Alumni Dissertations

 

Alumni Dissertations

Filter Dissertations By:

 
 
  • An empirical re-evaluation of the effects of education in nonmarket production

    Author:
    Christie Agioutanti
    Year of Dissertation:
    2009
    Program:
    Economics
    Advisor:
    Michael Grossman
    Abstract:

    Abstract AN EMPIRICAL RE-EVALUATION OF THE EFFECTS OF EDUCATION IN NONMARKET PRODUCTION by CHRISTIE AGIOUTANTI Adviser: Distinguished Professor Michael Grossman This paper tests the empirical validity of the productive efficiency hypothesis effect of education in the nonmarket sector via consumption expenditure patterns, which is essentially the foundation in Michael's (1972) behavioral model. Under the assumption of factor and commodity neutrality, the education effect is expected to be positive for luxuries, negative for necessities and zero for market goods inputs whose income elasticity is equal to one. Much of the innovation of this paper primarily stems from a preliminary and broad attempt to update Michael's empirical work, while focusing on the demand for inputs in the production of "good health", the incorporation of both grouped and individual observations on consumer expenditures household data as an additional comparison tool, inclusion of all four geographical regions and allowing for zero value market good outlays. By and large, the empirical analysis of this study fails to support the predicted qualitative relationship between income and education elasticities, regardless of sample classification and item categorization. Consequently, it rejects the accuracy of productive efficiency hypothesis in the neutrality framework. Whether these findings are the byproduct of the limitations set by the theoretical model, the type and idiosyncratic nature of the dataset used or by the econometric approach employed, the effect of education on household productivity remains a puzzle and a topic that requires more rigorous investigation. "The most important of all capital is that invested in human beings." Alfred Marshall

  • An empirical re-evaluation of the effects of education in nonmarket production

    Author:
    Christie Agioutanti
    Year of Dissertation:
    2009
    Program:
    Economics
    Advisor:
    Michael Grossman
    Abstract:

    Abstract AN EMPIRICAL RE-EVALUATION OF THE EFFECTS OF EDUCATION IN NONMARKET PRODUCTION by CHRISTIE AGIOUTANTI Adviser: Distinguished Professor Michael Grossman This paper tests the empirical validity of the productive efficiency hypothesis effect of education in the nonmarket sector via consumption expenditure patterns, which is essentially the foundation in Michael's (1972) behavioral model. Under the assumption of factor and commodity neutrality, the education effect is expected to be positive for luxuries, negative for necessities and zero for market goods inputs whose income elasticity is equal to one. Much of the innovation of this paper primarily stems from a preliminary and broad attempt to update Michael's empirical work, while focusing on the demand for inputs in the production of "good health", the incorporation of both grouped and individual observations on consumer expenditures household data as an additional comparison tool, inclusion of all four geographical regions and allowing for zero value market good outlays. By and large, the empirical analysis of this study fails to support the predicted qualitative relationship between income and education elasticities, regardless of sample classification and item categorization. Consequently, it rejects the accuracy of productive efficiency hypothesis in the neutrality framework. Whether these findings are the byproduct of the limitations set by the theoretical model, the type and idiosyncratic nature of the dataset used or by the econometric approach employed, the effect of education on household productivity remains a puzzle and a topic that requires more rigorous investigation. "The most important of all capital is that invested in human beings." Alfred Marshall

  • PRICING COLLATERALIZED DEBT OBLIGATIONS WITH PURE JUMP LÉVY PROCESSES: A DYNAMIC BOTTOM-UP APPROACH

    Author:
    Uluhan Basaga
    Year of Dissertation:
    2011
    Program:
    Economics
    Advisor:
    Liuren Wu
    Abstract:

    The Gaussian copula model is the industry standard in pricing CDO tranches because of its easy implementation and speedy calibration. However, it has several well-known shortcomings: It leads to the so-called correlation smile", generates symmetric and light-tailed asset return distributions and it is static. This dissertation proposes a dynamic bottom-up model based on a pure jump Lévy process, a path rarely taken in the credit pricing literature, and makes a comprehensive empirical analysis of bottom-up CDO pricing models. Owing to its ability to capture asymmetric heavy-tailed return distributions and to accommodate different degrees of dampening for positive and negative jumps, empirical evidence shows that the proposed model significantly outperforms the models commonly employed in the industry and frequently referenced in the literature in fitting CDX and iTraxx tranche spreads. As such, it constitutes an important addition to the credit pricing literature.

