Finance - Theses
Permanent URI for this collection
Search Results
1 - 10 of 37
-
ItemNew Approaches of Utilizing Firm Characteristics in Empirical Asset Pricing( 2022)Firm characteristics carry very important information about stock prices and can be widely used in various aspects of asset pricing. This thesis contains 3 new applications of firm characteristics combined with some new technologies, which brings new understandings to several important questions: In chapter 3, my co-authors and I apply the General Empirical Likelihood (GEL) estimator to ask and answer the question of, which characteristic-sorted portfolios provide valuable information about the stochastic discount factor (SDF). We estimate a non-parametric SDF from a set of portfolios, then test whether excluding a portfolio changes the implied SDF. Though related to traditional asset pricing tests, our approach has several advantages: we test all portfolios jointly and can incorporate trading costs easily. We show four portfolios provide independent information about the SDF after accounting for trading costs: the Market and Profitability factors, an Investment-based portfolio, and the Value-Momentum-Profitability anomaly portfolio. The remaining portolios are redundant. We show both the joint testing and transaction cost adjustments are important for inference, and provide a simple way to implement our tests. In chapter 4, I introduce a Machine Learning (ML) based approach to construct comparable groups that researchers often use to compute the abnormal part of stock returns. The characteristics shown in chapter 3 to be important to asset pricing are included in this study as candidate characteristics. In order for stock expected returns to be similar within groups and disperse across groups, I use the ML based approach to select characteristics that best distinguish expected returns, and cutoffs points where returns are most sensitive to the underlying characteristics. I show that: 1) the combination of chosen characteristics changes over time; 2) fewer fund managers are identified to be stock pickers once the time-variation in comparable groups is incorporated; 3) and the resulting portfolios exhibit desirable properties as basis assets. In chapter 5, I adopt the ML based comparable groups from chapter 4 to analyze the market timing and stock picking skills of actively managed mutual funds. Compared to traditional studies which are prone to model mis-specification due to the difficulties in handling large dimensions and non-linearity, my approach can extract information about returns from a large number of characteristics, and allows for complex and time-varying relations between stock returns and firm characteristics. To the contrary of the consensus in prior literature, I find strong evidence in support of fund timing skills. On average, funds exhibit significantly positive timing performance. Cross-sectionally, the best timers continue to outperform others for at least three years after the ranking period. There is some picking return for the average fund but out-performance in picking does not persist. These results have real-world implications. I further show that investors can use my timing measure to identify funds with high future risk-adjusted performance.
-
ItemComputational complexity of decisions: Quantifying computational hardness and its effects on human computation( 2021)Humans are presented daily with decisions that require solving complex problems. In many cases, solving these problems is computationally hard. This raises a tension between the computational capacity of the agent and the computational requirements of a task. Whilst the underlying invariants of this mechanism remain unclear in cognition, they have been widely studied in computer science. I build on theoretical and empirical work in computational complexity, which characterizes the intrinsic computational hardness of problems. I first present an adaptation of this theoretical framework for the study of human cognition by introducing a set of metrics of hardness of instances of problems. I do this in a way that is independent of any algorithm or computational model and that can be generalized to other problems. Based on this, I explore empirically, in a set of lab experiments, how these task-independent metrics of hardness affect human problem-solving. I do this at two levels of analysis. Firstly, I study how these metrics affect human performance at the behavioral level in three canonical computational problems: the knapsack problem, the traveling salesperson problem and the Boolean satisfiability problem. Secondly, I examine the relation between computational hardness and the neural processes associated with problem-solving, employing ultra-high field functional MRI. I find that the metrics of intrinsic hardness put forward here predict human performance and time-on-task across the three computational problems in a similar way. Moreover, I identify the neural correlates of computational hardness in the knapsack task, a complex problem-solving task. I show that this framework can be used for the study of the neural underpinnings of problem-solving by providing a generic definition of cognitive demand. The results of these studies provide support for the conceptual premise that the quantification of intrinsic hardness is fundamental in the development of more refined theories of human decision-making and its neural underpinnings. Critically, they provide a framework to study how humans adapt to computational complexity and how intrinsic hardness of tasks affect the reliability of human decision-making. This could inform public policy by identifying which decisions over products involve solving problems that require computational resources beyond those available to an agent, and how this affects decisions.
