Not long ago, there was a debate in professional circles about the predictive value of central bank forecasts. They have been desperately inaccurate in the last few years, a time of severe price instability. The question has therefore arisen as to whether the models underlying these forecasts are sufficiently robust and realistic.
Chimera of Realistic Models
The task of science is to investigate and then attempt to provide explanations for phenomena observable in the real world. And this applies not only to the very advanced fields of natural sciences but to every field of study, not excluding economics. No human being, however intelligent, is capable of understanding such phenomena in their full complexity – of taking all aspects of them into account at the same time. And therefore we have no choice but to create imaginary simplified representations of this endlessly complex reality – to construct models.
The essence of models is that they take into account only a limited number of aspects of the phenomena under study and deliberately ignore the others. It follows that they necessarily simplify reality. And this in turn means that they are by definition unrealistic. The only absolutely realistic model would have to be a perfect reflection of reality, in other words, reality itself. But this would bring us to the initial problem, which is that the human brain is unable to grasp reality in its integral form. The realistic model is therefore a mere chimera.
Central Bank Models And Forecasts
The above description of the role and pitfalls of models in science can now be applied to the macroeconomic models used by central banks around the world in their decision-making. On the basis of these models, monetary authorities subsequently produce forecasts of selected macroeconomic variables.
To some extent, central bank forecasts are similar to weather forecasts. Even these are far from accurate. But it is usually better to know it than to ignore it completely. Meanwhile, both weather forecasts and central bank forecasts are becoming more and more accurate and reliable thanks to our improved technical knowledge.
However, they should be taken with a pinch of salt, as they are now still some way from perfect reliability. And they will be for a long time to come. But it would be unfair to blame the central bank for using its macroeconomic forecasts in its decision-making, which in retrospect can sometimes seem out of touch with reality. The fact that – in the words of Czech National Bank (CNB) Deputy Governor Jan Frait – these models “represent a highly imperfect description of the economy” should therefore come as no great surprise.
The central bank, however, is not open to the possibility of having absolutely precise and error-free models. The question is therefore not whether the central bank has perfect forecasts or not. As CNB board member Tomas Holub argues,
“The relevant question is whether monetary policy that can rely on such analyses and forecasts is any better than a purely backward-looking policy, and I think this is simply true in the practice of the Czech National Bank.”
Although the modeling apparatus of central banks is constantly improving, it is still far from flawless. It is unfortunately almost by their very nature that they are often grossly inaccurate, especially in times of unprecedented shocks or structural change (when their reliability is essentially needed most). But as Tomas Holub says: “We have nothing better at the moment“.
Written by Stepan Drabek, analyst at CETA.