https://hdl.handle.net/10419/264607 2026-10-13T02:42:18Z https://hdl.handle.net/10419/318435 Title: Dual interpretation of machine learning forecasts Authors: Goulet Coulombe, Philippe; Göbel, Maximilian; Klieber, Karin Abstract: Machine learning predictions are typically interpreted as the sum of contributions of predictors. Yet, each out-of-sample prediction can also be expressed as a linear combination of in-sample values of the predicted variable, with weights corresponding to pairwise proximity scores between current and past economic events. While this dual route leads nowhere in some contexts (e.g., large cross-sectional datasets), it provides sparser interpretations in settings with many regressors and little training data-like macroeconomic forecasting. In this case, the sequence of contributions can be visualized as a time series, allowing analysts to explain predictions as quantifiable combinations of historical analogies. Moreover, the weights can be viewed as those of a data portfolio, inspiring new diagnostic measures such as forecast concentration, short position, and turnover. We show how weights can be retrieved seamlessly for (kernel) ridge regression, random forest, boosted trees, and neural networks. Then, we apply these tools to analyze postpandemic forecasts of inflation, GDP growth, and recession probabilities. In all cases, the approach opens the black box from a new angle and demonstrates how machine learning models leverage history partly repeating itself. 2025-01-01T00:00:00Z https://hdl.handle.net/10419/318509 Title: Health and long-term care insurance wealth in Austria Authors: Koman, Reinhard; Hofmarcher, Maria M.; Holzmann, Robert Abstract: Estimates of social insurance pension wealth are available for a number of Western economies, including Austria for the year 2017. Such wealth may be compared with conventional wealth in terms of size and distribution and when we add such estimates to measures of wealth in the conventional sense, we arrive at measures of augmented wealth. In this paper, we estimate health and long-term care insurance wealth in Austria and add that to conventional wealth estimates for Austria to achieve a further measure of augmented wealth. To our knowledge, this is the first attempt worldwide to do this. The resulting magnitude of health and long-term care insurance wealth is substantial, namely EUR 238,000 at the household level. It is comparable in scale to both pension insurance wealth (EUR 245,000) and net wealth in the conventional sense, i.e. property plus financial wealth minus debt (EUR 250,000). As regards distributive characteristics, health and longterm care insurance wealth is rather equally distributed (Gini coefficient of 0.31), compared to pension insurance wealth (0.45) and conventional wealth in Austria (0.73). The Gini coefficient for the new augmented wealth distribution (conventional wealth plus health and long-term care insurance wealth) is 0.47. 2025-01-01T00:00:00Z https://hdl.handle.net/10419/318520 Title: How Phillips curve dynamics enhance business cycle synchronization analysis in Central and Eastern Europe Authors: Petz, Nico; Zörner, Thomas Abstract: This paper analyzes business cycle synchronization and the Phillips curve (PC) relationship in Central, Eastern, and Southeastern European (CESEE) economies relative to the euro area. We find an overall increase in business cycle synchronicity, particularly among Euro adoption candidates, with notable heterogeneities during the early 2000s, the global financial crisis, and the euro crisis. Using a Kalman filter to extract business cycles and various measures of synchronicity, we show that CESEE EU countries align more closely with the euro area than non-EU countries. The unemployment-inflation relationship, analyzed with time-varying parameter (TVP) models, reveals a steepening of the Phillips curve postCOVID-19, with negative slope coefficients across all countries. We observe a growing convergence of the PC slope toward the euro area, especially in candidate countries. These results highlight the role of EU membership in fostering economic synchronization and emphasize the importance of considering time-varying dynamics in assessing economic convergence amid major shocks. 2025-01-01T00:00:00Z https://hdl.handle.net/10419/312869 Title: Improving disaggregated short-term food inflation forecasts with webscraped data Authors: Beer, Christian; Ferstl, Robert; Graf, Bernhard Abstract: This study examines the effectiveness of using webscraped data to predict price developments in the Austrian food retail sector. We calculate monthly nowcasts of price changes based on daily price data collected by the OeNB since mid-2020, using Eurostat methodology for price index calculation, along with further details provided by the national statistics office. We assess the quality of our nowcasts by comparing them with various baseline models and more advanced time series methods also covering machine learning approaches. Our findings indicate that webscraped data are a useful way to obtain more accurate nowcasts with a time advantage, amounting to several weeks, over traditional data sources. In addition, we are the first, to our knowledge, to explore the possibility of using the improved accuracy of the nowcasts as a basis for disaggregated short-term forecasts that extend up to one quarter. While direct forecasts at higher levels of aggregation produce slightly more accurate overall metrics, indirect forecasts derived from disaggregated data provide superior insights into the underlying dynamics of specific sub-components. Our results show that more advanced time series models have trade-offs in terms of computational efficiency while performing very similarly to more traditional methods. These findings have implications for policymakers who aim to develop an effective system for real-time monitoring of inflation dynamics at a very granular level. 2025-01-01T00:00:00Z