Continuous innovation in factor credit strategies

There is an large body of academic literature that documents the existence of factor premiums that enable investors to construct portolios that generate better risk-adjusted returns than passive investing. To ensure that such factor-based investment strategies continue to deliver on their objectives, continuous innovation is required, for example by further enhancing factor definitions and investment processes and investigating alternative sources of data. In this article we show two examples of recent innovations in the field of factor investing for corporate bonds.In the first study, we document how factors can be successfully applied to hard-currency emerging market credits. Size, low-risk, value and momentum factor portfolios outperform the emerging market credit universe on a risk-adjusted basis, for example shown by higher Sharpe ratios. A multi-factor portfolio moreover benefits from diversification between the individual factors and obtains the highest information ratio. The factor portfolios benefit from bottom-up allocations to countries as well as from bond selection within these countries. Analysing the break-even transaction costs shows that these factors remain attractive after taking transaction costs into account. While generic factors offer a good starting point, more complete risk-assessments limit the exposure to unrewarded risk. In the second study, we extend previous research that showed that the text from annual and quarterly corporate reports filed with the Securities and Exchange Commission can be used to predict equity returns and volatility. We examine whether the text in these reports is informative for corporate bond returns and volatility as well. Particularly, we test if bonds of ‘non-changers’, companies with few changes in their reports over time, outperform bonds of ‘changers’. In a large sample of corporate bonds over a twenty year period we find that results for equities do carry over to corporate bonds: in both the investment grade and high yield sub-samples non-changers outperform changers by more than 50bps annually.