Remittances, investment, and portfolio allocations: An analysis of remittance usage and risk-tolerance

This dissertation shows how remittances affect household investment decisions. The investment decision is broken into two components. First, the household allocates expenditures between aggregate consumption and investment bundles. Second, the household determines which investment assets to purchase based on the investment's expected return and the household's risk-tolerance. Each decision stage is independent and is modeled separately.; Previous empirical analyses of remittance effects on household investment purchases have contradictory results. Most studies evaluate remittance effects by monitoring the accumulation of specific goods. Problems arise when motivations for purchasing specific goods are caused by factors other than remittances. The first section of this dissertation attempts to isolate remittance effects by evaluating a household's budget allocation when composed of aggregate consumption and investment bundles versus single goods. Using panel data from rural Pakistan, households are split into three groups: households receiving remittances from abroad, households receiving remittances from within Pakistan, and non-remittance households. An LES demand system is used to evaluate a household's budget allocation by analyzing prices and expenditures.; The receipt of remittances shows a direct effect on the allocation of total expenditures to investment goods. The estimates show households without remittances value investments as non-essential goods, the opposite of how remittance receiving households value investments. The elasticity of substitution between investment and consumption shows no sign of being affected by remittances. A fixed-effects logit regression further shows remittances have a positive, statistically significant effect on the probability of purchasing investments.; The second section of the study evaluates risk-aversion levels of the three types of households. To detach a household's risk-aversion from remittance effects on the investment decisions for individual goods, the entire portfolio is modeled. Returns to eight investment choices (agricultural land, tractors, tubewells, livestock, urban land, human capital, and two types of banking accounts) are calculated on a yearly basis using aggregate country wide data for a total of 15 years. A mean-variance portfolio is modified to reflect Islamic banking constraints and is used to calculate the optimal portfolio frontier.; The results suggest remittances have little effect on a household's portfolio selection. Remittances do not affect a household's risk-aversion level. The portfolio allocations for all household groups are nearly identical, resulting in negligible differences in the portfolio returns and risk levels. All household groups could better optimize their portfolio as all the allocations were found to be inefficient. Despite inefficiencies, the results suggest remittances do not influence the purchase of any specific asset.