top of page

Five key take away points from the SEC’s approach to surveillance

Updated: Oct 24, 2022


It’s no secret the SEC invests vast sums into its surveillance capabilities. There seems to have been many lessons learnt post 2008/9, some clever people hired, more data gathered and in 2010 a whopping $50m fund allocated for multiyear projects*.

So what can we learn from their approach to catching abusive behaviour on Wall Street and importantly what can implement at a more local level?

An article in the FT published in 2014 gave a useful insight of the key measures being employed by the SEC and some wins (and potential wins) that came from it. In two short years the world of reg tech has moved on at rapid pace with many firms now (for example) incorporating various machine learning methodologies into their surveillance and fraud analysis. Are there still some important indicators to consider outside of trade analysis in isolation? The SEC seems to think so and so do I.


1. Follow the lines of communication


The SEC employed a tactic called parallel trading to uncover a $37m insider trading scheme. This involved focussing on the traders themselves rather than the movement of stocks. In some cases they scoured phone books and bank records once initial data analysis was carried out to make the links. There are two take away points here. One I will come back to, but clearly there is emphasis placed on knowing who is talking to and cooperating with whom. The regulator clearly has the benefit of seeing all trades across all firms. Something financial services groups cannot do. However you can start to look at networks in emails, phone records, participants in corporate transactions to look at pre and post transaction collusion.

This transformed the way the SEC was looking and their investigations and

“By flipping the investigative focus to people instead of stocks that “opened up the possibility of making connections between traders across multiple securities””.

If during your trade analysis you notice fund managers x, y and z always buying into a coincidental long position just before a new strong buy recommendation goes out, isn’t that a useful network to watch for other abusive practices?


2. Build up a picture of events – before, during and after – enrich with data you have mastered.


“If the hedge fund manager bought at $10 and sold at $9.50, it may not stand out to the human eye as suspicious. But with the new system the SEC will be able to see that the stock later fell $5, giving the hedge fund manager who anticipated the negative news a chance to avoid a big loss.”


Build up the picture. Set the scene and know the state of play. Enrich your data with as much as you know about what has and is going on. There is almost no excuse for not doing this. If you are running a deal or putting out research, you should have a lot of the data you need. Enrich with market data for depth and you can start to paint a picture.

How much data you put into this will of course depend on which lines of business you are involved in. As a base line you could look across one asset class at a time and start to incorporate obvious red flag indicators, e.g. CDS or shorting.


3. Look out for coincidental events around events


This is where ‘traditional’ trade surveillance comes into play. Are there any coincidental patterns around the corporate deal? Suspect trading? Position movements? Suspect hedging or PA dealing? This should all become more apparent once you’ve started to build the picture as in point (2). Rely on alerts as indicators and don’t not dig deeper if something doesn't smell right. Depending on the system and set up you use, it might return a very granular view. Take a step back and look for bigger trends. Does your monitoring team incorporate any trade surveillance in their review? It might yield suspect trends if you look across other similar corporate transactions.


4. Bring in multiple data sources


Another quote from the FT:


“Mr. Ceresney [director of enforcement at the SEC] says it has identified links by accessing multiple data sources. For example, it can match stock trades made one day with money transfers executed three days later. “It’s not stuff we couldn’t have done before. It would have taken weeks and now it takes minutes””.


I touched on mastering your data in an earlier blog post and cannot emphasize it enough

again. Know Your Data and don’t not experiment to see if there are any correlations in there. It may not be causation, but it may point to something else. Ask intelligent questions. Do you suspect someone is wash trading? Look at their PnL. Isn’t it suspect that someone with such high volumes is generating such thin profits? A compliance team will no doubt approve PA deals. So why not bring that data in. G&E? I’ve found this to be very useful in generating analytics around PA dealing and MNPI (including large orders). Add this to the visualisation in point (2) to create rich detail of potential cracks in the Chinese Wall.


5. It’s not all down to the machines (yet).


Some at the SEC cautioned that new tech will not answer all of their problems and until the data is where is in the state it needs to be. I think there is still an argument for some good old fashioned detective work. Sometimes you might pick up on something which doesn’t feel right via alerts or near misses and need to dig deeper and elsewhere. It could be your data might not be to the standard you require. It might be that it’s all still a manual process at your firm. The advantage of a carrying any kind of manual investigative work is that it starts to help you understand which data items are useful and if you were to one day design an automated system, which data feeds you might need. Are corrupt traders always going to try to abuse the market in the same way? A degree of agility and depth of understanding will always be required in analysing the output.

In conclusion although analysing trades is key in any surveillance exercise, there is clearly value in looking beyond this and seeking patterns elsewhere. Ultimately knowing your business and data come into play and, it feels to me, those who master the art of bringing the two together will have the advantage. Over time as methods of executing business evolves there will be a clear need to adapt surveillance techniques accordingly. While rules based tests may work today, they may rapidly become redundant as the banking world wakes up to trend analysis.


Final note from a former SEC chairman:


“Technology really requires concerted, consistent spending to make sure systems stay up to date as well as developing new tools when they’re needed”


Keep that project budget topped up!

*Source: The Financial Times

All quotes used in this post are from an article which appeared in the Financial Times titled: “SEC: With the program” on 08 May 2014 by Kara Scannell.


10 views0 comments
bottom of page