Three essays on institutional investors' holding and trading activities
The purpose of my dissertation is to demonstrate the effects of institutional) investors' preferences on their trading and holding activities. To be specific, I look at the choices of bonds in certain maturity ranges of insurance companies, their tendency not to frequently trade. I also look at institutional investors' responses to rating changes. In the first manuscript, I present direct and comprehensive evidence on preferred habitat by investigating the Treasury bond duration choices of property liability insurance companies. I model insurers' portfolio durations as a partial adjustment process toward unobservable target durations. The findings confirm the presence of preferred habitat. First, I find evidence consistent with the duration matching hypothesis that durations of insurers' Treasury bond portfolios and expected claim payments are positively correlated. Second, durations of insurers' Treasury bond portfolios are positively correlated with insurers' risk management ability. Finally, firms may deviate from targets if yield premiums offered at non-preferred maturities are sufficiently high. The second manuscript extends the first study from the Treasury bond market to the corporate bond market. Insurance companies collectively own one third of outstanding corporate bonds (Campbell and Taksler, 2003). They are conventionally viewed as buy-and-hold passive investors. I confirm this by demonstrating a strong negative relationship between insurance ownership and liquidity. Also, there is a robust positive relation between insurance ownership and bond yield spreads after controlling for bond liquidity. This suggests either the presence of unpriced risk factors correlated with insurance ownership or insurers have a superior ability in picking bonds. The third manuscript investigates how institutional investors trade on the analyst recommendations change and examine stock performance following their trading activities. I find institutional investors' trades react to rating downgrades more quickly and in a greater magnitude than to upgrades. Regardless of upgrades or downgrades, most heavily purchased stocks beat those least purchased while least sold stocks beat those most heavily sold in the portfolio formation quarter. Further, upon stock upgrades, stocks least sold or heavily purchased by institutional investors with superior past performance persistently outperform those heavily sold or least purchased by those investors.