In the recent years, data science methods have been developed considerably and have consequently found their way into many business processes in banking and finance. One example is the review and approval process of credit applications where they are employed with the aim to reduce rare but costly credit defaults in portfolios of loans. But there are challenges. Since defaults are rare events, it is—even with machine learning (ML) techniques—difficult to improve prediction accuracy and improvements are often marginal. Furthermore, while from an event prediction point of view, a non-default is the same as a default, from an economic point of view much more relevant to the end user it is not due to the high asymmetry in cost. Last, there are regulatory constraints when it comes to the adoption of advanced ML, hence the call for explainable artificial intelligence (XAI) issued by regulatory bodies like FINMA and BaFin. In our study, we will address these challenges. In particular, based on an exemplary use case, we show how ML methods can be adapted to the specific needs of credit assessment and how, in the case of strongly asymmetric costs of wrong forecasts, it makes sense to optimize not for accuracy but for an economic target function. We showcase this for two simple and ad hoc explainable ML algorithms, finding that in the case of credit approval, surprisingly high rejection rates contribute to maximizing profit.

Artificial Intelligence (AI) is one of the most sought-after innovations in the financial industry. However, with its growing popularity, there also is the call for AI-based models to be understandable and transparent. However, understandably explaining the inner mechanism of the algorithms and their interpretation is entirely audience-dependent. The established literature fails to match the increasing number of explainable AI (XAI) methods with the different stakeholders’ explainability needs. This study addresses this gap by exploring how various stakeholders within the Swiss financial industry view explainability in their respective contexts. Based on a series of interviews with practitioners within the financial industry, we provide an in-depth review and discussion of their view on the potential and limitation of current XAI techniques needed to address the different requirements for explanations.

No abstract

In two-pass regression-tests of asset-pricing models, cross-sectional correlations in the errors of the first-pass time-series regression lead to correlated measurement errors in the betas used as explanatory variables in the second-pass cross-sectional regression. The slope estimator of the second-pass regression is an estimate for the factor risk-premium and its significance is decisive for the validity of the pricing model. While it is well known that the slope estimator is downward biased in presence of uncorrelated measurement errors, we show in this paper that the correlations seen in empirical return data substantially suppress this bias. For the case of a single-factor model, we calculate the bias of the OLS slope estimator in the presence of correlated measurement errors with a first-order Taylor-approximation in the size of the errors. We show that the bias increases with the size of the errors, but decreases the more the errors are correlated. We illustrate and validate our result using a simulation approach based on empirical data commonly used in asset-pricing tests.

A semi-strong efficient market incorporates relevant new information immediately. Using an event study, we investigate whether and to what extent regular earnings announcements of Swiss companies listed on the Swiss Market Index show the expected effects in share prices. For this purpose, we test for abnormal returns caused by earnings announcements in the period from 2012 until 2022. In contrast to previous studies of the Swiss market, we find that deviations from analysts’ expected earnings lead to pronounced immediate movements in stock prices, as predicted by the semi-strong efficient market hypothesis. Pre- and post-announcement abnormal returns are modest and generally not statistically significant.