-
ItemEssays on Political Economy of Finance( 2021)This thesis explores mechanisms that connect firms to politicians, as well as new channels through which firms benefit from being politically connected, under different ecopolitical environments. It contains three essays examining various aspects of finance at its intersection with political science. The first essay exploits Donald Trump’s nonpolitical background and surprise election victory to identify the value of sudden presidential ties among S&P 500 firms. In our setting firms did not choose to become politically connected, so we identify treatment effects comparatively free of selection bias prevalent in this literature. Firms with presidential ties enjoyed greater abnormal returns around the 2016 election. Since Trump’s inauguration, connected firms had better performance, received more government contracts, and were less subject to unfavorable regulatory actions. We rule out a number of confounding factors, including industry designation, sensitivity to Republican platforms, campaign finance, and lobbying expenditures. The second essay finds that borrowers from the same state as the Chairman of the US Senate Banking Committee, whom I term "connected", are able to borrow at spreads 14 basis points lower than other borrowers. Connected borrowers’ contributions toward the Chairman are influenced by their cost of loans, but the same is not true for nonconnected borrowers. Findings suggest the Chairman is incentivized by reelection to actively help connected borrowers obtain cheaper loans. Banks that offer a larger fraction of connected loans enjoy higher future excess stock returns. Results are consistent with the existence of a quid pro quo relationship triangle between firms, banks, and politicians. The third essay examines how changes in local political leadership affect firms’ gover- nance structures. Using a novel dataset, I document that following the appointment of a new city-level Chinese Communist Party (CCP) secretary, local firms increase (decrease) the fraction of directors who share a common birthplace with the incoming (departing) secretary. This appears to be a channel through which Chinese firms establish political connections. Firms with a higher percentage of birthplace-connected directors exhibit higher abnormal returns around secretary appointments. These firms enjoy superior accounting performances and attract institutional fund flows. I reject an alternative hypothesis that these directors are appointed to company boards on the "orders" of the politician, rather than actively recruited by firms.
-
ItemThree Essays on CEO Compensation and the Agency Problem of Debt( 2020)Jensen and Meckling [1976] propose that compensating a manager with debt-like payment, a.k.a. “inside debt”, can be a mechanism to mitigate the agency problem of debt. Recent empirical papers document that higher inside debt held by a chief executive officer (“CEO”), which is measured by the present value of their supplemental retirement benefits, is associated with more conservative management and lower cost of debt. However, the grant of inside debt might originate from rent extraction purposes, and happens to encourage risk aversion among CEOs. Therefore, it remains unclear whether companies use CEO inside debt as a mechanism to mitigate the agency problem of debt. To address this question, my thesis examines how U.S. public companies adjust their use of CEO inside debt in three corporate events that involve changes to the agency problem of debt. In the first essay (Chapter 4), I show that new active blockholders adjust CEO inside debt-equity ratios to increase total firm value, not just equity value. These investors are more like to arise when CEO inside debt-equity ratios are not properly set up to maximize total firm value. The speed of adjustments towards the appropriate ratios triples in the presence of active blockholders, but returns to the normal level once these investors “exit”. Such compensation adjustments are associated with positive stock and bond abnormal returns over the active block holding period. I also find that new active blockholders arise and restructure compensation when the old structures over-align CEOs’ incentives to either shareholders’ or debtholders’ interests. I argue that superb stock-picking skill, a mean-reverting process of compensation changes, or co-founding firm characteristics cannot explain the large compensation adjustments during active blockholders’ presence. Instead, these blockholders actively affect CEO compensation structures by appointing their favored directors into the targeted firms’ boards, especially into the compensation or governance committees. In the second essay (Chapter 5), I show that companies raise their CEO inside debt to address the heightened agency problem of debt due to increased leverage, after they remove anti-takeover provisions (“ATPs”). By using a difference-in-difference-in-difference analysis, I document significant increases in CEO inside debt-equity ratios after companies remove ATPs. Inside debt also rises significantly after ATP removals, which accounts for 70% of increases in the overall ratios. In contrast, inside equity significantly decreases after companies remove ATPs, as these companies reduce the stocks and