In our business you could get the impression that the idea of individual talent is an Anglo-American pathology. As a metaphor for how talent is deployed in an investment bank let’s take a pre-Volcker prop desk. A bunch of traders run different strategies whose overall returns are balanced by the strategies’ lack of correlation. While the traders share infrastructure, their intellectual efforts are deliberately kept separate. The returns – and compensation – that the traders individually enjoy are highly variable but the variance of returns for the entire desk is stabilised by the number of independent contributors and their (hopefully) low covariance.
You can see the logic in this but also you can see the risks and limitations.
A friend of mine, who is the most able mathematician I know outside of academia, provides a good contrast. He heads R&D at one of the large reinsurers. A chunk of his working life is spent authoring high-quality papers on actuarial mathematics. In addition, his abilities are leveraged to prepare pitches to clients and then to meet with them to win business, and in management – directing the work of other researchers. His expertise is meshed into the institution in a much deeper way than that of the traders on our imagined prop desk.
Whereas the models of our prop traders were kept separate as a policy, there are often more mundane reasons why the best ideas aren’t shared. For example, the work of another talented mathematician friend of mine is going unutilised, despite demonstrable commercial value, because his relevant colleagues don’t really get it, and he has an anti-talent for marketing. At least this is only wasteful, not dangerous. Calamities have famously occurred when traders and quants respect and use but fail to understand the insights of the other discipline: combining as a mixture rather than a compound, to use an analogy from Chemistry. If a trader doesn’t know where a model breaks down or a quant doesn’t appreciate the real-world subtleties of products and markets it can get expensive. Similarly, managers can only manage what they can understand. In the crisis, the quant/traders who were packaging up CDO-squareds and CDO-cubeds were every bit as sharp as they needed to be but who could competently supervise them? Talent on the trading floor can be a machete in a crowd.
Bringing together a set of talented individuals does not make a talented organisation. To mesh together in an organisation talented people need to communicate in a shared language. The lexicon of risk management should be a part of this. To an extent, this happens: for example, traders have to live within risk limits set at a departmental and group level. However, the intraday risk measures that are often more vividly felt by the trader are typically different. These desk-level measures that are tailored to the way that a particular book trades are in one language while the risk measures of record are in another. In the best case, there is a process of aggregation akin to a formal translation between these languages. More often, the two languages just co-exist side-by-side. Again, this is dangerous as the talents of the trader and the talents of the risk manager abut each other with insufficiently deep engagement.
This danger clearly motivates the BIS paper, Principles for effective risk data aggregation and risk reporting, which mandates yet more targets for the big banks in an effort to reduce the odds of another systematic crisis. While the 14 principles given in the paper are important, I suspect that it misses the mark. A bank could, as I read it, conform in full to every one of these principles yet still have desk risk as managed by the traders run on spreadsheets with middle office staff generating the risk reports of record in wholly separate big systems. Sure, the traders would have to stick to all of the risk limits coming from the official system and bank management would have to attest that these limits cover risks in all material dimensions. But to the extent that the traders and the risk managers speak in different languages the total talent directed coherently at risk management is much less than the sum of the parts, and the scope for unforeseen problems remains correspondingly much higher.
According to the Financial Times, this is one reason why banks are keen to see more trading go electronic. I also suspect that it has been a background motive for some of the periodic attempts to migrate aspects of trading away from New York and London. For example, a few years ago Credit Suisse, following UBS, moved its main FX trading floor to Zurich. There are all sorts of rational reasons for a Swiss bank to want to do this but I believe that the sense that the Swiss culture might be more collaborative, organised and prudential was a part of it. As it transpired, the benefits of being in Zurich were not strong enough to fight the reality of FX markets being centred on London and the move was subsequently reversed. Maybe it was just ahead of its time. Like talented people, good ideas need friends.
The futility of lone talent holds equally in technology. There is no point having a fantastic trading system if you can never get the trades you want done because of slow infrastructure or if your trade processing systems can’t reliably handle your volumes. To be excellent at anything you have to be excellent at everything; or more accurately, the set of supporting competences you need to achieve a target competence is often far wider than you think.
One response to this reality is that of Henry Ford who (as reported by Jared Diamond) sent out staff to scour junkyards to learn which parts of the Model T didn’t routinely wear out during the life of the car so that he could spend less money making them. eCo also aims to make the supporting competences less expensive but in a different way. We try to help talent to find friends and to help them mesh effectively. In part this is achieved through measuring the suitability of software for collaborative use and socialising the user experience of software and support quality. Based on this, the talent available to an organisation can be more easily enhanced by contributions from the wider community.