2007

PROJECTS FUNDED IN 2007

Liquidity and Credit Default Swap Spreads

Link to PDF File: Liquidity and Credit Default Swap Spreads
Hong Yan, University of South Carolina
Dragon Yongjun Tang, Kennesaw State University

We propose an empirical study on the pricing effect of liquidity level and liquidity risk in the credit default swaps (CDS) market. CDS is the key constituent of the fast growing credit derivatives market that has $34.4 trillion in total notional value by the end of 2006. Credit derivatives play an important role in today’s financial market by facilitating the transfer of credit risk. Credit derivatives are over-the-counter contracts executed through bilateral search. Trading motives include both credit risk management and, probably more notably, informed speculation. Government regulators around the globe have repeatedly expressed their concerns over the opacity and lack of comprehension of the credit derivatives market. A better understanding of the liquidity structure and its impact on the pricing of credit derivatives is critical to improving the efficiency and stability of financial markets and the overall health of the economy, as evidenced by the ongoing subprime mortgage crisis.

Our study represents the first systematic investigation of the effect of CDS liquidity characteristics and liquidity risk on CDS spreads, above and beyond the credit risk component. We first construct a set of liquidity proxies to capture various facets of CDS liquidity, such as adverse selection, search frictions, and inventory costs, using a comprehensive database on CDS transactions. We carefully analyze the determinants and interactions of the liquidity proxies to validate that they reflect aspects of bilateral matching, funding constraints, and informed speculation in CDS trading. Then we analyze the effect of these liquidity characteristics on CDS spreads. Our analysis completed thus far shows that the liquidity effect on CDS spreads is significant with an estimated liquidity premium on par with those of Treasury bonds and corporate bonds. We also find cross-sectional variations in the liquidity effect highlighting the interplay between search friction and adverse selection in the CDS market. Furthermore, using liquidity betas and volume respectively to measure liquidity risk, we provide supporting evidence for liquidity risk being positively priced beyond liquidity characteristics in CDS spreads. Our estimates indicate that liquidity characteristics and liquidity risk together could on average account for about 20% of CDS spreads.

Credit default swaps are fast becoming as important as the corporate debt instruments upon which they are defined. Instruments that often are highly illiquid characterize both markets. The proposed work should help practitioners better understand how illiquidity affects the pricing of these instruments.

The Troves of Academe: Asset Allocation, Risk Budgeting and the Investment Performance of University Endowment Funds

Link to PDF File: The Troves of Academe: Asset Allocation, Risk Budgeting and the Investment Performance of University Endowment Funds
Keith C. Brown, Texas Lorenzo Garlappi Texas
Cristian-Ioan Tiu, SUNY-Buffalo

In the proposed research, our objective is to study the asset allocation decisions of university endowment funds. The salient features of these institutional investors, such as an infinite investment horizon, tax exemption, decentralized management, and ability to implement relatively unrestricted investment strategies, make them ideal candidates for better understanding the effect of asset allocation models on portfolio performance. The problem is particularly interesting because, given their high tolerance for short-term volatility, university endowments have been at the vanguard lately in investing in alternative assets such as hedge funds, private equity, and real estate.

Our contribution will be to provide what we believe is the first comprehensive time-series and cross-sectional analysis of the investment practices and performance of an exhaustive sample of college and university endowment funds in the United States , Canada and Puerto Rico over a period spanning 1984 to 2005. Our data sources include an annual data set obtained from the National Association of College and University Business Officers (NACUBO).

The outcome of the analysis so far completed reveals that while the average endowment does not produce statistically significant risk-adjusted return, the subset of “active” endowments (i.e., those for which asset allocation is a less important component of total return) tend to outperform “passive” endowments. This out performance happens despite the fact that asset allocation is the most important determinant of return variation in the time series. We document that this discrepancy between cross-sectional and time-series importance of asset allocation emerges because endowment funds in our sample tend to take a remarkably similar level of passive risk. After developing and calibrating to existing data a risk budgeting model, we argue that despite the return potential that endowments can extract from asset allocation, this self-imposed “risk budget” is ultimately hurting the performance of the most passive funds. The access to richer data will allow us to analyze further aspects of the endowment investing process, such as the role of internal versus external managers and the effect of managerial turnover on performance.

Academic endowment funds represent a large and growing fraction of institutionally managed investment funds. This research will examine how these funds make investment allocation decisions. Investment managers will be interested in these results because they serve this market and because these decisions can be large enough to affect the market. Donors and the endowments themselves also will have a strong interest in the results as stakeholders in these funds.

When Benchmark Indices Have Alpha: Problems with Performance Evaluation

Link to PDF File: When Benchmark Indices Have Alpha: Problems with Performance Evaluation
Martijn Cremers, Yale
Antti Petajisto, Yale
Eric Zitzewitz, Dartmouth College

From 1980-2005, the one, three, and four-factor alphas of the S&P 500 and S&P Midcap 400 were approximately +0.4% and +2.4% per year, respectively, while those of the Russell 2000 were about -2.2%. Preliminary results suggest that the positive alphas of the S&P indices occurred disproportionately during periods during which the indexing and “closest indexing” of these indices increased. In contrast, about 70 percent of the negative alpha of the Russell 2000 was concentrated in June and July, suggesting it is related to index reconstitution effects, while the remainder occurred disproportionately during periods in which its popularity as a benchmark declined.

