, market makers and clearinghouses are required to maintain and provide electronically to the SEC. For instance, the SEC asked the (in)famous broker Robinhood for this data during the GameStop frenzy of last year, among other things.
As you can imagine, all this information means millions of datapoints generated for each trading day, which means that your regular Microsoft Excel is not enough to do a proper data analysis.
For that reason, the MAU has its own big-data analytics tool called the Advanced Relational Trading Enforcement Metrics Investigation System (ARTEMIS), which focuses on the analysis of suspicious trading patterns and relationships among multiple traders and combines “about 10 billion equity and options trade records from SEC and FINRA and uses advanced analytics, created by Division staff, to rank trades bases on different metrics”. The metrics of course are not public.
For example, this tool was used to uncover a insider trading scheme set by three Netflix’s engineers:
“The three Netflix software engineers traded based on changes, especially growth, in the Netflix subscriber base. They used this information BEFORE it was disclosed in periodic earnings announcements. After the lead engineer left Netflix in 2017, he continued to obtain confidential subscriber information from a friend and former colleague. Another former colleague tipped the lead wrongdoer concerning the July 2019 earnings announcement. The engineers used encrypted messages and paid cash kickbacks in an attempt to avoid detection, but the MAU identified the pattern of “improbably successful” trading in Netflix stock.”
SROs & SONAR
Another source of information for the SEC are the SROs (Self Regulatory Organizations, a.k.a the exchanges) which have primary responsibility for surveillance of their markets and provide the SEC with hundreds of reports of suspicious trading each year.
This surveillance mainly involves monitoring for unusual volume and price action, especially before a market-moving event (companies are required to provide the relevant press-release in advance of releasing it to the general public).
One of the most famous SROs is the NASDAQ, which has its own big-data tool called the Securities Observation, News Analysis, and Regulation (SONAR) system.
SONAR is used to monitor NASDAQ exchanges, “processing approximately 10,000 news wires stories and SEC filings, evaluating price/volume models for 25,000 securities, and generating 50–60 alerts (or “breaks”) per day for review by several groups of regulatory analysts and investigators”.
It also makes use of several AI and statistical techniques, “including NLP text mining, statistical regression, rule-based inference, uncertainty, and fuzzy matching”, which greatly improves the quality of the alerts generated.
The funny thing is that these cited features are from a paper dated in 2003, when the artificial intelligence field was not as widespread as it is today. SONAR is operating since 2001, I can’t imagine how sophisticated the system is now.
Conclusion
As we have seen, the lead for an illegal insider trade case can come from an human source or from a very sophisticated quantitative analysis of -mainly-market and customer transaction data.
The technological capabilities of the SEC are much more advanced than people think, and their quantitative techniques and statiscal tools are way more complex that those used at many fancy hedge funds.
That being said, catching illegal insider trading schemes is no easy task if the insider really knows what is doing and keeps a low profile. Even if the SEC is able to detect the trading anomalies, they still have to prove the “intent” before a court of justice.
In another post we will talk about the SEC investigation process and how they have developed new techniques to gather evidence: from traditional methods such as witness interviews, to using wiretaps or accesing your Google search history (only to find that you literally searched “how sec detect unusual trade” and “insider trading with international account” days before effectively committing illegal insider trading).
Please clap the post if you liked it and follow me in Medium and Twitter for more content on insider trading and investing!
Thank you for reading!
