<p> Economists, finance professionals, businesses, and governments analyze economic activity to forecast recessions to put in place plans that will minimize the impacts of a recession. A tool that could identify how long a recession might last could reduce the unwanted effects of a recession. While researchers have identified that the interest rate term spread, a component of the yield curve, can effectively predict if a recession is probable and approximately when it might begin, it is not known whether the components of the interest rate yield curve might also be able to predict how long a recession might last. The purpose of this quantitative correlational study was to analyze the components of the yield curve to determine if they could be used to forecast how long a recession might last or when a recession might end. Data for the study was gathered for the U.S. economy for the period from 1942 to 2012. The data set for the entire time period and for each of the twelve recessionary periods was analyzed using the autodistributive lag and vector autoregression models. It was found that there was a statistically significant correlational relationship for two yield curve components, the interest rate term spread and the three-month T-bill rate, that could be used to forecast when a recession might end. Prior to these recessions the interest rate term spread drops rapidly and typically turns negative. The term spread began to increase during the recession and for the twelve recessions analyzed rose to the level of 100 basis points on average about seven months before the end of the recession. It was also found that the three-month T-bill rate rose significantly before the recessions, peaked, and then fell throughout the recessionary time period. For the recessionary periods analyzed the peak in the T-bill rate occurred on average about eleven months before the end of the recession.</p>

We examine a continuous-time structural model of debt valuation with the possibility of default and Chapter 11 bankruptcy. In doing so, we derive Chapter 11 duration and allocations to the debtor and bondholders in Chapter 1 I as the outcomes of a bargaining game between the debtor and the bondholders. The absolute priority rule (APR) violations arising in equilibrium are then embedded into closed-form solutions for the values of equity, finite-maturity debt, and credit spreads. It has been recently documented that existing credit risk models explain only a fraction of the observed yield spreads when confronted with the data on default rate and recovery rate at default (e.g., Collin-Dufresne et al. (2001) and Elton et al. (2001)). Taking the exclusivity period as an approximation to a legal environment of the bargaining process, we model Chapter 11 as the debtor's ultimatum offers to the bondholders and calibrate the model using an approach similar to that of Huang and Huang (2002). We obtain credit spreads that are twice to three times as large as those produced by the model in Leland and Toft (1996). The reason why this result holds in our model is that when the debtor obtains a non-zero allocation in bankruptcy, her option to default is worth more and exercised sooner than in Leland and Toft's model. Therefore, the debt value is smaller, and consequently, credit spreads are higher. Calibrated credit spreads are high for firms expected to be more solvent at default and those with large absolute priority rule violations. Finally, our model predicts a significant cross-sectional variation in Chapter 11 duration. Indeed, such heterogeneity is seen in actual bankruptcy experiences. We discuss several new empirical implications of the model with regards to the expected time in bankruptcy as a function of different firm characteristics. The model predicts that firms with a higher fraction of intangible assets, lower pre-bankruptcy volatility of asset value, and lower average maturity of debt in their capital structure spend less time in Chapter 11.

Researchers have continually questioned the economic viability of the Central Arizona Project (CAP). The completion of CAP in 1993 triggered the repayment obligation to the federal government. Despite enormous federal subsidies and the existence of several revenue sources other than from water sales the annual repayment obligation caused CAP's total expenses to exceed total costs for the two years following 1993. Using various combinations of repayment terms this study solves for pricing schedules that generate annual revenues equal to annual costs. Using estimates of water demand and water supply, price schedules are determined by maximizing consumer surplus subject to a revenue constraint. Under the current repayment terms the initial increases in water prices are less than twenty percent. After the first year increase, prices decline in real terms over the 50 year repayment horizon. In some cases future CAP prices actually drop below current water prices in real terms. Using alternative repayment terms that partially remove large federal subsidies results in first year percentage increases of up to fifty-two percent. In summary, the massive federal subsidies inherent in CAP keep the second-best prices from becoming exorbitantly high. The relatively small impact to the individual water user from the initial increase in prices suggests that CAP will likely remain financially viable. Based on the decreasing trends exhibited by the second-best prices it is possible that CAP may someday generate significant excess revenues to be used to further develop, improve and maintain Arizona's water resources.

