The enhanced access of developing countries to the international financial market since the seventies has been characterized by boom-bust cycles of unfettered external borrowing followed by abrupt financial crises. The first chapter analyzes the macroeconomic effects of volatile capital flows to a developing country. The analysis shows that investment, consumption, and the current account deficit depend positively on the expected availability of external finance. If international investors may unexpectedly decide to reduce their exposure to financial assets issued by the country, the optimal cost of external borrowing should exceed the interest rate paid by domestic residents in the international financial market. In the absence of insurance markets for this type of risk, a tax on capital inflows can be optimal.

Recent endogenous growth models characterize a firm's technology as a commodity which is both partly excludable and associated with some production inputs, such as human capital and equipment. The second chapter explores the nature of the link between equipment investment and technology at the plant level in a large sample of Colombian manufacturing establishments. The results support the endogenous growth model's notion that technology is associated with the production inputs. Larger plants that invest more in machinery and equipment and employ higher levels of human capital tend to be more efficient.

Models of investment with non-convex costs of adjustment predict that microeconomic time series of investment may be characterized by infrequent investment spurts and prolonged periods of little or no investment. In the third chapter I study the pattern of investment at the plant level in different categories of capital goods. As in the U.S., plant-level investment in Colombia is lumpy, and the probability of observing a large investment episode depends positively on the time elapsed since the latest large investment episode. As a contribution to the literature, I propose and implement two alternative econometric methods for the estimation of a simple model of irreversible investment. The results show that increases in the real exchange rate (pesos per dollar) have a consistently negative effect on investment, regardless of the type of capital good.