Securitized Real Estate is known for extreme return movements in secondary capital markets. Accordingly, the Gaussian assumption of normality has been empirically falsified for public equity positions including REITs. Thus, literature has intensified the research on Extreme Value Theory and Generalized Pareto Distributions for exceedances above a certain threshold. These observations in the tail of the return distribution can be empirically characterized by scale and shape parameters. Nonetheless, descriptive statistics are regularly enriched by inductive models to provide explanatory power of independent covariates. The central aim of the present study is the establishment of statistical significance between explanatory covariates to model scale and shape parameters statistically across time. Time-variant parameterization appears to be highly important, since empirical literature has shown volatility clustering, implying differently volatile market phases.