In Star Trek: The Next Generation, when Captain Jean-Luc Picard says "make it so", 'it' usually happens. If life were that simple, regulators around the world could tell capital markets participants to get a grip on trading risks. Regulators could utter the phrase "make it so" and traders, risk and compliance officers would jump to ensure that trading exposure and algorithms were continuously monitored, with no threat of rogue trading, out-of-control volatility or nasty flash crashes. This would reassure investors that, despite all the bad press from 2008 to the present day, banks now have everything under control.
Enter UBS stage left and the banking industry takes a sudden gasp of breath as another intergalactic crisis unfolds. Is it the Romulans or the Borg? Worse in fact -- it's a rogue trader! According to Reuters, a trader at UBS lost the Swiss bank around $2 billion through unauthorized trades. At the same time, police in London arrested a trader from UBS' ironically named Delta One trading desk; Delta One desks focus on hedging for customers, and are by nature relatively low risk.
This is not a first of course. Other high-profile rogues were caught and prosecuted. Most recently in January 2008 rogue trader Jerome Kerviel (like his predecessor Nick Leeson of Barings Bank fame) woke the world up to the fact that trading without proper risk controls and supervisory oversight can lead to unintended consequences. In that case the consequences were massive losses (approximately $6 billion) for Société Générale, and all because Kerviel had knowledge of and abused risk management systems at the bank to hide losses.
In UBS' case it transpires that the alleged culprit, 31-year-old Kweku Adoboli, was a trade support analyst prior to his becoming a trader. Trade support analysts are in charge of looking after the bank's electronic trading and post-trade platforms. Deja vu all over again...
It is not only rogues and scoundrels that banks need to be aware of. Trading positions that push firms beyond a safe risk threshold are a classic problem. Trading algorithms can go out of control, like one of Deutsche Bank's in Japan that went wild in June 2010, going into an infinite loop and taking out a $183 billion stock position. Intentional market manipulation -- by traders or algorithms -- is also an issue.
Detection of criminal fraud or market abuse is something that must happen in real-time before any unauthorized trading has a chance to lose a firm money or move the market. There are a number of best practices that can be used to mitigate these possible problems. In the course of writing my forthcoming book "Business Attention Deficit (BAD)" I created a prescription; it's a recipe that empowers trading firms, brokers and regulators, to keep up with high speed markets and detect or predict problems in real-time. This enables risk and surveillance groups to be responsive and take mitigating action - before problems get out of control. Below I've taken an excerpt for how to See, Respond and Improve to risk in real-time. 'See' is all about plugging in to market events, as well as portfolio data; 'Respond' is all about continuous intelligence and detecting pre-trade risk threats; 'Improve' is about learning as you go along and refining the way the firm runs.
Responsive Business Prescription for Real-time Risk*Step 1: See!
- Monitor all real-time quotes and trades for the markets on which you are trading.
- Have easy access to your portfolio data.
- Monitor events about the behavior of your trading algorithms, such as what state they are in, trades that are placed etc.
- Continuously recalculate analytics, such as portfolio value and current exposure.
- Monitor for patterns that indicate portfolio risk exposure thresholds have been breached or will be breached if certain trades are made. If detected, block subsequent trades and alert the chief risk officer (CRO) and trading desks.
- Within a brokerage house, monitor for patterns that indicate an institutional client has breached risk limits. Block subsequent trades and alert the CRO and the client.
- Monitor for patterns that indicate a trading algorithm has gone wild, e.g., continuous orders in close proximity outside of its specification indicate it may have entered an infinite loop. In this circumstance, alert the CRO, block the algorithm's trades or, if possible, place the algorithm in quarantine.
- Investigate circumstances that have historically caused risk spikes and set up rules to identify early warning signs or predict upcoming breaches.
- Ensure all trading algorithms are tested in a wide variety of realistic situations before going live.
- Expand policies for back-testing and simulating algorithms to include all circumstances in which algorithms have gone wild.
- Collaborate: A CRO can communicate with institutional customers or traders to manage a potentially risky situation.
- Social Media: Monitor global market sentiment to anticipate potential market spikes that could affect risk exposure.
- Mobility: Make a risk dashboard available on mobile devices, so the CRO gets early warning regardless of location.
We don't have to wait until the 24th century to arm trading firms. The technology to police in real time is here and with the right forward-looking prescription we can proactively address problems before they become a crisis.
By taking the above steps, we can begin to see risks lowered and confidence in the financial markets restored. Make it so.
*Excerpted from Dr. Bates' upcoming book Business Attention Deficit: How to plug in and profit in the face of constant crisis.