On the Valuation Skills of Corporate Bond Mutual Funds

Abstract

We introduce a new measure to assess the valuation skills of investment-grade corporate bond funds. Our measure recognizes funds that ex-ante hold a higher fraction of undervalued bonds as having better valuation skills. The measure predicts future fund performance, is stable over time, and is unrelated to other sources of skill. Fund investors recognize such skill by responding more to the past performance of funds with better valuation skills. Consistent with the equilibrium model of Gârleanu and Pedersen (2018), our evidence suggests that as growing capital gets allocated to skilled bond funds, the corporate bond market is becoming more efficient.

The Performance of Corporate Bond Mutual Funds and the Allocation of Underpriced New Issues

“…Fund age, net flows, expense ratios, turnover ratios, size, and family size all exhibit considerable variation, underscoring the need for controls in our empirical design. Also showing considerable variation are two measures that prior research has linked to fund performance: the Valuation Accuracy Score of Cici and Zhang (2021), which is higher for funds with better valuation skills, and the Liquidity Score of Anand et al (2021), which is higher for funds that profit from providing liquidity.…”

“…table reports summary statistics of the sample of 338 investment-grade corporate bond mutual funds. TNA is fund total net asset, TO is fund annual turnover ratio; EXP is fund expense ratio, Age is fund age measured in years, Flow is the monthly net flow ratio, RET is the monthly post-fee fund return, FmSz is the size of corporate bond holdings of the fund's family; VAS is the valuation accuracy score ofCici and Zhang (2021), and LiqScore is the liquidity score ofAnand et al (2021).…”

“…The dependent variable is monthly fund pre-fee abnormal return estimated over the prior 36 months based on the four-factor model. Independent variables include the offering return sensitivity (ORS) or its quintile ranking (ORS Quintile), valuation accuracy score (VAS) ofCici and Zhang (2021), liquidity score (LiqScore) ofAnand et al (2021), natural logarithm of fund total net asset (TNA), natural logarithm of size of corporate bond holdings of the fund's family (FmSz), natural logarithm of fund age measured in years (Age), monthly net flow ratio (Flow), expense ratio (Exp), and annual turnover ratio of the prior year (TO). All independent variables except for Exp are lagged.…”

“…Ind_Exp is the industry expertise of a fund family measured as described in Section 4.1. VAS is the valuation accuracy score ofCici and Zhang (2021) aggregated at the family level and measured at t. Underwriting is the fraction of the dollar amount of new offerings during [t+1, t+24] underwritten by underwriters with which a family had a significant relation. The control variables include natural logarithm of the size of corporate bond holdings of the fund's family (FmSz); flows (Flow), expense ratios (EXP), and turnover ratios (TO) of the prior year averaged across all bond funds in the family and weighted by fund assets in the previous month.…”