Friday, May 16, 2008

What has happened to Gillian Tett?

(Naked Capitalism) A year ago, I found Gillian Tett, then the Financial Times' capital markets editor, to be the single most useful financial reporter by a considerable margin. She gave insights into areas that were important but badly neglected elsewhere, such as CDOs, credit default swaps, SIVs, all well before they entered the mainstream lexicon.

She was promoted. While she may add value behind the scenes, her stories this year are a shadow of her former work. And that's being polite.

Consider her offering du jour, "How talking can help cut the risk of a lemming fall." Here's the set-up:
Imagine for a moment that you are a banker, who stumbles across a juicy new instrument called the "lemming" product that your sales team could sell to retail clients - for a fat profit......even though it has been rubber-stamped by your compliance department....investors will suffer big losses if stock markets fall more than 30 per cent.

....over 70 per cent of the audience [in a Securities and Investments Institute conference] voted to block the lemming deal in an anonymous poll, taken after the participants had discussed the issue with neighbours.

But then the organisers presented a chart which highlighted a more sobering point: when the SII has done these tests before, it has typically found that the proportion of bankers who block risky trades falls dramatically when participants do not discuss the issue with their neighbour first - even if they vote anonymously.

Tett uses this example to conclude that what investment banks need isn't better incentives, but (to use that horrid American term) more dialoguing:
....bankers should be forced to talk about their business with a wide pool of colleagues, including those outside their immediate silo, rather than just their bosses alone.

Rubbish. A conference is such an artificial setting that to generalize its findings to the day-to-day operations of a company is fantasy. People want to look good before their peers, and in a weird bit of self-deception, once someone takes a position publicly, they typically find it difficult to recant privately. And here, the tradeoff has been framed in uncharacteristically black and white terms: big profits versus big downside to clients in relatively low-odds situations. Would the response had been different if the question has included: "the odds of the market falling 30% in the next X year is Y"? Yes. Survey results are HIGHLY influenced by the wording of the question. So just imagine how susceptible real world situations are to subtle and shifting pressures.

Take the lemming. The response of a manager/department head in the real world no doubt will also be shaped by:
How tough standard disclosure language would be

What leadership in the lemming might do for league table rankings

How close your team is to being on track for its targets for the year

Whether your boss is satisfied with you these days

Whether your firm has had a major compliance/litigation problem in the last two years

Whether you sell to retail clients directly (ie, you own them) or primarily via other firms' salesforces

Whether other firms are selling lemmings actively. This is probably the biggest single consideration. There is far less perceived risk if others are already in the pool

Tett also argues that Goldman, an example of better practice, engages in just this sort of debate:
Institutions such as Goldman Sachs, for example, try to ensure that different business silos have ways of watching what each other does. They also invest heavily in creating a holistic risk management culture: Goldman Sachs' risk systems, for example, are run by Gerry Corrigan, the former New York Federal Reserve president, who makes a virtue out of sticking his nose into as many dark corners as possible - and trying to encourage companywide debate.

Tett has the causality backwards. Goldman still carries the legacy (now weakened since it went public and is dominated by the trading side) of being extraordinarily risk averse and image conscious. The firm when it was a partnership went to unusual lengths to make sure that even very junior staff understood the finer points of legal and practical liability. This was pragmatism; the partners regarded it as a cheap form of insurance.

The firm in the 1980s was also cautious about delegating decision-making authority to client facing staff that other firms delegated routinely (such as not letting investment bankers quote indicative prices for financings) and about putting capital at risk (it was late to book interest rate swaps). When its practices became visibly uncompetitive, it generally came up with a solution, or at least a finesse. Goldman is an inwardly-focused firm that takes few mid-career hires, so its culture is reasonably intact. The structure is a product of the culture, not vice versa.

Something this misleading from a formerly keen observer is surprising and disheartening. She needs to get back out and mix it up with her source more often.

No comments: