Central European Journal of Economic Modelling and Econometrics

House prices are of special importance for monetary policy since their sudden falls are usually associated with credit crunch followed by long-lasting and painful recessions. Despite several spectacular episodes of such events, each time house prices exhibit long-lasting growth trend with little volatility around it, it is argued that this pattern is a “new normal”. This paper shows that a central bank following this view would increase the volatility of inflation and output as compared to a policy that assumes high volatility of house prices. In the former case the monetary authority would conduct too accommodative monetary policy during abrupt house price expansions significantly increasing output and inflation fluctuations. In the latter situation, in turn, the policy would work well irrespective of the realized house price volatility.

We built a logistic regression Early Warning Models (EWM) for banking crises in a panel of 47 countries based on data from 1970-2014 using candidate variables that cover macro and financial market indicators. We find that VIX, a proxy of global risk-premium, has a strong signalling properties and that low VIX (low price of risk) increases likelihood of crisis. It does not only mean that stability leads to instability, but that this tends to be a global rather than a domestic phenomenon. We also find that particularly high contribution of financial sector to GDP growth often precedes crises, suggesting that such instances are primarily driven by excessive risk taking by financial sector and may not necessarily be sustainable. Other variables that feature prominently include credit and residential prices. Models using multiple variable clearly outperform single variable models, with probability of correct signal extraction exceeding 0.9. Our setting includes country-specific information without using country-specific effects in a regression, which allows for direct application of EWM we obtain to any country, including these that have not experienced a banking crisis.

The work contains discussions and simulation analyses of the expectation formation processes, taking account of the data revisions. In particular, it contains results of simulations examining statistical properties of the rationality tests and extrapolation processes, with particular focus on their behaviour in the case of short samples and data with measurement errors. The conclusions indicate that the rationality test based on the optimal regression and the proposed adaptive and accelerating tests are the most efficient and flexible. The tests showcasing best properties have been applied to a new set of macroeconomic forecasts for Poland. The results show that there are no grounds for rejecting the hypothesis on the rationality of forecasts derived from the National Bank of Poland (NBP) and the Organisation for Economic Cooperation and Development; however, this property was rejected for the European Commission. What is more, the comparative analysis indicates that only the national institution (NBP) may potentially aim the final readings of the macroeconomic data as the forecasting target. Finally, it transpires that the extrapolative models, albeit simple and intuitively interpreted, generally fail to correctly explain the forecast formation processes regarding the Polish economy.

We propose a Bayesian approach to estimating productive capital stocks and depreciation rates within the production function framework, using annual data on output, employment and investment only. Productive capital stock is a concept related to the input of capital services to production, in contrast to the more common net capital stock estimates, representing market value of fixed assets. We formulate a full Bayesian model and employ it in a series of illustrative empirical examples. We find that parameters of our model, from which the time-path of capital is derived, are weakly identified with the data at hand. Nevertheless, estimation is feasible with the use of prior information on the production function parameters and the characteristics of productivity growth. We show how precision of the estimates can be improved by augmenting the model with an equation for the rate of return.