Using natural language processing to score companies by the news frequency of terms related to private equity, the authors create an index weighted by theme exposure and liquidity, whose returns are highly correlated with non-traded indexes.
Abstract
Using natural language processing, we score companies based on the frequency with which news articles contain both their names and terms private equity and leveraged buy-out. An index is then created and can be updated seamlessly at high frequency. The weights are set as a function of the relative exposure to this theme. We add liquidity constraints to ensure minimal transaction costs. Even though the algorithm does not optimize on either return or correlation, this listed private equity index is highly correlated to commonly used private equity fund market indices: nearly 90% correlation with Burgiss LBO fund index. In addition, our index has similar returns as non-tradable Leveraged Buy-Outs (LBO) fund indices. Our approach can be generalized to many other investment themes.