The impact of pre-selected variance inflation factor thresholds on the stability and predictive power of logistic regression models in credit scoring

Background: There are various studies that confirm the efficiency of the Johannesburg Stock Exchange (JSE), implying that there are no opportunities for active portfolio managers to earn excess returns over the long run.Aim: The aim of the research is to prove that the sub-indices on the JSE go through cycles of efficiency and inefficiency even though the JSE as a whole might be considered informationally efficient.Setting: Although the JSE as a whole can be considered to be weak-form efficient, portfolio managers are not bound to investing in large liquid stocks alone. Many aggressive funds allow managers to also allocate a portion of their portfolio to smaller stocks. This has implications when considering the efficiency of the stocks being selected.Methods: Given the impact efficiency has on portfolio selection, we test for the adaptive market hypothesis using a representative sample of stock indices by means of the automatic variance ratio test, the Chow–Denning joint variance ratio and the joint sign test on the JSE.Results: Our results confirm that some of the smaller, and in some instances younger, indices are not always as efficient as the all share index, thus allowing portfolio managers with an active management approach some opportunities to profit from informational inefficiencies in the market.Conclusion: The practice of active management by portfolio managers in the South African market seems to defy logic if one considers the fact that the JSE as a whole is at the very least weak-form efficient. By proving that some of the sub-indices that make up the all share index are inefficient most of the time, this article shows that the phenomenon of active portfolio managers is less of a surprise.

In this paper, we investigate the performance of a variety of frequentist estimation techniques for the scale and shape parameters of the Lomax distribution. These methods include traditional methods such as the maximum likelihood estimator and the method of moments estimator. A version of the maximum likelihood estimator adjusted for bias is included as well. Furthermore, an alternative moment-based estimation technique, the L-moment estimator, is included, along with three different minimum distance estimators. The finite sample performances of each of these estimators are compared in an extensive Monte Carlo study. We find that no single estimator outperforms its competitors uniformly. We recommend one of the minimum distance estimators for use with smaller samples, while a bias-reduced version of maximum likelihood estimation is recommended for use with larger samples. In addition, the desirable asymptotic properties of traditional maximum likelihood estimators make them appealing for larger samples. We include a practical application demonstrating the use of the described techniques on observed data.