The impact of liquidity shocks on capital

This thesis sheds some light on factors that affect the level and stability of investment in economies with undeveloped capital markets. In particular, we focus on shocks to the demand for bank liabilities (checks). Banks are the major source of finance for firms, and fiat money and checks are the main media of exchange for households. To what extent is investment stimulated by liquidity shocks that increase the demand for bank liabilities? Our analysis is less relevant for economies with developed capital markets. However, to understand how the evolution of capital markets affects the stability and growth of investment, we focus on economies where banks are still important and capital markets are undeveloped. The thesis is composed of an empirical and a theoretical analysis. We first show with linear regression estimates that the effect of the variance of demand deposits on that of investment is statistically significant in countries with undeveloped capital markets. Such is not the case for developed countries. We then proceed to provide a theoretical analysis of how liquidity shocks affect investment in developing countries. Throughout our thesis, we model liquidity shocks as resulting from changes in the amounts of relocation of spatially separated agents searching for trading partners. Spatial separation is a trading friction which endogenously gives rise to fiat money as a medium of exchange. The relocation of agents has the effect of changing the demand for liquidity. This changes the demand for investment-backed inside money and therefore impacts investment. This contrasts with the Clower constraints/infinite horizon models in the literature where the real sectors of an economy, and hence capital, are insulated from monetary effects. We provide two spatial separation/overlapping generations models. We show that equilibrium capital is affected by liquidity shocks due to the overlapping generations (OLG) structure. OLG is analogous to the case of households borrowing constraints, a feature of economies with undeveloped capital markets. Specifically how investment responds to liquidity shocks depends on the timing of the shock. In the first model, we show that when the young are hit with a liquidity shock, capital is a non-monotonic function of the shock. In the second model, old agents are subject to a liquidity shock. In this case, capital is an increasing function of the liquidity shock