The Random Walk Hypothesis (RWH) when applied to stock prices makes strong statements about such things as serial correlation and momentum in stock returns and about the information carried in the path of the stock price. Its significance to Finance has motivated many tests of the RWH. This paper, using data from 1963-1997, tests the RWH and then evaluates those tests using simulated data. We find that the data reject the RWH and that rejections are stronger using NASDAQ firms than they are with NYSE/AMEX firms. The time series properties of the rejection of the RWH suggest that recently prices more closely conform to the properties of a random walk than they did in the more distant past. The simulation results show that if the true underlying price process is a random walk, market imperfections such as price discreteness are sufficient to reject the random walk hypothesis. The time series properties of the rejection of the RWH are consistent with a narrowing of the bid ask spread through time. Covered call strategies are touted as win-win strategies. If share prices rise strongly then shares are sold and the gain is locked in. If share prices are flat or fall then the premiums from the sale of the call options act as additional income to supplement the poor performance of the stocks. The last two chapters in this paper compare covered call strategies to a simple buy and hold strategy, using data from 1989-1998. The results show that the buy and hold strategy out-performs the covered call strategy over this period even after adjusting for systematic risk and co-skewness with the market. Empirical comparisons of buy and hold and covered call strategies are period dependent and difficult to generalize as market performance plays a key role. Bootstrapping and simulation provide control over market returns. Both simulated and empirical analyses consider transactions costs and taxes as well as account for the different risk assumed under each strategy.

This dissertation studies intraday and daily foreign exchange market volatility. First, we address how best to model the intraday seasonality and the serial correlation in return volatility. We find there is no gain from smoothing the intraday seasonal volatility pattern. A model that jointly estimates the intraday seasonal pattern and conditional heteroskedasticity underperforms models that remove seasonal variance through deseasonalization and then model conditional heteroskedasticity with a GARCH model. Secondly, we show how intraday data can be used to create daily volatility estimates. Results show intraday data allow for daily volatility estimates which are independent of a volatility dynamics specification. Lastly, we show that intraday data improve the performance of one-step ahead forecasts based on a one year sample and show that the results are consistent with Monte Carlo simulations.

This paper explores the relationship between strategic trading and the clustering of volatility commonly found in financial asset returns. The aim is to provide a theoretical foundation for the volatility clustering which has been documented in the econometric literature of ARCH and stochastic volatility models. The analysis builds on the notion that, while exogenously occurring events generate new information, investor activity ultimately determines how quickly information becomes incorporated into prices. In the model developed, a single investor obtains superior information about the implications of news events for an asset's value and is able to exploit his informational advantage over time by trading with a market maker. The model extends the market microstructure literature by allowing the trader to receive a flow of information through time and by making uncertain the length of time the trader's information remains private. Results show how an investor's trading behavior can influence the arrival of information into the market and cause above average volatility to cluster and persist, even when the arrival of events does not. A key finding is that, while the level of volatility is sensitive to the rate at which private information flows, its persistence is driven by the nature of the trader's uncertainty. The model and its extensions are tested in a computer laboratory setting with human subjects, which permits control of the underlying distributions of values and events. In testing the model's hypotheses experimentally, two underlying issues are addressed: whether trading will tend toward the theoretical dynamic equilibrium and whether agents' preferences are suitably modeled as risk neutral. Analysis of the experimental data reveals that market makers' pricing was more sensitive, informed trading was more cautious, and price volatility was higher than the risk neutral model predicts. Estimation of an alternative model specification reveals that behavior was consistent with risk aversion.

