Ruminations on Rogue TradingO
Andy (mod note): 9/23/11 article from Damodaran - but the talk on Rogue
We are in the midst of a "rogue " scandal and the media loves it. It started with this report in the Wall Street Journal about an unnamed trader who had lost $ 2 billion for the Swiss banking giant, . The trader was quickly identified and named as Kewku Abodoli, a director at the Delta One desk (more on that later). Today's story has more details, with a comment from Abodoli's about how much he regrets his actions (or at least getting caught). If you have a sense of deja vu, it is because you have seen this story play out before. Just to refresh your memory, here are some memorable rogue traders from times past, with one being famous enough (Nick Leeson) that he had a movie (and not a very good one) made about him.
- Trading is addictive: Anyone who has traded knows that the process can be addictive, where trades lead to more trades, and at least for some people, there is no stopping the .
- The search for the "big" payoff: It is human nature to aspire for the pot of gold at the end of the rainbow, the mega million dollar prize on the lottery and the trifecta at the race track. In , that big payoff is (or at least seems) closer than in any other profession and many rogue traders start down their chosen path hoping to make the "big trade"...
- House money and Breakeven effects: In an earlier post on Jerome Kerviel, I noted that two well documented tendencies in behavioral finance: the tendency to take more risk than you should with other people's or house money and the proclivity to reckless risk taking, once you start losing money, to get back to break even. As you look across the rogue episodes, they all share this characteristic. These traders all started with small losses, which they tried to recover from with bigger bets, and the process kept escalating until you get to hundreds of millions of dollars.
Let's start with the first question. For a trader to lose hundreds of millions, he or she has to "lever" up and there are at least three ways this can be accomplished:
- Buy on borrowed money: You can borrow immense amounts of money on a small capital base, especially if you work at a large bank, and invest the amount in risky assets. You increase the upside on your equity investment, if you are right, but you magnify the downside, if you are wrong.
- Derivatives: You can make bets on derivatives that have potentially unlimited losses: this is the case when you buy or sell a futures contract on a commodity or a currency or when you sell options (either calls or puts).
- Long-short strategies: You can sell short on some risky assets, collect the proceeds and buy other risky assets, i.e., the hedge fund strategy. If the asset prices move in the right direction (your short sold assets have to drop in value while your long assets have to increase in value), you can make dramatic returns on small investments. However, if asset prices move in the wrong direction, your losses can be many times your equity investment.
Coming back to Kewku Abodoli, the rogue trader of the moment, it is worth noting that he traded on the "Delta One" desk. On the surface, Delta One desks are relatively placid and profitable places, where traders trade derivatives and exchange traded funds (ETFs) to take advantage of movements in the underlying assets (thein the desk name references the option , i.e., the percentage change in the option value for a unit change in the stock price). However, access to derivatives and ETFs can be a double edged sword, allowing a rogue trader to take very large risk exposures.
- Time lags in risk measurement/management systems: Given how quickly prices move in financial markets, there can be time lags in marking investments to market and learning about risk exposures. These problems are exacerbated with ETFs, since they are themselves often portfolios of traded assets which have to be marked to market.
- Systematic measurement error in risk management systems: All risk management systems are based upon risk measures that are estimates. Thus, Value at Risk ( ), a widely used risk measure at banks, has measurement error on many of its inputs, and some of these errors are systematic. For instance, a that is based upon the assumption that asset prices move in a normal distribution will understate the risks of assets whose prices are discontinuous and tend to jump. Traders that learn about these systematic errors can then exploit them to hide real risks in their portfolios.
- Avoidance and Denial: It is possible that those who monitor the rogue trader get a sense that something is wrong much earlier than the final denouement. However, in a very human response, the first response is to deny that a problem exists and avoid it, until it blows up.
- Hire the right traders: In my post on Jerome Kerviel, I pointed to the folly of entrusting to young men, a group that tends to take bigger and more reckless risks than any other subgroup of the population. I also suggested, only half in jest, that hire a few traders' mothers to trade alongside the traders, since older women are the perfect counterweight to young men in risk taking. I am sure that Mr. Abodoli would have been more cautious in his risk taking, if his mother had been sitting at the next desk.
- Real time and dynamic risk measurement systems: Risk management systems should track prices in real time and capture bad risk exposures early.
- Restrict in illiquid assets: Even the most sophisticated risk management investments have trouble dealing with assets that are illiquid, where the prices are appraised values and not transaction values. As investing choices widen from traded stocks and corporate bonds to ETFs, derivatives and mortgage backed securities, risk management systems have come under more strain.
- Simple and focused risk management systems: Since inputs into risk measurement and management systems have systematic measurement error that traders exploit, simpler risk management systems that have fewer bells and whistles are more difficult to game. In addition, can borrow a technique that El Al, the Israeli airline, has used for years to keep terrorists off their planes. Rather than spreading their resources wide and check every passenger, they profile passengers and focus on those most likely to create problems. Banks can adopt a similar practice: rather than have risk management systems that track every trade in the bank, they can identify those areas, where rules are most likely to be broken and focus attention on them.
- Stress testing: Every risk management system will fail. It is a question of when, not whether. The key to good risk management is how you respond to failures in the system, not successes. Rather than assume that everyone is playing by the rules and measuring the consequences, it makes more sense to assume that some will not play by the rules and prepare for the consequences.
- Look at outcome and process, not just outcome: I started this post by noting that rogue traders who make millions are feted and celebrated. As long as we continue to do that, we will incite traders to take unconscionable risks. The best way to bring home the point that you will not put up with rogue is to fire a rogue trader who makes millions and to deny him his bonus.
- Top managers at banks: The six suggestions that I have above are all directed at management, but they will reduce, not eliminate, the likelihood of rogue . Top managers at banks have to consider rogue to be one of the risks that comes with proprietary . When allocating capital to different businesses (corporate banking, investment banking, proprietary ) should incorporate this risk. (Proprietary will have to make a higher return on the equity invested in it to break even than commercial banking...)
- Investors in these banks: By the same token, investors in banks have to be cognizant of the risks that come with proprietary . A bank that generates a higher proportion of its profits from proprietary is riskier, other things held equal, than a bank that generates its revenues from traditional banking. If these banks trade at the same multiple of earnings, I would pick the latter over the former. In practical terms, I am suggesting that when screening for bargains among banks, we look at the percentage of profits from proprietary as a risk measure.
- Regulatory authorities: If rogue Volcker rule in the US and the new banking rules that are being discussed in the UK, both of which would bar commercial banks from proprietary . I agree. is part and parcel of proprietary , then it follows that institutions where the government provides a backstop should not be allowed to indulge in it. This is the basis for the