We show that fund families allocate their most skilled managers to market segments in which manager skill is rewarded best. In efficient markets, even skilled managers cannot generate excess returns. In less efficient markets, skilled managers can exploit inefficiencies and generate higher performance than unskilled managers. Fund families seem to be aware of the relation between skill, efficiency, and performance, and allocate more skilled managers to inefficient markets. They pursue this strategy when hiring new fund managers and when reassigning managers to funds within the family. Overall, we conclude that fund families allocate fund managers in an efficient way.
This paper is the first to study whether fund families allocate fund managers to market segments so that manager skill is rewarded best. This issue is vital since fund performance depends crucially on the fund manager (e.g., Baks, 2003) and determines the money inflow into the fund (e.g., Sirri and Tufano, 1998). As a fund family typically charges a fixed percentage fee on its assets under management, manager allocation ultimately determines the profitability of the fund family.
Our first main hypothesis is that manager skill is rewarded more in less efficient markets. If a market is fully efficient, prices reflect all information and even highly skilled managers cannot generate excess returns. In less efficient markets, however, skilled managers can exploit inefficiencies and generate excess returns, which unskilled managers are unable to do. Given that skill is rewarded more in less efficient markets, labor economics theory suggests that fund families should allocate more skilled managers to less efficient market segments. This is our second main hypothesis.
We test these hypotheses using data from the investment grade (IG) and high yield (HY) corporate bond market segments. We choose these market segments because they differ with respect to their efficiency: The HY segment is less efficient than the IG segment.
We first test whether skill pays off more in the less efficient HY segment. Our regression analysis supports this hypothesis, even after we control for various manager and fund characteristics. Skill pays off in the less efficient HY segment, but not in the more efficient IG segment. In our second set of tests, we analyze whether fund families allocate managers to market segments so that their skill pays off best. We find strong evidence for such behavior: Fund families allocate more skilled managers to the less efficient HY segment.
Our paper is related to two strands of the literature. First, it contributes to the literature analyzing the impact of manager skill on fund performance (e.g., Golec, 1996, Chevalier and Ellison, 1999b, Gottesman and Morey, 2006 and Li et al., 2011). Our study reconciles contradictory evidence on the impact of skill on performance provided in earlier studies by showing that this impact depends on market efficiency.
Second, our paper contributes to studies that explore how fund families allocate managers. Khorana (1996) and Chevalier and Ellison (1999a) analyze hiring and firing of fund managers. Drazin and Rao (2002) study how fund families allocate already-employed managers to newly founded funds. We add to this literature by demonstrating that fund families allocate managers to market segments according to manager skill and market efficiency.
The rest of the paper is structured as follows. In Section 2, we outline a conceptual model to substantiate our hypotheses. In Section 3, we provide evidence that the HY market is less efficient than the IG market. In Section 4, we describe our data and present summary statistics. In Section 5, we test our first main hypothesis: Manager skill pays off more in less efficient market segments. Section 6 presents results on our second main hypothesis: Fund families allocate more highly skilled managers to less efficient market segments. In Section 7, we provide several robustness checks. Section 8 summarizes and concludes.
In this paper, we show that fund families allocate fund managers in a sensible way. They allocate more skilled managers to less efficient market segments since skill is rewarded more there.
We come to this conclusion by studying US fund managers in the investment grade (IG) and the high yield (HY) corporate bond market. We show that the impact of skill (measured as SAT score) on fund performance is larger in the less efficient HY segment than in the more efficient IG segment. We even find that skill is rewarded only in the less efficient HY segment. Fund families seem to be aware of this relation between skill, efficiency, and performance, and allocate more highly skilled managers to HY funds. Furthermore, they reallocate IG managers to HY funds according to the expected alpha (add-on) managers can generate in the HY segment. All these results are robust over time and remain stable when we use different ways to measure skill and performance and different estimation approaches.
The manager allocation strategy that we show is highly sensible since it increases HY fund alphas without destroying IG fund alphas. Hence, average alpha in the fund family increases and, as a consequence, the family attracts new money inflow and fee income.
