logo

Forecasting, Forecrusting?

There is no more practical thing than a good theory. Furthermore, the strong predictive power of such theory – as Milton Friedman (1953) argued[i] – is a necessary precondition to consider the given theory as an instructive one.
A non-negligible proportion of prominent economists argue that forecasting is the raison d’être of the science of economics. When it comes to forecasting, we enter the sphere of not being satisfied with the proper descriptive way we look at the past and the present processes, but we are to be barded with knowledge that can guide our future actions and can help us to evaluate the impacts of our current actions (e.g. policymaking).
The great deal of experience with forecasting however gives us the impression that being a cautious economist is a must.
For example, almost two years ago, on 22 April 2014, a news appeared from Bianco Research[ii] which pointed out that according to their survey of 67 authoritative economists during that month showed every single one of them expected the 10-year Treasury yield to rise in the next six months. It did not happen, it was falling. This dispiriting forecasting power is now widely documented. Philip Tetlock, a Berkeley psychiatrist, made a 20-year long study on predictions made by experts on television and get quoted in newspapers and found that they are no better than the rest of us at prognostication.[iii] With the benefit of hindsight it seems that experts could deal only with the crust (i.e. do some ’forecrusting’) of the reality. Dealing only with the crust of a tree will not say much about how it has developed (i.e. understanding the inside phenomena is a must by looking at growth rings).
To my best knowledge, one of the most insightful forecasts was done by Friedrich August von Hayek, an Austrian economist, during the 1920s about the forthcoming crisis of 1929-1933. Albeit, pursuing better and better forecasting is a noble and honourable undertaking, one must not forget what Hayek himself said during his Nobel Prize lecture in 1974. Hayek spoke about the pretence of our knowledge and warned the profession to pursue the avoidance of really bad outcomes rather than pushing policymaking to perpetually optimise the good ones. It means that policymaking does not have the necessary and comprehensive knowledge to do policy optimisation, but it certainly has the potential to bring hectic movement into the life of the society through scientism-based[iv] policy optimisation interventions, which might depress the trust infrastructure. And as the Nobel-laureate George Akerlof together with his co-author Robert J. Shiller emphasised in the book Animal Spirits, the societal and economic development relies heavily on the trust infrastructure of the given country.
This joins the line of thinking of Daniel Kahneman (Nobel Laureate) who emphasised that the predictive power of our knowledge reaches its diminishing marginal returns relatively fast. There is no gainsaying the fact that this psychological finding holds especially in case of predicting non-linear processes evolving in the complex global system interspersed with growing interconnectedness and interdependency and the current dominance of uncertainties (e.g. estimating ex ante the value of fiscal multiplier precisely is particularly cumbersome as it was admittedly the case documented by Olivier Blanchard and Daniel Leigh (2013)). Unsurprisingly, the fiscal policy related forecasts’ predictive power has never reached the 50%, either (Tóth, 2014)[v].
Of course, with forecasting – on future market shares, potential interest rates, GDP growth numbers etc. – we recurrently create a series of events based on non-linear behaviour of public-private and civic sectors’ agents. By analysing and collecting statistical data and information, the observer does nothing, but synthetically manufacturing reality, which does not necessarily equal to the “real” reality. And these manufactured realities are differing across observers along the complex process of unconscious perception and conscious information processing. This can easily lead to mishaps in understanding what is really going on in the economy. In an economy, which has now become intensively dominated by financial sector of which intermediation role has been to some extent relegated with respect to supporting the real economy. It implies that the financial sector’s productivity per se a non-existent category, its real function shall be deciphered along the issue of to what extent and how it is conducive to the real socio-economic development of the real economy (e.g. in terms of productivity, decreasing excessive income inequality, supporting quantitative as well as qualitative growth pervaded by the view of greening out the economy etc.). By modelling and forecasting financial market phenomena and variables, we influence our behaviour in a non-linear fashion leading to the outcome that could not be predicted (i.e. a philosophical and non-testable question arises namely that is it the case that our models might have been not so bad, but we resulted in detours from their predictions via our series of interactive and reflexive actions in pursuing optimisation?)[vi].
In summary, it seems that the legitimacy of our profession partly comes from its forecasting activity. Still, recognising that forecast is not a panacea is a must since forecasts are always ‘If, then’-type considerations with a rather wide spectrum of assumptions. It was rightly accentuated during the Spring Seminar 2016 organised by wiiw[vii] (The Vienna Institute for International Economic Studies) last week where the institution’s official forecast was presented for the Central and Eastern European states (e.g. the institute expects a 3% rate of GDP growth in case of the region if certain conditions and assumptions are taken into account). As a corollary, financial markets, policymakers should build more on comprehensive and deep case studies rather than cultivating only the quantitative and forecast oriented way of functioning if they are to avoid really bad outcomes through the creation of a chain of events unintendedly. Policymakers should pursue micro-realism rather than macro-idealism. Related to this, remembering to the old-findings of Pitirim Sorokin would be of paramount importance. Sorokin criticised 57 years ago the rampant quantification methods/models built on the short term and fresh data in trying to get some predictive considerations. Taking into account the world’s nuances and the complexity we live in (Sorokin, 1956)[viii] by contemplating longer term socio-economic phenomena arising potentially in a way of ‘creeping normalcy’ are essential to avoid ‘forecrusting’ by looking beneath the surface.

[i] See: Friedman, M. (1953): The Methodology of Positive Economics. In: Friedman, M. (1953): Essays in Positive Economics. University of Chicago Press, Chicago, pp. 3-43.

[ii] See: http://blogs.marketwatch.com/thetell/2014/04/22/100-of-economists-think-yields-will-rise-within-six-months/ Accessed on: 14.04.2016

[iii] See: Tetlock, P. E. (2006): Expert Political Judgment: How Good Is It? How Can We Know? Princeton University Press; New Ed edition (August 20, 2006). p. 352

[iv] Hayek, F. A. von (1952): Scientism and the Study of Society. Economica, Vol. 8, No. 35, August 1942, reprinted in The Counter-Revolution of Science, Glencoe, Ill., 1952, p. 15 of this reprint.

[v] Tóth, G. (2014): The Forecasting Capacity of Indicators Measuring Budget Sustainability. Public Finance Quarterly 2014/4 (p. 511-528.)

[vi] Welch and Goyal (2008) reexamined the performance of variables that have been suggested by the academic literature to be good predictors of the equity premium. The authors found that by and large, these models have predicted poorly both in-sample (IS) and out-of-sample (OOS) for 30 years now; these models seem unstable, as diagnosed by their out-of-sample predictions and other statistics; and these models would not have helped an investor with access only to available information to profitably time the market. See: Welch, I. – Goyal, A. (2008): A Comprehensive Look at the Empirical Performance of Equity Premium Prediction. Review of Financial Studies, Vol. 21, No. 4, pp. 1455-1508

[vii] See: http://wiiw.ac.at/growth-stabilises-investment-a-major-driver-except-in-countries-plagued-by-recession-p-3822.html Accessed on: 14.04.2016

[viii] Sorokin, P. A. (1956): Fads and Foibles in Modern Sociology and Related Sciences. Henry Regnery, Chicago.