Uncertainty and firm investment

This thesis explores effects of uncertainty on firm investment that are described in estimates of firm level investment specifications which include proxies for uncertainty over expected future firm profitability. A panel data set of UK firms covering the period 1987-2000 is used to estimate firm level investment specifications. Within year volatility in stock returns - a common proxy for firm specific uncertainty in previous literature - is compared with covariance measures between stock returns and market returns representing un-diversifiable risk from the CAPM; and with alternative uncertainty proxies based on volatility in I/B/E/S securities analysts' forecasts of earnings per share. Within estimates of firm level investment specifications, the thesis investigates the sensitivity of coefficients on uncertainty terms to the choice of underlying investment specification: error correction model between the natural logarithms of capital and sales; or the Hayashi (1982) Q model of investment. Coefficients on stock return volatility measures of uncertainty terms are found to vary significantly between estimates of error correction and average q specifications. Differences between coefficients estimated on uncertainty terms across estimates of these two investment specifications are supported with simulated data. Uncertainty measures based on volatility in I/B/E/S securities analysts' forecasts of earnings per share are found to be much more informative of investment behaviour than within year stock return volatility in estimates of both error correction and average q specifications. Coefficients on I/B/E/S uncertainty proxies imply more consistent investment-uncertainty relationships across estimates of error correction and average q specifications for the UK panel data set.This thesis is not currently available via ORA