Essays in development economics and finance

Abstract

This thesis is a collection of three essays on development economics and finance. The first chapter studies the 1992 presidential impeachment in Brazil to evaluate the impact of an anti-corruption drive on politically connected companies. I identify two types of firms: companies owned by friends and relatives of the impeached president ('family-connected') and firms proven to be connected to him in a parliamentary investigation ('other-connected'). Using an event study procedure, I establish that family-connected firms have on average negative daily abnormal returns of 2 to 9 percentage points when damaging information about the president is released. However, the 'other-connected' companies do not experience a decline in their stock market valuation during the impeachment. Furthermore, the stock market decline experienced by 'family connected' companies was reversed entirely within a year. The impeachment had limited success in reducing corruption. The second chapter evaluates the effects on multinational firms of the OECD "Convention on Combating Bribery of Foreign Public Officials in International Business Transactions". I compare the balance sheet performance of foreign companies in 24 developing host countries whose source countries have implemented the convention with the performance of firms whose source countries have not yet implemented it. I find that the OECD convention had a negative impact on profit and sales growth of multinational companies. This effect is amplified in countries with less efficient bureaucracies. In economies where bribery is more valuable to firms, the OECD convention has a larger negative impact on multinational firms. The third chapter studies in detail the distribution of one type of financial market participant: mutual funds. The essay documents that their size follows a regularity observed in several other area of economics, Zipf's law: the number of funds with size greater than x is proportional to 1/x. This chapter extends previous theories of random growth to explain why this is the case: Zipf's law arises when mutual funds grow at the highest speed allowed by constraints in the system, something we call a "maximum growth principle." We investigate empirically the key features of the theory, and show that they are validated by the data.