In this thesis I study the effects of institutional trading on the comovementof financial assets. In the first chapter, joint work with Christopher Polk, we connect stocks through common active mutual fund ownership, and use these connections to forecast cross-sectional variation in return covariance, controlling for similarity in style and other pair characteristics. We argue this covariance is due to contagion based on return decomposition evidence, cross-sectional heterogeneity in the extent of the effect, and the magnitude of average abnormal returns to a cross-stock reversal trading strategy exploiting information in these connections. We show that the typical long/short hedge fund covaries negatively with this strategy suggesting that hedge funds may potentially exacerbate the price dislocation we document. In the second chapter I study the sources of change in the systematic risks of stocks added to the S&P 500 index. Firstly, using vector autoregressions (VARs) and a two-beta decomposition, I find that I cannot reject the hypothesis that all of the well-known change in beta comes from the cash-flow news component of a firm's return. Secondly, I study fundamentals of included firms directly to reduce any concerns that the VAR-based results are sensitive to my particular speciffcation. As ownership structure cannot directly influence fundamentals, these results challenge previous findings, as they are consistent with the change in beta being due to a selection effect. In the third chapter, joint work with Daniel Bergstresser, we explore index-based comovement in the market for Credit Default Swaps (CDS). We exploit the additions of individual CDS contracts in the Markit CDX Index, a major credit derivative benchmark. We find that for single name CDS contracts, comovement increases after inclusion in the index. Comparing movements in the CDS spreads to movements of the bonds of the same issuers, the CDS spread comovement increases significantly more than the bond spread comovement. This pattern of evidence is consistent with the excess comovement in equity markets documented by Barberis et al (2005) and others.

I show that risk sources such as unexpected demographic changes or shocks to the agent's subjective time preferences may have stronger implications and be of greater importance for asset pricing than risk in the (aggregate) consumption growth process. In the first chapter, I discuss stochastic changes to time preferences. Shocks to the agent's subjective time discounting of future utility cause stochastic changes in asset prices and the agent's value function. Independent of the consumption growth process, shocks to time discounting imply a covariation between asset returns and the marginal utility process, and the equity premium is non-zero. My model can generate both a reasonably low level and volatility in the risk-free real interest rate and a high stock price volatility and equity premium. If time discounting follows a process with mean- reversion, then the interest rate process is mean-reverting and stock returns are (at long horizons) negatively auto-correlated. In the second chapter, I analyze the asset pricing implications of birth and death rate shocks in an overlapping generations model. The interest rate and the equity premium are time varying and under certain conditions the interest rate is lower and the equity premium is higher during periods characterized by a high birth rate and low mortality than in times of a low birth rate and high mortality. Demographic changes may explain substantial parts of the time variation in the real interest rate and the equity premium. Demographic uncertainty implies a large unconditional variation in asset returns and leads to stochastic changes in the conditional volatility of stock returns. In the last chapter, I illustrate how shocks to the death rate may affect expected asset returns in the cross-section. An agent demands more of an asset with higher (lower) payoff in states of the world when he expects to live longer (shorter) and marginal utility is high (low) than an asset with the opposite payoff schedule. In equilibrium, the first asset pays a lower expected return than the latter. Empirical evidence supports the model. Out-of-sample evidence suggests that a strategy, which loads on uncertainty in the death rate, pays a positive unexplained return according to traditional market models.

This thesis contains three essays on disclosure of holdings by institutional investors. Chapter 1 presents a theoretical model that examines the impact of confidential treatment requests made by institutional investors to the Securities and Ex- change Commission (SEC) to delay disclosure of their holdings. Chapter 2 presents another theoretical model that analyses how an informed trader trades strategically in the presence of copycats who track his disclosed trades. Chapter 3 is an empirical study that examines the impact of more frequent portfolio disclosure on mutual funds' performance.

This thesis examines settings where multiple decision makers with conflicting interests benefit from cooperation in joint combinatorial optimisation problems. It draws on cooperative game theory, polyhedral theory and graph theory to address cost sharing in joint single-source shortest path problems and joint weighted minimum colouring problems. The primary focus of the thesis are problems where each agent corresponds to a vertex of an undirected complete graph, in which a special vertex represents the common supplier. The joint combinatorial optimisation problem consists of determining the shortest paths from the supplier to all other vertices in the graph. The optimal solution is a shortest path tree of the graph and the aim is to allocate the cost of this shortest path tree amongst the agents. The thesis defines shortest path tree problems, proposes allocation rules and analyses the properties of these allocation rules. It furthermore introduces shortest path tree games and studies the properties of these games. Various core allocations for shortest path tree games are introduced and polyhedral properties of the core are studied. Moreover, computational results on finding the core and the nucleolus of shortest path tree games for the application of cost allocation in Wireless Multihop Networks are presented. The secondary focus of the thesis are problems where each agent is interested in having access to a number of facilities but can be in conflict with other agents. If two agents are in conflict, then they should have access to disjoint sets of facilities. The aim is to allocate the cost of the minimum number of facilities required by the agents amongst them. The thesis models these cost allocation problems as a class of cooperative games called weighted minimum colouring games, and characterises total balancedness and submodularity of this class of games using the properties of the underlying graph.

