[Skip to content]

Sign up for our daily newsletter
The Actuary The magazine of the Institute & Faculty of Actuaries

Book review: Lecturing Birds on Flying

In this book, Pablo Triana discusses what he sees as the primary cause of recent financial crises: over-reliance on models that do not work. He proposes that a widespread problem exists, that problem being the quantification of finance and the subsequent abandonment of ’rules of thumb’.

There are many strands to Triana’s argument. He makes many good points, but these do not always justify the conclusions that he reaches. For example, he asserts that finance does not have immutable laws in the same way as, say, physics. This is true – but it does not automatically follow that it is pointless to try and determine patterns in data, as he implies.

Triana also takes issue with many of the theories developed in the finance departments of business schools, which depend on assumptions that are simply not valid in the real world. I have sympathy this point of view, and I do question the usefulness of much of the research generated along these lines. However, my main objection is that this type or research, no matter how well peer-reviewed, is often pointless; Triana’s issue, on the other hand, is that it leads to legions of business school graduates applying these theories in the real world and causing financial mayhem.

There is some evidence that simplified models can escape from the university campus into the real world. The initial pricing of collateralised debt obligation (CDO) tranches relied on assumptions that can best be described as heroic. However, there were factors other than quantification and model naïvity to blame here. Not least, there were significant agency issues around pricing and rating CDO tranches, with little financial incentive for those involved to try to analyse the risks and rewards more accurately. More importantly, though, Triana later asserts that CDO traders were not actually using quantitative models at all.

A similar point is made later in relation to the Black-Scholes model. The flaws in this model are explored in great detail, perhaps leading us to believe that it has been to blame for many financial woes. However, we are later told that in the real world no-one uses it for trading. So why do the flaws in the Black-Scholes model matter?

As a final salvo against quantification, we are told that investment banks’ requirements for doctorates in quantitative disciplines are used only as barriers to entry for some segments of investment banking, and as ways of impressing clients. Once the doctors are in, they trade off-the-cuff, not using unrealistic models. Given that the starting point of the book is that finance academics are teaching unrealistic formulae to students who then use their knowledge to plunge the economy into chaos, this last argument appears somewhat self-defeating.

This is not to say that the Triana does not make plenty of good points. Excessive reliance on models was, and still is, a problem – but it is not the only one. Triana makes the important observation that a risk model that does not work in times of financial stress is not particularly helpful. Overall, however, there is too much hyperbole and too great a willingness to blame everything on a narrow range of issues. Bad models are clearly bad, but only if they are used in practice. The lack of understanding by executives of good models, remuneration structures encouraging traders to take excessive risks or to follow the herd, regulatory failure and many other factors are at least as important. Recognising this would have made the arguments in this book more convincing.

Paul Sweeting is professor of actuarial science at the University of Kent