Improving Investment Operations through Data Science: A Case Study of Innovation in Valuation

Abstract

New technologies in data science are allowing long-term investors to bring much more rigor to their operations. In this article the authors provide empirical examples in support of these data-driven advances, demonstrating their practical applications. They use the UC Investments office as their case study and discuss how adoption of advanced data science techniques can move organizations past the current unsatisfactory state of the art and toward an unprecedented level of operational finesse. Specifically, the authors focus on a methodological innovation in fair valuation of illiquid assets that is supported by an automated, rigorous process. They test this process in a real-world setting and find, at least in this case, that these advances can enhance roll forward outputs in terms of timeliness, accuracy, and granularity. This finding has several potential impacts, not only for reporting, but also for investment, risk management, actuarial purposes, and even personal compensation of teams.

TOPICS: Big data/machine learning, performance measurement