This thesis contributes to economic research in microfinance and behavioural economics and bridges the gap between the two fields. Chapter 2 compares three lending mechanisms used by microfinance organizations: individual lending, simultaneous group lending and sequential group lending. The results are that the optimal choice of lending mechanism depends on the underlying distribution of project returns and on the level of available official contract enforcement. If contract enforcement is weak, sequential group lending unambiguously achieves the highest repayment rate. Hence sequential group lending can operate in settings in which simultaneous group lending and individual lending are not feasible due to weak contract enforcement. Chapter 3 shows that multiple price lists, currently the standard way of eliciting time preferences, will give biased estimates when income is uncertain. This is first shown theoretically and the resulting hypotheses are then tested empirically. The experiment finds that income risk causes participants to make more patient choices when choosing between two payments in the future. When choosing between an immediate and a future payment, however, income risk has no significant effect. As a result, participants with uncertain income appear more present-biased and less future-biased. Finally, only estimates obtained under safe income show a significant correlation with real-world financial outcomes. The fourth chapter shows how several features of the microfinance industry can be explained by projection bias over habit formation. Humans have a tendency to 34 underestimate to what extent their future preferences will differ from their current preferences and this systematic bias is known as projection bias. With the help of a formal model, this paper demonstrates that the prevalence of flat interest rate calculations, the high frequency of repayments, and the problem of over-investment can all be explained by this particular bias. Importantly, the policy implications resulting from this theory differ from those of other prevailing models, such as hyperbolic discounting.

This thesis includes three self-contained chapters whose common theme is the analysis of house price and rent movements, and how these movements influence the economic actions of individuals. In Chapter 1, I analyse a micro dataset on housing sales and rentals in Central London. I show that the ratio between prices and rents differ across property types: bigger and better located properties have higher price-rent ratios. These differences in price-rent ratios can be explained through a hedging model where households avoid rent risk by increasing their demand for homeownership. Consistently with this hypothesis, I find that rental prices for bigger properties and properties in more expensive neighbourhoods are not growing significantly faster than for other properties, but are more volatile. In Chapter 2, together with my two co-authors Christian Hilber and Olmo Silva, I study the relationship between homeownership and entrepreneurship by exploiting the longitudinal dimension of the British Household Panel Survey (BHPS) and constructing a detailed monthly-spell dataset that tracks individuals' job histories and tenure choices, coupled with other time-varying characteristics. Our fixed-effect estimates show that purchasing a house reduces the likelihood of starting a business by 20-25%. This result is driven by homeowners with mortgages and persists for several years after entering homeownership. The negative relationship can be rationalised by portfolio considerations: leveraged housing investments crowd out entrepreneurial investments. Alternative explanations based on credit constraints find little support in our data. In Chapter 3, I analyse the duration of house price upturns and downturns in the last 40 years for 19 OECD countries and provide two results. First, upturns display duration dependence: they are more likely to end as their duration increases. Second, downturns display lagged duration dependence: they are less likely to end if the previous upturn was particularly long. Both these facts are consistent with a boom-bust view of housing price dynamics, where booms represent departures from fundamentals that are increasingly difficult to sustain, and busts serve as readjustment periods.

This thesis seeks to understand how shifts in global economic power affect the policies and practices of multilateral development banks (MDBs). The study proposes three hypotheses. First, the “business” of development lending has changed radically in recent years as a result of the rise of middle income economies that now have a variety of options for sovereign borrowing—a reality thus far largely overlooked in academic research on MDBs. Second, a key factor defining the operational characteristics of the 20-odd MDBs in existence is the relative balance of power between borrower and nonborrower shareholders in MDB governance. Third, the financial pressures inherent in MDBs’ organizational models limit the options for MDB operations and shape how they will react to evolving market conditions. To test these hypotheses, the study examines several inter-related areas of MDB history and current operations, using as cases the World Bank (controlled by non-borrowing countries), the Inter-American Development Bank (control divided between borrowers and non-borrowers) and the Andean Development Corporation (controlled by borrowing countries). The analysis of the MDBs is complemented with case studies of Colombia and Ecuador, two countries with extensive borrowing histories with all three MDBs, to understand the demand side of development lending from the point of view of borrowing country government officials. The thesis finds compelling evidence in support of all three hypotheses, which suggests that the prevailing academic view that MDBs can be understood by focusing on the organization itself while ignoring the views of borrowers is not sufficient to understand the complexities of multilateral lending. MDBs are not all-powerful, but rather one resource among many at the disposal of governments to further their development, with varying competitive characteristics that impact the demand for their loans by borrowing countries in the current global context.