Firms can choose to restructure by using a carve-out, sell-off, spin-off, or exchange, or by issuing tracking stock. The purpose of this study is to determine how firms choose among these methods when restructuring to "unlock value." Using a sample of Compustat firms during the period from 1990-1999, I estimate the joint likelihood that a firm will restructure and the likelihood it will raise external capital. I find that firms are more likely to either restructure only or restructure and raise capital as the percentage of unrelated segments increases, as the abnormal industry sales decrease, as firm size increases, and as the firm's current ratio decreases. In addition, the likelihood of restructuring and raising capital increases as the firms leverage ratio increases. The probability that the firm will only restructures is negatively related to the firm's growth rate and return on assets, and positively related to the number of segments and the firm's free cash flows. Finally, I find evidence that the restructuring and need for capital decisions should be considered simultaneously. Conditional on the estimated probabilities, I estimate logit equations to determine how restructuring choices are made. I find that the likelihood of choosing a spin-off or issuing tracking stock rather than an exchange increases when the segment is unrelated and as the probability of only restructuring decreases. This result is consistent with firms having the capacity to use means other than restructuring to raise cash. As the probability of restructuring only decreases, exchanges, tracking stock issues, or sell-offs are more likely to occur than carve-outs. In addition, segments that experience growth in sales are more likely to be carved-out than exchanged or sold-off. Firms are more likely choose a sell-off rather than issue tracking stock, or exchange or carve-out a segment when leverage ratios are increasing, the firm is growing, and the probability of restructuring only decreases. This finding is consistent with the notion that firms look to sell-off assets as a way to raise cash to pay off existing debt or because additional debt issues are too costly.

This dissertation is a collection of three essays related to the initial public offerings (IPOs) in Malaysia. The IPO market in Malaysia has historically been tightly regulated. However, with the formation of the Securities Commission (SC) in 1993, the regulations have eased slowly. The first essay investigates the relationship between underpricing and regulations by looking at 546 IPOs from 1990 to 2002. Underpricing refers to the initial return that an investor earns if he buys shares of the IPO at the offer price and sells it at the end of the listing day at the market price. Regulations are measured by the relaxation of the pricing method, the required allocation to indigenous investors, the mechanisms to protect minority shareholders, and the length of time periods. The first three features of regulations are unique to Malaysia. The findings are mixed regarding the relaxation of the pricing guideline in 1995 since it does not lead to lower underpricing for the period from 1996 to November 7, 1997 or before the Asian financial crisis. Fraction of shares set aside for indigenous investors does not affect underpricing; length of time from price setting to listing date relates negatively to underpricing. Finally, the protective mechanisms lead to more underpricing for firms that go public between 1996 and November 6, 1997 or those that go public after 1998, i.e., after the Asian financial crisis. The second essay looks at the relationship between the universal banking facility and the costs of going public for 546 initial public offerings listed on the Kuala Lumpur Stock Exchange from January 1990 to December 2002. It is hypothesized that by sharing private information about a firm that is going public with its affiliated commercial bank, an investment bank could lower the costs of going public for the firm. Costs of going public are measured by the degree of underpricing and gross spread, or the fees paid to the underwriters as a fraction of gross proceeds. The result in this essay is that, for the period under study, firms do not reduce the costs of going public when they use the universal banking facility. The third essay looks at the three-year performance of the IPOs in Malaysia from 1994 to 2000. Evidence from most studies in different countries finds that initial public offerings (IPOs) underperform their benchmarks or matches in the long run. However, our evidence regarding the long run performance of IPOs in Malaysia is that the IPOs do not underperform their matches. The returns of the IPOs are adjusted by using either a market index or firms with similar characteristics to the IPOs. Two different matching estimators are employed to identify the firms of similar characteristics. One of the contributions of this essay is the use of a new methodology to identify the matches for the IPOs. (Abstract shortened by UMI.)

<p> Technical analysis, namely the moving average rule and the channel rule, is applied to the currency of an Asian managed floating exchange rate regime (USD/SGD) to see if opportunities for profitable trading exist. Instead of using only daily or monthly data, higher frequency time frames of 10, 15, 30 and 60 minutes are analyzed. Profitable strategies (if any) will be broken down and analyzed within smaller time frames to see if the profits are specifically in sample.</p>