  • PRICING COLLATERALIZED DEBT OBLIGATIONS WITH PURE JUMP LÉVY PROCESSES: A DYNAMIC BOTTOM-UP APPROACH

    Author:
    Uluhan Basaga
    Year of Dissertation:
    2011
    Program:
    Economics
    Advisor:
    Liuren Wu
    Abstract:

    The Gaussian copula model is the industry standard in pricing CDO tranches because of its easy implementation and speedy calibration. However, it has several well-known shortcomings: It leads to the so-called correlation smile", generates symmetric and light-tailed asset return distributions and it is static. This dissertation proposes a dynamic bottom-up model based on a pure jump Lévy process, a path rarely taken in the credit pricing literature, and makes a comprehensive empirical analysis of bottom-up CDO pricing models. Owing to its ability to capture asymmetric heavy-tailed return distributions and to accommodate different degrees of dampening for positive and negative jumps, empirical evidence shows that the proposed model significantly outperforms the models commonly employed in the industry and frequently referenced in the literature in fitting CDX and iTraxx tranche spreads. As such, it constitutes an important addition to the credit pricing literature.

  • Sovereign Debt and Tax Collection Dynamics in Argentina

    Author:
    Kyle Bauser
    Year of Dissertation:
    2014
    Program:
    Economics
    Advisor:
    Merih Uctum
    Abstract:

    This manuscript examines how the dynamic macroeconomic effects from shocks to taxes and inflation differ between the United States and Argentina. On the fiscal side, wages, private capital, and consumption tax cuts have long-run revenue growth effects (in both countries) that mitigate initial tax receipt losses. These growth effects, however, are larger in Argentina - a country where both the consumption tax rate and sensitivity to wage changes are higher. Specifically, Chapter 2 finds that growth from a U.S. capital tax cut pays for roughly 60% of the initial static loss, whereas the corresponding effect in Argentina is 80%. On the monetary side, multiple regimes are then considered with money in the utility function to determine optimal scenarios, holding tax revenues constant. Chapter 3 concludes that distortions from taxes on wages, private capital, and inflation outweigh the efficiency losses from a consumption tax, and as such, an economy whose government places more emphasis on consumption to generate tax receipts achieves higher utility. The tax frameworks introduced in Chapters 2 and 3 build from the neoclassical Ramsey growth models. Inflation's role as a source of revenue via seigniorage in Chapter 3 is extended to the Argentine fixed income market in Chapter 4. Using proprietary pricing data and a structural vector autoregression framework, Chapter 4 finds that inflation as a predictor of the probability of default in Argentina is much larger than the government claims it to be; despite non-investment-grade government bonds, Argentina's fixed income market actually became more attractive during the U.S. mortgage crisis; and global risk aversion has predictive power in explaining sovereign spreads.

  • STOCK'S PRICE BEHAVIOUR AROUND CORPORATE MERGER AND ACQUISITION ANNOUNCEMENTS: EVIDENCE FROM THE NYSE

    Author:
    Nataliya Bershova
    Year of Dissertation:
    2010
    Program:
    Economics
    Advisor:
    Robert Schwartz
    Abstract:

    Efficient price discovery is one of the most important qualities of a financial market. Assessing a market's efficiency of price discovery is a challenging topic because efficient prices are not directly observable. It is not clear how to measure the extent to which actual prices conform to efficient prices. In this study, we use a methodology based on a state space model which deals naturally with short-term microstructure noise and enables the estimation of the unobservable prices and by implication, the pricing error. We investigate the contribution of non-instantaneous price discovery to intraday stock's price volatility around US corporate merger and acquisition announcements for a sample of 53 NYSE stocks from 2004 to 2008 using the TAQ data. Specifically, we model and estimate non-instantaneous price discovery effects that are associated with partial price adjustments to merger and acquisition (M&A) announcements for target names. We estimate price impacts from order imbalances that create pressure on a stock's price and cause a stock's price to move in a direction of order flow using one minute differencing intervals in the day of the merger news and the following day excluding the first 30 and the last 30 minutes of the trading session. Partial price adjustments may result in market over/under reaction to M&A news. To address this issue we also model and estimate market over/under reaction to merger news. We analyze the estimates of price impact and market over /under reaction to M&A announcements by contrasting the results in an M&A environment with the estimates observed in no-news days and before M&A news days. Our results shed light on price discovery around M&A announcements and its contribution to accentuated stock's price volatility. We find evidence of a protracted price discovery process following M&A news that takes at least two days after public merger announcements. Our findings reveal a significantly more informative period in two days following M&A announcements with the information flow per unit of time a factor 2 higher compared to the pre-announcement period and no-news days.