For many, microfinance is about building inclusive financial systems to help the poor gain direct access to financial services. Hundreds of grassroots have specialised in the provision of microfinance services worldwide. Most of them are adhoc organisations, which suffer severe organisational and informational deficiencies. Over the past decades, policy makers and consortia of microfinance experts have attempted to improve their capacity building through ICTs. In particular, there is strong emphasis on open source software (OSS) initiatives, as it is commonly believed that MFIs are uniquely positioned to benefit from the advantages of openness and free access. Furthermore, OSS approaches have recently become extremely popular. The OSS gurus are convinced there is a business case for a purely open source approach, especially across international development spheres. Nonetheless, getting people to agree on what is meant by OSS remains hard to achieve. On the one hand scholarly software research shows a lack of consensus and documents stories in which the OSS meaning is negotiated locally. On the other, the growing literature on ICT-for-international development does not provide answers as research, especially in the microfinance context, presents little empirical scrutiny. This thesis therefore critically explores the OSS in the microfinance context in order to understand itslong-term development and what might be some of the implications for MFIs. Theoretically I draw on the 3rd wave of research within the field of Science and Technology Studies –studies of Expertise and Experience (SEE). I couple the software ‘biography’ approach (Pollock and Williams 2009) with concepts from Simondon’s thesis on the individuation of technical beings (1958) as an integrated framework. I also design a single case study, which is supported by an extensive and longitudinal collection of data and a three-stage approach, including the analysis of sociograms, and email content. This case provides a rich empirical setting that challenges the current understanding of the ontology of software and goes beyond the instrumental views of design, building a comprehensive framework for community participation and software sustainability in the context of the microfinance global industry.

The behaviour of a nation's stock market is increasingly seen as a barometer of its economic growth, strength and stability. While the behaviour of well established equity markets is well researched and documented, the behaviour of small and developing exchanges is still not much studied. This thesis examines and analyses some aspects of the behaviour of an emerging equity market known as the Kuala Lumpur Stock Exchange (KLSE) - the national stock exchange for Malaysia. To serve as the groundwork for our empirical investigation, this study begins with a survey of the related literature. The literature on efficient market hypothesis (EMH); the literature on various theories and models which are complementary and contradictory to the EMH - are reviewed. Empirically, four major aspects of the behaviour of the KLSE are examined. Using both share price indices and individual company share prices/returns for a sample period 1984:01 through 1994: 12, we study some statistical properties of stock returns, correlations with other markets, stock market forecastability and the presence of mean reversion/mean aversion in stock returns. The behaviour of the KLSE market indices are compared in several respects with the indices of selected developed markets. Our study has resulted a number of findings, some of which could be considered as intriguing and novel for a relatively unresearched market like the KLSE. Similar to many previous researches, our study has provided evidence that the distributions of stock returns are not normal. Rather, they are leptokurtic. Variances/standard deviations of stock returns on the KLSE were found to be large compared with, for example, the New York and London stock exchanges, but the realised returns were not significantly different for the period of study. The KLSE is found to be positively correlated with most foreign exchanges, although these correlations are far from unity. These correlations however, are not constant/stable through time. Our evidence also suggests that the Malaysian market tends to exhibit strong regional links. Additionally, the KLSE appears to have significant lagged correlations with a number of developed exchanges. Three equity markets are identified as the most influential foreign exchanges to the KLSE in terms of their comovements (and/or lagged correlations). They are, the Stock Exchange of Singapore, the Hong Kong Stock Exchange and the New York Stock Exchange. We found no evidence that the "forecastability" of stock prices/returns on the KLSE could be improved when an 'out-of-sample' forecasting procedure known as the multi-process models was employed. Moreover, we have found that the returns for some stocks are more forecastable than others. Variance ratio tests indicate that over long horizons, some stocks listed on the KLSE tend to exhibit mean reversion, some are mean aversive and the rest seem to follow a random walk. The present research has also raised a number of issues which might be interesting for further study. These issues are discussed in Chapter Seven of the thesis.

This thesis aims to revise the current understanding of the behaviour of different groups of traders in financial markets. Research involves statistical analysis of historic 'Commitment of Traders' reports, a U.S government dataset providing the long and short positions of core groups of traders, reported at weekly intervals over 17 years. Empirical work identifies a surprising level of consistency amongst different groups across 31 markets. A specific pattern is identified: speculators are found to increase their buying interest when prices are rising whilst commercial traders (or 'hedgers') increase their selling; the opposite pattern of behaviour occurs when prices are falling. The thesis explores the implications of this behaviour for existing models of financial markets by referencing a number of peer-reviewed studies. The agent-based computational model of Alfarano, Lux, and Wagner (2005) is implemented and analysed. A lack of validity is demonstrated in the interactions between the different types of traders in this model. These theoretical components are further shown to be typical of much of the literature in this area. An objective for the thesis is to correct this oversight by incorporating genuine patterns of trading behaviour into an existing computational model. The approach of Mike and Farmer (2008) is used for this purpose, being currently unique in that core components are calibrated from real-world data and no group-level representations are assumed. This model is extended to observe groups of traders with different levels of order-aggression: speculators are found to rely on market orders whereas commercial traders rely on limit orders. These preferences, in the absence of deeper theoretical considerations, are sufficient to account for the identified behaviour. A discussion is offered on the relevance of this finding for financial market regulators, who have typically focused on regulating types of traders, specifically speculators, rather than on types of trades.