This thesis comprises of three essays on pricing credit and commodity derivatives. The first essay estimates recovery rates from CDS spreads using three years of daily data on 152 corporates. We use a quadratic pricing model which ensures nonnegative default probabilities and recovery rates. The estimated cross-section of recovery rates is plausible, with an average recovery rate of 54%, and substantial cross-sectional variation. Estimated five-year default probabilities are on average 67% higher than default probabilities obtained using the standard 40% recovery assumption. This finding critically impacts the valuation of structured credit products. Larger firms and firms with more tangible assets have higher recovery rates.In the second essay, we find that the state price densities (SPDs) implicit in the crude oil market display a time varying U-shape pattern. This implies that investors assign high state prices to both negative and positive returns. We use data of the crude oil market, where speculation and short sales are not regulated, to document how the SPDs are dependent on investor beliefs. Investors assign higher state prices to negative returns when there are more net short positions, higher dispersion of beliefs in the futures market, and higher demand for out-of-the-money put options. The increase in speculation after 2004 reinforces this effect.The third essay investigates the variance risk premia in the crude oil market analyzing horizons beyond one month. The existing literature restricts attention to one-month maturity, even though commodity hedging plans very often have much longer horizons. Furthermore, we provide the first futures return predictability study in the literature using the variance risk premia as complementary information variables. As suggested by theory, storage level and hedging pressure are determinants of futures returns. Importantly we find that the variance risk premia significantly increase the predictability of short maturity futures returns. This finding is robust across various implementations of predictors and time periods. We provide economic interpretations for the predictability of futures returns using the variance risk premia in a simple economic setting with heterogeneous agents. / Cette thèse comporte trois essais portant sur la tarification des dérivés de crédit ainsi que les dérivés sur commodités. Dans le premier essai, nous estimons le taux de recouvrement à partir des CDS (Credit Default Swap) en utilisant des données portant sur 152 entreprises et couvrant une période de 3 ans. Nous utilisons un modèle quadratique qui permet d'assurer la non-négativité des probabilités de défaut et des taux de recouvrement. Les résultats montrent que les taux de recouvrement estimés sont plausibles avec une moyenne de 54% et varient considérablement d'une entreprise à une autre. Les probabilités de défaut pour un horizon de 5 ans sont, en moyenne, 67% plus élevées que celles obtenues avec un taux recouvrement de 40% communément utilisé. Ce résultat a un impact important sur la tarification les produits dérivés et structurés de crédit. Les grandes entreprises et les entreprises qui possèdent beaucoup d'actifs tangibles ont un taux de recouvrement plus élevé.Dans le second essai, nous montrons que le facteur d'actualisation stochastique (stochastic discount factor) implicite dans le prix pétrole brut présente une forme quadratique. Ce résultat implique que les investisseurs affectent un facteur d'actualisation stochastique plus élevé aussi bien pour les rendements négatifs que positifs. En utilisant des données sur le marché du pétrole brut où les spéculations et les ventes à découvert ne sont pas réglementées, nous montrons que le facteur d'actualisation stochastique dépend des opinions des investisseurs. Les investisseurs affectent un facteur d'actualisation stochastique plus élevé pour les rendements négatifs quand il y a plus de positions nettes courtes, une plus grande dispersion des opinions des investisseurs dans le marché des contrats à terme et une demande élevée pour les options put «out-of-the-money». Cet effet est plus prononcé après 2004, dû à une augmentation des spéculations.Dans le troisième essai nous étudions la prime de risque de la variance dans le marché du pétrole brut en considérant des horizons dépassant un mois. Les travaux existants se concentrent principalement sur un horizon d'un mois, bien que les besoins de couverture dans le marché de commodité dépassent l'horizon d'un mois. Nous étudions aussi la capacité de la prime de risque de la variance à prédire les rendements dans le marché du pétrole brut. La théorie stipule que le niveau de stockage and les besoins de couverture déterminent les rendements futurs. Nous trouvons que la prime de risque de la variance prédit les rendements futurs à court terme. Ce résultat est robuste à plusieurs tests de robustesse. En se basant sur un modèle économique avec agents hétérogènes, nous fournissons quelques interprétations économiques pour expliquer la capacité de la prime de risque de la variance à prédire les rendements dans le marché du Pétrole brut.