options awarded to their CEOs. These findings are robust to different explanatory variables, matching samples, and removals of certain ATPs. Further analysis displays that inside debt-equity ratios increase continually for the first three years after ATP removals. This trend coincides with the after-ATP-removal spike in leverage, which is caused by increased debt issuance. I also show that increasing inside debt-equity ratios, especially increasing inside debt, helps companies reduce the cost of borrowing after ATP removals. However, increasing inside debt can only address part of the heightened agency problem of debt, as companies still face more costly debt financing after ATP removals. Despite the increase in inside debt and decrease in inside equity, CEOs take more risks after their companies remove ATPs. This result suggests that removing ATPs can substitute for the role of inside equity in providing CEOs with more risk-taking incentives. In the third essay (Chapter 6), I explore the timing strategies for bond issuances based on disclosures of CEO inside debt. During the months before proxy statement releases, new changes in CEO inside debt are private information to companies’ insiders. Companies can exploit this information asymmetry to issue bonds when the market, based on publicly available information of inside debt, perceives these companies’ debt agency problems as relatively insignificant. I find that companies cluster their bond offerings in the immediate quarters after (before) disclosures of positive (negative) inside debt changes. The tendencies to time bond issuances based on inside debt disclosures also increase with the magnitudes of the disclosed changes. In addition, the adoption of these timing strategies is more observable when the issuing firms are regular issuers or when the new issues do not include covenants, especially the debt-restriction covenants. Finally, I verify that these timing strategies help reduce the cost of borrowing. The bonds issued at favorable times, i.e. right after positive change disclosures or before negative change disclosures, have lower offering yield spreads than those issued at non-favorable times.
-
ItemEssays on Perfect Foresight in Asset Pricing( 2020)Through experimental and theoretical analysis, this thesis addresses the question of how the context of information in capital markets can affect the occurrence of perfect foresight equilibria. It contains three essays that build on theoretical and experimental applications of the perfect foresight assumption. The first essay contrasts theoretical and experimental strains of information aggregation – the ability of prices to aggregate disparate pieces. It contains a methodological guide for designing robust information aggregation experiments with a detailed description of the pilot studies that were used to develop the experimental study in the second essay. The second essay introduces two new theoretical concepts to the analysis of rational expectations equilibrium models. These concepts stem from “stability” characteristics inherent to the perfect foresight state-price mapping. Differences in stability characteristics are shown to arise from differences in the initial information structure underlying the aggregation problems. In the experiment, we test information aggregation with two fundamental information structures in continuous time double auction asset markets: The first information structure is motivated by the canonical information aggregation model in theoretical asset pricing. In this setting, the asset traded pays according to the average privately held information signal in the market. This setting has a stable state-price mapping and is shown to aggregate information well. The second information structure is motivated by prediction markets and studies in experimental finance. Both feature winner-take-all contracts where binary pay depends on the signal type held by the majority of agents. This setting is unstable under the theoretical stability concepts and is shown to aggregate information less efficiently. The third essay examines the use of perfect foresight when modelling disagreement in financial markets. In particular, we examine the conditions under which the perfect foresight approach can be used in a rational expectations equilibrium model. We show that an agent’s perfect foresight may be inconsistent with their own beliefs (based on subjective probabilities) unless their higher-order beliefs (about other participants’ beliefs) are correct.
-
ItemTime-varying efficiency of internal capital markets( 2019)This thesis examines the time-varying efficiency of diversified firms. The first essay documents flipped Q-sensitivity of investment of conglomerates over the business cycle. I find that the internal capital markets affect investment and cash retention differently from stand-alone firms. The result is robust to the degree of diversification, matching sample and difference-in-differences analyses. The second essay examines the effect of compensation incentives on intra-firm capital allocation, of both CEOs and divisional managers. I provide novel evidence that reconciles the efficiency explanation with the theory prediction that managers require extra incentives to relinquish capital in control and finance investment.
