Greater use of judgement

Long ago, bands of English artisans, Luddites, rioted against mechanization and destroyed machinery. Clearly, they were opposed to the increased industrialization and the use of new technology. Perhaps, in a way, we should emulate them.

Some years back, a few professional stock traders began to deal with the aid of theoretical prices derived from a model. Others derided that, saying that one should “trade like a man,” and stop relying on mechanical models. Nevertheless, many persisted in trading such things as stock and bond derivatives, calculating the value of the items traded, thanks to models.

More recently, those who are dependent on mechanical models for financial transactions remain front and centre. Someone accused those “technophiles” of killing the real financial markets.

More and more critics of risk modelling are speaking out, claiming that those who use models as the basis of financial transactions are “worse than nothing,” taking a dangerous course of action.

What happened in the last few years was that those economic models went awry when mortgage-backed securities were further packaged in debt obligations. Heretofore with traditional products, such as corporate debt, rating agencies used basic credit analysis and personal judgement. Other financial items became so complicated that especially designed models were used, but soon they developed all kinds of flaws.

The problem was worsened by credit agencies that issued unjustifiably favourable ratings to reward the issuers who paid them. 

Furthermore, it turns out that the big use of models actually changed the markets that they were supposed to map, thus undermining the value or their own predictions.  Too, many others also were using the same risk models, thus confirming the forecasts that were so flawed. In addition, government regulators based some of their actions on those flawed models.

All that points to the need for greater use of judgement and less reliance on models’ projections. Obviously, it is unfair to condemn all models.

Models, together with their users, are only partly at fault. In this era, modelling is not going away. Number crunchers are working on new ways to ensure that models become more helpful.

The way forward, therefore, is not to reject high-tech finance, but to be more realistic about its limitations. Those who are thoroughly devoted to technical interests should not be cast aside, but they should be coupled with more everyday, sensible assistance.

A professor at Columbia University summarized that very well when he said, “You may need a plumber, but you get now a professor of fluid dynamics.”

In other words, let us not throw technology overboard, but we should rely more on human judgement, and less on obscure technology.


Bruce Whitestone