  • The Effects of Fiscal Policy on Savings and Investment

    Author:
    Andrew Bossie
    Year of Dissertation:
    2014
    Program:
    Economics
    Advisor:
    Michael Edelstein
    Abstract:

    This dissertation contains three essays exploring the effect of fiscal policy shocks on savings and lending behavior of the private economy. Two essays focus on World War II and one essay focuses on the entire postwar period. Essay I looks at the response of monetary policy to fiscal policy shocks and the effect of fiscal policy on the private sector's balance sheet over a period that covers 1954 and 2007. I find a minimal response of the Federal Reserve to fiscal policy shocks. I also find the the main response of the private sector to fiscal policy shocks manifests itself in household assets. Long term assets in particular react very strongly. Essay II establishes the important role of savings during and immediately after World War II. I show that the unexpectedly high savings rates that persisted after the war ended was driven by the fact that housing purchases are counted as a kind of savings. Treating housing as a durable consumption good produces the negative savings expected. Essay III looks at the behavior of commercial banks in the period 1940 to 1955. I find that there is a--both economic and statistically--significant negative effect of war spending on total assets, mortgage lending, and commercial, industrial and agricultural loans made by commercial banks throughout the war and until 1949. The response of bank assets during the immediate postwar period is indicative of the unusual pattern of output and the money supply between 1946 and 1950. All this is taken as evidence that reconversion was not as smooth and robust as commonly argued.

  • Essays in Market Efficiency

    Author:
    Juan Cabrera
    Year of Dissertation:
    2009
    Program:
    Economics
    Advisor:
    Tao Wang
    Abstract:

    This dissertation investigates the market inefficiencies of both foreign exchange and equity markets. On the one hand, the efficiency of foreign exchange markets is explored through the measurement of the contribution to price discovery of the spot and futures market, and the its effect on intermarket mispricing. On the other hand, the efficiency of equity markets is tested by examining the martingale behavior of recently popular international stock index ETFs. The first chapter provides a comprehensive analysis of the dynamic intraday price discovery process of the Euro and Japanese Yen exchange rates in three foreign exchange markets based on electronic trading systems: the Chicago Mercantile Exchange (CME) GLOBEX regular futures, E-mini futures, and the EBS interdealer spot market. Contrary to evidence in equity markets and more recent evidence in foreign exchange markets, the spot market is found to consistently lead the price discovery process for both currencies during the sample period. Furthermore, E-mini futures do not contribute more to the price discovery than the electronically traded regular futures. In the second chapter, we examine the daily return predictability for eighteen international stock index ETFs. Out-of-sample tests are conducted, based on linear and various popular nonlinear models and both statistical and economic criteria for model comparison. The main results show evidence of predictability for six of eighteen ETFs. A simple linear autoregression model, and a nonlinear-in-variance GARCH model, but not several popular nonlinear-in-mean models help outperform the martingale model. The allowance of data-snooping bias also substantially weakens otherwise apparently strong predictability. The final chapter investigates the relationship between the deviations of prices from their no-arbitrage value and the differences in informational efficiency across foreign exchange markets trading the same underlying asset. This relationship is examined by jointly modeling the dynamics of the futures-cash basis and information share differential across futures and cash markets. Evidence of two-way Granger causality between the no-arbitrage futures-cash basis and the relative speed of adjustment measure is found. Shocks to the no-arbitrage basis predict future differences in the speed of adjustment, and vice versa. The evidence is robust to different currency markets and different degrees of liquidity.

