We investigate the determinants of institutional investors' holdings in IPO firms and the post-issue relation between these holdings and operating performance. We find that institutions' initial holdings strongly depend on IPO characteristics, in particular, the public float. After controlling for public float, initial institutional holdings are unrelated to pre-IPO operating performance. During the first year after the IPO, average institutional holdings increase from 24% to 36% of shares outstanding and stabilize at about 42% by the end of the second year. Furthermore, post-IPO operating performance is positively related to institutional holdings. This relation holds also in lags, but diminishes towards the third year after the IPO. Overall, our findings indicate that institutional ownership is a valid indicator of the firm's operating performance in its initial years as a public company.

We analyze the benefits accruing to a risk averse investor of adding commodity futures to her portfolio of traditional assets using an out-of-sample framework that efficiently incorporates the information contained in commodity characteristics. We identify the futures carry as a key characteristic for tactical allocation among commodities. Adding futures to the traditional portfolio by conditioning on the carry leads to a significant increase in the Sharpe ratio of 427%, without an adverse impact on the distribution of portfolio returns. Indeed, a stochastic dominance methodology, which explicitly takes into account the investor's aversion towards negative skewness, indicates that the investor would strictly prefer the commodity futures carry augmented portfolio over the benchmark (stocks and bonds only) portfolio. 

We examine the effects of learning by experience on portfolio decisions of U.S. investors. Our evidence suggests that knowledgeable investors benefit the most from learning. That is, as knowledgeable investors gain experience, they improve portfolio construction and performance. However, there might be limits to learning since investors are heavily influenced by prior choices made near the time when the investor starts trading (primacy effect) and in the most recent period (recency effect). Overall, our evidence suggests that investment knowledge can enhance learning, which may be impeded by the impact of prior choices

Alliances come in many shapes and forms. We show in a simple formal model that alliances can resemble two broad types, depending on the degree of product market competition: Low competition alliances will resemble a stag hunt game in which contributing to the alliance is in the partners’ interests. High competition alliances are akin to a prisoner’s dilemma in which alliance partners have an interest in withholding effort, at least if the alliance game is played only once. We match these two alliance environments with different alliance configurations and vary the number of alliance partners and the level of commitment. We find that alliance configuration matters most in high competition alliances, while performance of low competition alliances is relatively unaffected by different alliance configurations. This points towards important contingencies for the choice of alliance configuration.