Essays on delegated asset management in illiquid markets

UnrestrictedThis dissertation consists of three chapters of interrelated work in which I investigate the implications of money management incentives to delegated asset allocation and to asset pricing in the context of illiquid markets.; In the first chapter, using a continuous-time dynamic portfolio choice framework, I study the problem of an investor who exogenously decides to delegate the administration of her savings to a risk-averse money manager who trades multiple risky assets in a thin market. I consider a manager who is rewarded for increasing the value of assets under management, which is the product of both the manager's portfolio allocation decisions, taken over the investment period, and the money flows into and out of the fund, as a result of the portfolio performance relative to an exogenous benchmark. The model proposed here shows that, whenever the manager can substitute between more illiquid and less illiquid risky assets, she is likely to choose to hold an initial portfolio that is skewed toward more illiquid assets, and to gradually shift toward less illiquid assets over the investment period. The model further shows that several misalignments of objectives between the investor and the manager can lead to large utility costs on the part of the investor, and that these costs decrease with asset illiquidity. Solving for the shadow costs of illiquidity, the model indicates that delegated rather than direct investing is likely to lead to larger price discounts.; The second chapter is joint work with Juan Sotes-Paladino. It builds upon the first chapter and extends the theoretical framework to analyze a liquidity-constrained dynamic asset allocation problem in which investors delegate their portfolios to mutual funds that operate under a family organization. The funds are allowed to cross their trades of illiquid common holdings in response to the interests of the family as a whole, whereas investors' flows are assumed to respond asymmetrically to funds' performance, rewarding good performers disproportionately more than they penalize bad ones. We then study a previously unexplored channel through which fund families can play favorites among their affiliated funds: to have some funds avoid the costs of illiquidity by making others adopt suboptimal investment decisions. We find that families' ability to cross-trade among member funds allows them to save on transaction costs but at the same time elicits higher risk-taking by affiliated fund managers, compared to their standalone counterparts. We show that the additional costs of agency that investors incur under such a money management structure are likely to increase with asset liquidity. However, by imposing position limits on their funds' portfolios, we show that investors can improve their welfare outcome. We further find that families' optimal strategies can induce a negative correlation between their affiliated funds' after-flow returns, creating diversification benefits to the family's overall portfolio. Nevertheless, the correlation between portfolios' returns can still be higher for affiliated funds than for comparable standalones, due to the overlap in holdings. Finally, the model in this chapter suggests that, as market liquidity increases, fund families are likely to favor more correlated asset holdings within each member fund's portfolio.; The third chapter is joint work with Breno Schmidt. In this third chapter we propose to test some of the empirical implications derived from the results of the second chapter. In particular, we propose to test the hypothesis that mutual funds affiliated with large families tend to coordinate trades so as to mitigate the damaging effects that liquidity shortfalls can have on funds' performance and consequent shareholder redemptions. From this co-insurance hypothesis, we intend to derive important implications for the price reaction to asset fire sales and for managerial risk-taking incentives. First, we use a bootstrap-based approach to show that offsetting trades among funds affiliated with larger families are more likely to be the result of coordinated trade strategies. Second, consistent with internal coordination, our preliminary results indicate that there is weak or no price pressure on traded securities mostly held in common by distressed funds affiliated with large families. Finally, we make use of the known result that, due to strategic complementarities among investors' share redemptions, the generally convex relationship between flows and past performance does not appear to hold true for funds invested in less liquid assets. Consistent with our working hypothesis, our preliminary results indicate that this is true only for funds affiliated with small families. As a result, by improving the convexity of their implicit payoff structures, co-insurance strategies at the family level can potentially encourage individual fund managers to take extra risks