  • Essays in Market Efficiency

    Author:
    Juan Cabrera
    Year of Dissertation:
    2009
    Program:
    Economics
    Advisor:
    Tao Wang
    Abstract:

    This dissertation investigates the market inefficiencies of both foreign exchange and equity markets. On the one hand, the efficiency of foreign exchange markets is explored through the measurement of the contribution to price discovery of the spot and futures market, and the its effect on intermarket mispricing. On the other hand, the efficiency of equity markets is tested by examining the martingale behavior of recently popular international stock index ETFs. The first chapter provides a comprehensive analysis of the dynamic intraday price discovery process of the Euro and Japanese Yen exchange rates in three foreign exchange markets based on electronic trading systems: the Chicago Mercantile Exchange (CME) GLOBEX regular futures, E-mini futures, and the EBS interdealer spot market. Contrary to evidence in equity markets and more recent evidence in foreign exchange markets, the spot market is found to consistently lead the price discovery process for both currencies during the sample period. Furthermore, E-mini futures do not contribute more to the price discovery than the electronically traded regular futures. In the second chapter, we examine the daily return predictability for eighteen international stock index ETFs. Out-of-sample tests are conducted, based on linear and various popular nonlinear models and both statistical and economic criteria for model comparison. The main results show evidence of predictability for six of eighteen ETFs. A simple linear autoregression model, and a nonlinear-in-variance GARCH model, but not several popular nonlinear-in-mean models help outperform the martingale model. The allowance of data-snooping bias also substantially weakens otherwise apparently strong predictability. The final chapter investigates the relationship between the deviations of prices from their no-arbitrage value and the differences in informational efficiency across foreign exchange markets trading the same underlying asset. This relationship is examined by jointly modeling the dynamics of the futures-cash basis and information share differential across futures and cash markets. Evidence of two-way Granger causality between the no-arbitrage futures-cash basis and the relative speed of adjustment measure is found. Shocks to the no-arbitrage basis predict future differences in the speed of adjustment, and vice versa. The evidence is robust to different currency markets and different degrees of liquidity.

  • THE EFFECT OF FIRM-SPECIFIC RETURNS VARIATION ON R2: FROM THE PERSPECTIVE OF THE ACCRUAL ANOMALY

    Author:
    Wei-Hsin Chung
    Year of Dissertation:
    2009
    Program:
    Economics
    Advisor:
    Thom Thurston
    Abstract:

    R-square, calculated as CAPM of stock returns regressed on a market index, is constructed to explain stock price change by market-wide information. In my dissertation, I have analyzed the behavior of R-square and its decomposed variations (firm-specific and market-wide variation) with regard to the accrual anomaly, domestically and internationally. My major results in the US are as follows. First, the effect of accrual anomaly significantly lowering future R-square associated with higher firm-specific variations is robust and is not affected by (1) size, firm age, and other control variables; (2) industry risk and market risk; (3) other alternative explanations of accrual anomaly (value-glamour anomaly, bankruptcy risk, and arbitrage risk). Second, earning management is likely to be the reason why R-square decreases the accrual anomaly. Third, the difference in R-square between portfolios of good/poor accrual quality is subject to firm-specific variation. Fourth, the robustness of accrual anomaly significantly lowering future R-square even applies to the industry level. Using equity markets in 31 countries to investigate the relationship between the accrual anomaly and R-square, I find the occurrence of accrual anomaly significantly decreasing R-square is clustered largely in developed countries with relatively low R-square (USA, Australia, UK, Canada, Japan, Germany, Netherlands, South Africa, and France), casting doubt on the statement that the more efficient the capital market is, the lower R-square is. In terms of market efficiency, the actual R-square of developed countries should be higher. Finally, I observe that the phenomenon of accrual anomaly significantly lowering future R-square is more frequently found in countries with a low level of government corruption, those with a common law tradition, those with low quality of accounting standards, and those where shareholder ownership is widely dispersed. In addition, an analysis of ADRs (American Depository Receipts) helps to explain the effect of institutional and governmental structures on the relation between R-square and the accrual anomaly. Last, I also investigate the relation between stock return variation and extreme trading volume in the tail by demonstrating the asymmetry of the return and volume in six emerging countries.