If there are lessons to be learned from the recent foreign exchange markets scandal, then the need to monitor the behavior of traders must be on the top of the list.
As Swiss and U.K. regulators dig deeper into the manipulation of FX benchmarks, one of the things they are checking is whether banks have reduced the risk of trader misbehavior since the scandal was revealed.
I believe they will find that most banks have done very little to monitor and prevent dishonest traders from behaving the way they do.
The fact is, fraud and market manipulation are a form of antisocial behavior. In the U.K. when a person engages in antisocial behavior the courts can issue an order, known as the Antisocial Behavior Order (ASBO), which tells the person not to do it again. An ASBO can be issued in response to "conduct which caused or was likely to cause harm, harassment, alarm or distress" and is deemed necessary to protect relevant persons from further anti-social acts by the defendant.
The knock-on effect of the FX fixing manipulation is huge and the harm caused is both monetary and reputational. The 4pm WM/Reuters exchange rate benchmarks are used to set the value of trillions of dollars of investments, to calculate global bond and equity indexes, and as benchmarks for settling derivatives deals.
The alleged manipulation of FX prices during the one-minute 4pm fix period was done in dealer chat rooms, where traders openly make bids and offers and disclose deals. Their electronic fingerprints were everywhere; there were no secret meetings behind closed doors, or private-club handshakes. Any antisocial behavior on the traders' parts could easily have been detected given the right tools.
A pick-up in volumes, volatility and temporary spikes in currency prices had been cited as evidence of improper trading practices. Yet no seemed to be monitoring the patterns in the marketplace. Had anyone bothered they would have noticed the unusual flurry of activity before the 4pm fix. They may have spotted evidence of front-running or "banging the close" within the one-minute fixing period.
They could have seen that the messaging networks were busier than usual in the run-up to 4pm, and that certain banks appeared to be trading with their buddies at certain other banks a lot.
Following this scandal, it is clear that monitoring has to go much further than trades and data. The behavior of traders themselves must come under scrutiny. What time do they get in? When do they leave? Has this changed of late? Do they take any holidays? (Traders who do not take their holidays could be hiding something in their P&L's. Most banks insist that their traders take their vacations each year, so they can give the books a good going over.) Are they trading more than usual? Did their trades boost the market?
The technology now exists for self-learning monitoring systems that can detect potentially anti-social trader behaviour. This includes looking for known patterns of abusive behaviour, as well as patterns of behaviour that deviate by a significant degree from "the norm". Both types of pattern (or a combination) may indicate something for the compliance department to look into. An example is an unusually large trade in an instrument not usually traded by the trader, just before a news article moves the market by a significant amount. Could this be an insider trade? Was it preceded by messaging traffic between two parties that may have shared inside data illicitly?
Some forward-looking institutions have already deployed such monitoring systems, which can alert compliance managers, who can then take action soon after an occurrence or even while the issue is occurring. It is thus possible to be proactive, rather than reactive after discovering an incident when it is too late - days, weeks or months later. And the compliance department can then slap an ASBO on the offenders!