Three essays on the co-evolution of the real and financial sector during the growth process

In a seminal work by Gurley and Shaw, they established that economic development is typically accompanied by a transition from self-financed to intermediated investment. This work seeks to analyse the co-evolution of the real and financial sectors during this growth process. We study the biases generally observed in the AR(1) OLS estimation, and apply the result of these to causality tests between finance and growth. As originally identified by Sawa these biases are present in regressions using small samples, while OLS will normally be consistent if the error term is i.i.d. We assess and find analytical expressions for these biases. The phase-average methodology has been often used to study the growth effect of financial deepening; notably by Barro and Sala-i-Martin. We study the properties of phase averages of the variables when estimating parameters in growth regressions. We find that phase averages artificially increase the size of estimated parameters, introduce inefficiencies and bias the estimation by OLS. We conclude that phase averaging is more likely to produce problems than benefits, as aggregation from annual data entails loss of information and does not reduce other data problems such as serial-correlation. Chilean financial markets have experienced extensive development, increased participation, deregulation and new legislation over the last two decades. We analyse a financial database of equity, traded on the Chilean stock market between 1989 and 2000. We develop a model for the relationship between risk and return, under market frictions. We find that market frictions like compulsory holdings and short sale restrictions have a negative effect over asset prices for given levels of risk. We study the evolution of these frictions using the Chilean data