-
ItemCorporate bond market clientele( 2018)This thesis consists of two essays that explore research questions related to corporate bond market clientele. In the first essay, I examine the investment behaviour of corporate bond mutual funds, the fastest-growing clientele for corporate bonds in recent times. Specifically, I study the extent to which the funds’ investment decisions are driven by their investment mandates. In the second essay, I draw on anecdotal and empirical evidence that shows that clientele in corporate bond markets are fundamentally different from clientele in equity markets. I apply these insights to examine how corporate policy is affected by differences between corporate bond and equity investors’ pricing of risk. Theoretical and anecdotal evidence suggests that mandates are a key determinant of mutual fund investment behaviour. However, other evidence reveals that mandates are not legally binding, and funds may have incentives to deviate. In the first essay, I compare the investment expenditure of funds with different mandates within segments of the corporate bond market. The segments categorize bonds according to maturity or credit rating. I find that, in each segment, the funds that spend relatively more (less) are those whose mandates imply that they should invest more (less) in that segment. For example, if a fund states that it targets a longer portfolio maturity, the fund invests significantly more in long-term bonds, on average, relative to funds that target shorter portfolio maturities. These results apply not only to average investment expenditure, but also to how each incremental unit of flow is allocated across segments. Results hold in subsamples of secondary market bonds, as well as new issues, regardless of whether the issuer of a new bond is a firm in which the fund already invests. Overall, my findings suggest that the investment behaviour of corporate bond mutual funds is consistent with the funds’ mandates. In the second essay, I use a dynamic investment and financing model that accounts for differences between equity and corporate bond holders’ pricing of macroeconomic risk. In line with empirical and anecdotal evidence, I calibrate the bond investor’s price of risk to be unconditionally higher than the equity investor’s, as well as volatile and independent of the macroeconomy. Relative to a counterfactual scenario in which both investors price risk identically, average market (book) leverage is 2.8 (3.3) percentage points lower. These results reveal a new quantitatively significant channel to address the under-leverage puzzle. Also, in the scenario with heterogeneous risk pricing, firms issue equity more frequently as they substitute away from debt financing. Overall, relative to the scenario with homogenous risk pricing, firms exhibit a similar, albeit slightly lower, average rate of investment.
-
ItemEssays on naive diversification( 2018)The mean-variance model pioneered by Nobel laureate Harry Markowitz is the foundation of modern portfolio theory, and is widely applied in asset allocation and active portfolio management. However, the naive 1/N diversification rule (the equal-weight portfolio rule) has received much academic attention because of its superior performance relative to mean-variance portfolio rules. This thesis consists of two essays with respect to the naive 1/N diversification rule. The first essay examines the sample selection bias in portfolio horse races with the 1/N rule. Numerous studies have developed mean-variance portfolio rules to outperform the naive 1/N rule. However, the outperformance is often justified with a small number of pre-selected datasets. Using a novel performance test based on a large number of datasets, I compare various ``1/N outperformers" with the naive rule. The results show that although some ``1/N outperformers" outperform the 1/N rule on an average basis, a sample selection problem generally exists in claiming significant outperformance over the naive benchmark. To further understand portfolio performance, I explore the theoretical relations between assets' return moments and the performance of optimal versus naive diversification. These relations not only imply strong performance predictability, but also can be exploited to deliver out-of-sample portfolio benefits. The second essay studies how January seasonality affects the relative performance of the mean-variance and 1/N rules. I find that the good performance of the 1/N rule does not depend on the January seasonality when value-weighted indexes are used for portfolio construction. However, when individual stocks are formed into portfolios, a large proportion of the empirical success of the 1/N rule is attributed to enormous January returns. Mean-variance portfolio rules fail to outperform the naive rule on a risk-adjusted basis; however, the relative performance reverses when January returns are excluded. The results are robust with and without transaction costs, for various mean-variance rules, and across different industries and characteristics.
