Optimizing Policymakers’ Loss Functions in Crisis Prediction: Before, Within or After?
Early-warning models most commonly optimize signaling thresholds on crisis probabilities. The expost threshold optimization is based upon a loss function accounting for preferences between forecast errors, but comes with two crucial drawbacks: unstable thresholds in recursive estimations and an in-sample overfit at the expense of out-of-sample performance. We propose two alternatives for threshold setting: (i) including preferences in the estimation itself and (ii) setting thresholds ex-ante according to preferences only. Given probabilistic model output, it is intuitive that a decision rule is independent of the data or model specification, as thresholds on probabilities represent a willingness to issue a false alarm vis-à-vis missing a crisis. We provide simulated and real-world evidence that this simplification results in stable thresholds and improves out-of-sample performance. Our solution is not restricted to binary-choice models, but directly transferable to the signaling approach and all probabilistic early-warning models.
Does Machine Learning Help us Predict Banking Crises? ...
Urban Renewal in Saxony: A Need for a more Problem-orientated Allocation of Subsidies!
Wirtschaft im Wandel,
Currently, the political discussion of urban renewal in East Germany focuses more and more on new strategies to solve the problems of urban decline and vacancies on the housing market. Since 2001, the demolition of housing has been subsidized with approximately one billion Euros. Critics of this strategy argue that the continuation of demolition leads to a fragmentation of cities and a loss of urban functions. Therefore, they suggest to focus more on revitalization of residential quarters and to allocate more subsidies to improve neighbourhoods as well as residential amenities to lower housing vacancies.
This article argues that on the one hand, the overall housing vacancy-rates cannot be lowered with the current instruments of urban revitalization. Even though, there is a chance to attract citizens from the periphery of cities. This means to redistribute housing vacancies instead of an overall reduction. On the other hand, this strategy needs to be clearly focused on selected cities in which a potential of immigration exists. However, empirical results from Saxony suggest a different picture: The allocation of subsidies for urban revitalization shows no identifiable pattern. Therefore, the author proposes to refocus the policy of urban redevelopment.
The Loss Distribution of the Entrepreneurial Bad Debt Risk – a Simulation-based Model
IWH Discussion Papers,
The risk of bad debt losses evolves for companies which grant payment targets. Possible losses have to be covered by these companies equity and liquidity reserves. The question of how to quantify the level of risk of bad debt losses will be discussed in this paper. Input values of this risk are the probability of default, exposure at default and loss given default. It is shown how companies can derive probability functions to describe uncertainty and variability for each input value. Based on these probability functions a simulation model is developed to quantify the risk of bad debt losses. Based on an empirical study probability functions for probability of default and loss given default are presented.
Regional analysis of East Germany: A comparison of the economic situation of states, districts, and municipalities
Wirtschaft im Wandel,
A decade after the German unification we look at the extent of economic differentiation within East Germany. This is achieved by help of a set of selected statistical indicators for the years 1991 to 1998. Comparisons are drawn a) between the East German jurisdictions and b) between West and East German jurisdictions. On the federal state (Laender) level it can be shown that each state has developped its own specific economic profile. Brandenburg is characterized by a positive net migration (suburban function for Berlin), relatively low unemployment and high GDP values, but relatively low entrepreneurial activities. Saxony has achieved the lowest unemployment, a good endowment with human capital, modern industrial technology, infrastructure, and entrepreneurial activities. Special features of Thuringia consist of a relatively large number of patent applications and a stable industrial base. The economic state of Mecklenburg-Vorpommern is characterized by low industrial investment, negative net migration, and high unemployment. A special feature of this federal state is the intense investmenr in tourist services. Saxony-Anhalt registers the highest decrease in the numbers of industrial workers between 1991 and 1998 and the highest unemployment. On the other side it shows the highest amount of investment, especially in chemical industry and in mineral oil processing.
On the county level four clusters can be identified by means of a cluster analysis: A “cluster of counties with severe economic weaknesses” with a bias in the regions indutrialized in an early stage, a “cluster with a high human capital potential and suburbanization loss” consisting of 21 cities, a “cluster of counties with good economic results” predominantly surrounding the larger cities, and a “cluster of counties with SME growth potential” concentrating in Thuringia and Saxony.
The results at the city level show that the larger cities above 100.000 inhabitants, especially Dresden and Leipzig, do better than the smaller cities. Jena in Thuringia has specialized as a location for R&D, Zwickau in Saxony as a location for the automobile industry. Altogether the economic differences between the East German federal states, counties, and cities still are less pronounced than the degree of differentiation of their West German counterparts.