Unhappy with 23 years of uninterrupted economic growth, the Reserve Bank has conducted a review which is likely to recommend taking humans out of their forecasting and replacing them with equations.
“Models”, they call it, but it’s equations nonetheless.
That’s the message in an AFR article this morning which says that only a few senior officials have seen the report but that it has “raised concerns that Reserve Bank forecasters put too much store in ‘gut feel’ judgments over hard models”.
I have to declare an interest here. I consider myself a behavioural economics and finance guy. So, the idea that the RBA stops doing its important liaison work, where it goes out into the community to talk to businesses and investors to gauge the pulse of the economy, and stays in its Martin Place and Collins Street offices running models to me seems a dangerous step back.
The AFR says:
“The Reserve Bank has in recent years acknowledged that its forecasts are vulnerable to being wrong, moving in late 2013 to begin publishing in its quarterly outlook so-called fan-charts for its main economic growth and inflation forecasts that show a confidence range based on historical errors.”
But only someone who has never forecast anything would think that forecasts are anything but conditional guesses. Indeed, just last Friday Janet Yellen admitted the difficulty of forecasting when she said “I am describing the outlook that I see as most likely, but based on many years of making economic projections, I can assure you that any specific projection I write down will turn out to be wrong, perhaps markedly so.”
Will a model which is based on someone’s, or an institution’s, view of the world being hard coded into a “machine” really alleviate the risk of error? It’s simply a transferrence of risk from the “gut” to the model. Garbage in, garbage out they say.
Of course every business, no matter how well it is going, should conduct a review of its performance to ensure that it retains best practice.
But, if you were to objectively assess the RBA’s performance against say, the Fed, BoE, BoJ or ECB in the last 15 years, who would you choose to run monetary policy and your economy?
If you then overlay that performance with an small open economy, free float of its currency and a number one trading partner who has the ability to swamp every single industry in that economy and which drove a once in a century mining boom, who would you choose then?
The RBA is not perfect, but I hope the review still allows the RBA the leeway it needs to adjust policy based on both modelling and its interactions with humans.