Essays on portfolio choice and social security

Social security represents a fertile territory to be explored with the tools of modern finance, as social security systems in general affect the savings decisions of individuals in non trivial ways. This dissertation studies three issues of social security using the tools of continuous-time finance. All these issues are heavily motivated by the Chilean experience in the design of unemployment insurance and pension systems based on individual accounts.;The first chapter deals with the incentives embedded in the compensation scheme of a risk averse portfolio manager. The interest in this case is placed on how the compensation scheme changes the investment decisions of the portfolio manager. Using a compensation scheme based on a benchmark portfolio to define the penalties and bonuses of the scheme, it is shown that the scheme motivates the portfolio manager to imitate the investment strategy of the benchmark portfolio whenever this helps the manager to either obtain a bonus or avoid a penalty.;The second chapter focuses on the optimal design of benchmark portfolios. The analysis is carried out in the context of a classical Merton type portfolio choice problem. In particular, two related portfolio choice problems are studied, one dealing with the concerns of a representative worker (e.g., unemployment risk), and a second one dealing with the concerns of the portfolio manager in charge of a solidarity fund that finances the payment of a top-up monetary benefit to workers. The results suggest that the benchmark portfolio of the solidarity fund should take into account the optimal investment strategy of the representative worker's problem as the latter represents the funded portion of the liabilities of the former.;The third and final chapter focuses on the long term assessment of the financial risk in a defined contribution pension system. In particular, it analyzes a case in which the competitive incentives dissociate the investment objectives of the portfolio manager from those of the pension fund member. The results suggest that the common association between risk and stock volatility may be misleading.