-
ItemLockup Agreements during Equity Issuance( 2017)Information in the equity issuance market is highly asymmetric. Issuers have information advantages over investors and underwriters. Under asymmetric information, in the U.S., insiders from issuing companies and the underwriters voluntarily negotiate lockup agreements before the issuance of equity. Lockup agreements restrict insiders from selling their shares during the lockup period. However, underwriters have the right to release some or all of the locked-up shares, allowing insiders to sell their shares early at any time before the lockup expiration. Early sales refer to these insider sales during lockup periods. Lockups commonly exist in both initial public offerings (IPOs) and seasoned equity offerings (SEOs). This thesis investigates both the underwriters' incentives for early releases of locked-up shares during the IPO and the impact of IPO lockups on the decision to include SEO lockups. First, I study underwriters' incentives for early releases during an IPO. Ten percent of IPOs with lockup agreements have early sales by top executives. Early sales reduce the likelihood that IPO companies switch lead underwriters in their subsequent SEOs. IPO companies with early sales have better post-IPO performance than their counterparts without early sales. I argue that early sales reduce the signaling cost incurred by IPO lockups under asymmetric information. As information resolves after the IPO, good companies exercise early sales and directly benefit from the reduction in the signaling cost, while underwriters benefit from an increase in future business. Second, I examine the relation between IPO and SEO lockups. I find that underwriters are more likely to impose SEO lockups on issuers that have IPO lockups. I focus on a sample of issuers that conduct their first SEOs within four years after the IPO. I attribute the positive relation between SEO and IPO lockups partially to high correlations between company characteristics at the times of the IPO and the SEO. However, the commitment level of insiders in the issuing company does not offer an explanation for the positive relation between IPO and SEO lockups. Rather, the positive share price response to the announcement of the change from including lockups at the IPO to waiving lockups at the SEO implies that this change by underwriters conveys good news to the market, consistent with SEO lockups helping to reduce the information asymmetry in the equity issuance market.
-
ItemThe psychophysiology of time perception and temporal decision making( 2017)Time is a fundamental dimension of our perception of the world and therefore of critical importance to the organisation of human behaviour. Without the ability to perceive time we would not be able to navigate the world in an effective way. We would not be able to perceive the causality between events, nor would we be able to account for the future. However, despite its significance, our perception of time is not veridical, and appears to be labile to many external and internal factors. (A pertinent example is the apothegm “time flies when you’re having fun”, which reflects the temporal distortions we often experience during rewarding periods.) This thesis was aimed at characterising a novel source of volatility in time perception — the effect of reward consumption — and to assess the implications of non-veridical time perception for human decision making. The first study sought to characterise the effect of different primary rewards on interval timing. Participants performed a novel variant of a duration production paradigm, while they received different volumes of fruit juice on a trial-by-trial basis. The consumption of fruit juice lead to systematic overproductions of time (from 2-5 seconds) and this effect scaled with the volume of the consumed juice. Another four reward types were subsequently tested: money (a secondary reward), water (a tasteless, noncaloric reward), aspartame (a sweet, noncaloric reward), and maltodextrin (a tasteless, caloric reward). Maltodextrin also produced a similar effect on time productions. This pattern of results suggested that the observed effect was likely to be due to the common caloric content of both fruit juice and maltodextrin. In sum, the first study demonstrated a novel association between biologically relevant rewards and time perception. The second component of this thesis investigated the proposition that temporal decision making (i.e. decision making that involves time) operates on subjective representations of time. To explore this proposition, the second study investigated whether there was a relationship between individuals’ time perception and their temporal decision making. Participants performed both a temporal reproduction task and a temporal discounting task, while undergoing electrocardiography. The results provided no evidence that parameters from the temporal reproduction task were correlated with discount rates in the temporal discounting task. This did not support the idea that time perception and temporal decision making were related, at least as they were operationalised in this study. However, the behavioural measures from both tasks were independently related to some indices of autonomic nervous system function as measured by electrocardiography, suggesting distinct physiological correlates for both psychological processes. The third study was designed to assess whether factors that impact time perception also affect temporal decision making. Participants fasted for four hours, and then completed a task similar to a patch-leaving foraging paradigm, incentivised with monetary rewards. Participants gave up waiting for rewards significantly earlier when they experienced higher rates of reward. Participants who consumed a caloric drink in between blocks also gave up waiting significantly earlier, compared to those who consumed water (i.e. participants who consumed the caloric drink were less patient). These results suggest that the consumption of biologically relevant rewards altered time-dependent decision making. Overall, these findings support the notion that time perception can be affected by an individual’s homeostatic state, and further suggest that different homeostatic states can influence time-dependent decision making processes. Taken together, these experiments provide evidence that our experience of time may be part of a psychophysiological mechanism which may act to optimise ecological decision making.