The relationship between benchmark alphas and changes in benchmark popularity suggests that asset prices reflect a convenience or liquidity premium for holding assets in a benchmark index. If changes in benchmark popularity are hard to predict, then arguably returns due to exposure to a benchmark should be thought of as factor returns, rather than as risk-adjusted performance.

Non-zero benchmark alphas have implications for attempts to draw inferences about managerial skill from standard risk-adjusted performance measures, particularly given the prevalence of closet indexing by nominally active portfolio managers. Standard one, three, and four-factor alphas overstate the historical performance of large and mid-cap managers, particularly those whose portfolios most closely track the S&P indices. On the other hand, alphas understate the performance of small cap managers, particularly those whose portfolios overlap with the Russell 2000. Benchmark index alphas are persistent, so for both large and small-cap, failure to control for non-zero index alphas may also lead one to overestimate the persistence of performance.

A very large fraction of institutional money is either indexed to-or evaluated relative to-market benchmark indices. Accordingly, the properties of these indices are of paramount importance to investment managers and investment sponsors. This research will examine systematic performance characteristics of various indices in comparison to various market factors. The results will help explain performance shortfalls and gains that may be due to the benchmark measures rather than to active (or in some cases, passive) strategies.

A Multiple Lender Approach to Understanding Supply and Demand in the Equity Lending Market

Link to PDF File: A Multiple

Lender Approach to Understanding Supply and Demand in the Equity Lending Market
Adam C. Kolasinski, University of Washington
Adam V. Reed, University of North Carolina
Matthew C. Ringgenberg, University of North Carolina

Using a unique database of equity lending transactions, we analyze the structure of the equity lending market. Using various instruments for share loan demand, we find that the loan supply schedule is flat, and slightly downward sloping, at low quantity levels. At higher quantity levels, however, it has a positive slope. This helps reconcile seemingly conflicting findings in the literature. Some researchers find that lending fees are unresponsive to increases in quantity (Christoffersen, Geczy, Musto and Reed, 2007), whereas others find that large positive shifts in the demand for share loans cause increases in lending fees (Cohen, Deither, and Malloy, 2006). Our results suggest that the supply schedule is relatively flat for most quantity levels, explaining why price and quantity seem unrelated most of the time. However, the schedule slopes upward for very high quantities, explaining the findings of Cohen et al. We will refine this analysis to identify kinks and other non-linearities in the supply schedule.

We also explore the role of search costs in the market, which are likely significant since there is no central clearinghouse for share loans. Drawing on models in the Industrial Organization literature that relate search costs, average price levels, and price dispersion, we confirm that search likely plays an important role in the share lending market: we document a positive relation between lending fee level and the cross-lender dispersion in lending fee for a particular stock. We plan to further test the importance of search by examining the relation between price dispersion and the cross-sectional distribution of the parameters of individual stocks’ share loan supply schedules, as well as how these parameters vary with numerous firm-specific variables.

The share lending market is becoming increasingly important as long-holders seek to augment returns, as short speculators and hedgers seek to establish their positions at lowest cost, and as interested parties seek to obtain votes to cast in close corporate elections. Accordingly, investment managers and sponsors will be interested in how prices in this market are determined.

Differences in Governance Practices between U.S. Foreign Firms: Measurement, Causes, and Consequences

Link to PDF File: Differences in Governance Practices between U.S. and Foreign Firms: Measurement, Causes, and Consequences
Reena Aggarwal, Georgetown University
Isil Erel , Ohio State University
I>René M. Stulz, Ohio State University
Rohan Williamson, Georgetown University

In this paper, we compare the governance of foreign firms to the governance of comparableuntry-region w:st=”on”> U.S.firms using propensity scores. We find that it is quite important, when comparing the governance of foreign firms and U.S. firms, to do so by comparing apples to apples, namely firms with similar characteristics. Comparisons based on country averages of firm-level governance indices understate the magnitude of the differences in investment in internal governance across countries because small firms, which typically invest less in internal governance, are over-weighted in the

We call the difference in governance between a foreign firm and its matching U.S. firm the governance gap. For the typical foreign firm, the governance gap is negative in that the foreign firm invests less in internal governance than its matching

U.S. firm. A foreign firm is much less likely to have a negative governance gap in a country with good investor protection, so that there is clear evidence that investment in internal governance and investor protection are complements rather than substitutes.

We find that the governance gap is strongly related to firm value. Firms that invest less in internal governance than their matching

U.S. firm are worth less and their value shortfall increases with their internal governance investment shortfall. We conclude that a firm’s under investment in governance compared to its matching

U.S. firm cannot be explained by unobserved firm characteristics that would make it optimal for the foreign firm to invest less in internal governance. Country characteristics play an extremely important role in explaining why the typical foreign firm invests less in internal governance than its matching

U.S. firm. However, neither investor protection nor other country characteristics completely explain the relation between a firm’s internal governance investment and its value. It is quite likely that firms typically under invest in internal governance because doing so is optimal for their controlling shareholder and sub optimal for their minority shareholders. An increase in a typical foreign firm’s investment in internal governance would make minority shareholders better off, but would not make its controlling shareholder better off. Further, in countries that place greater weight on the interests of stakeholders, an improvement in internal governance might also adversely affect these stakeholders.

As sponsors and their investment managers continue to seek investment returns abroad, issues of corporate governance in the international cross-section are of increasing importance, especially since governance often determines how investment opportunities and returns are split between public investors and others. The proposed research addresses a problem in the previous literature: too often comparisons of the effect of governance have not controlled for other important determinants of performance. The results should allow sponsors and their investment managers to better understand the risks that they take when investing abroad.


?>