Media Briefing
Non-Technical Summary of my WP entitled "Strategic Judgment: its game-theoretic foundations, its econometric elicitation", presented at the ESEM2021 in Copenhagen.
Are professional forecasters and Central Banks sensitive to strategic cheating? The answer from new research by Emilio Zanetti Chini is simple: they are, and significantly more than suspected. This complicates the work to central bankers and, in general, to all the people interested in understanding the future dynamics of the economy.
In order to pursue their institutional duties, Central Banks adopt a variety of mathematical models to summarizes the whole available amount of informations on the past and present state of the economy and produce forecasts on the future. But the models produced by the same institutions may still be convey biased results. For this motivation, since many decades the Central Banks uses an additional source of information: the professional forecasters. These are experienced and recognized analysts that have their own informations, their own models, and report their own forecasts directly to the Banks, hence constituting a link among the Banks’ “institutional” forecasts and market expectations. A proper match among institutional and professional forecasts ensures all economic agents about the Banks’ deliberateness, coherence and objectivity.
Unfortunately, building a “consensus” forecast is far from being an exact science. In fact, professional forecasts do not agree with the institutional ones, nor among themself. Aggregating a so heterogeneous set of estimates is an activity highly sensitive to the evaluator’s inner policy. Not strangely, communicating economic forecasts or outlooks is one of the instruments that Banks adopt to address their policies. In turn, this may influence the future behavior of the same professional forecasters, thus enforcing a loop that may induce the Central Bankers (or any other policy-makers) in error.
The new study takes a deeper look at the mechanics of the entire forecasting process by considering that both the Central Banks and Professional Forecasters may pursue their own policy, their own models, use their own informations and, consequently, add their own (informed) opinions to their mathematical estimates.
But the key aspect of the new analysis is that the author consider a crucial point: these “opinions” (or judgments) are not the mere materialization of the fate, nor of the irrationality of professional forecasters, but an output of the complex, dynamic game among Central Banks, Professional Forecasters, their policy objectives as well as their learning costs. Thus, it is a ‘strategic judgment’. The study introduces a new methodology that links all these key aspects in one general mathematical object, the “Scoring (or Judgmental) Structure”. This enables the investigator to define and verify empirically, via statistical testing, the possible over-amount of bias due to strategic judgment in the professional as well in institutional forecasts.
The authors’ own calculations suggest that the number of cases in which strategic judgmental bias is substantial: all the three case studies in the study (Federal Reserve Bank, Bank of England and Bank of Norway) convey evidence of strategic bias in the forecasts of several variables (namely, the U.S. Real Gross Domestic Product and Unemployment Rate, the U.K. inflation rate and the Norwegian Output Gap).
In particular, the Author finds that the majority of tests applied on the above mentioned data reject the false hypothesis that the variable under investigation is not biased. The strength of this rejection varies across variables and across time horizons.
We interpret these results as follows: economic forecasts, like all the experiments that rely on basic Popper’s criterion, are manipulable. Governments should not rely completely on the estimates provided by Central Banks’ or professional agents’ communications. Viceversa, market operators and investors should always apply some skepticism about the institutional communications and correct their investment plans accordingly.
ENDS
CONTACT DETAILS:
Emilio Zanetti Chini
Sapienza University of Rome (University of Bergamo since September 1st)
emilio.zanettichini@uniroma1.it / emilio.zanetti@fastwebnet.it
Website: https://sites.google.com/site/ezcwebsite/