Endogenous Trading Constraints with Incomplete Asset Markets

The present project introduces the possibility of default on the trading contracts in an infinite horizon incomplete markets model, relaxing the usual assumption made in the literature with respect to the trading limits, which are chosen to be fixed or independent of the characteristics of the economy. In particular, it assumes that households can break their trading contracts by going into autarky, in which case they are seized from their total assets and only receive labour income from the default period on. Further, to avoid this happening in equilibrium, the trading limits are endogenously determined at the level where the participation constraint is satisfied at each possible date and state. Using this set-up, the main objectives are to quantify and characterize the endogenous limits resulting from the option of default and to study their real and financial implications along the growth path and during the business cycle. Our work is of considerable relevance, since it builds a bridge between several important strands of literature. First, it is related to the usual incomplete market models with ad hoc trading limits, such as Heaton and Lucas (1996), Marcet and Singleton (2001) or Telmer (1993), who study asset prices in two agent exchange economies, and Aiyagari (1994) or Krussel and Smith (1997) and (1998), who study production economies with a large number of households. In addition, it also relates to a recent literature, where a number of authors have introduced the possibility of default through participation constraints, resulting in agent and state specific endogenous trading limits. Among others, Kehoe and Levine (1993) and Alvarez and Jermann (2000) introduce this type of limits in exchange economies, while Kehoe and Perri (2002) analyse a production economy where investors are interpreted as countries. All these authors, however, introduce the option of default into a complete markets context, which they compare to an exogenous incomplete markets environment without explicitly characterizing the trading limits. In contrast to this, the present framework allows for a direct comparison of economies with the same asset structure, making it possible to isolate the effects of no default on the equilibrium allocations. Finally, our work is also related to the papers of Zhang (1997a) and (1997b), where the author derives the endogenous borrowing limits resulting from the possibility of default in a Lucas type exchange economy with one asset. In contrast to this, our work introduces different types of assets and incorporates a production sector, leading to state-dependent autarky values and trading limits, which considerable complicate the computations. In this respect, the contribution of the present project is also methodological, since it extends the standard policy iteration algorithm to incorporate state dependent limits and state dependent lower bounds for the grid of the endogenous states. To summarize, the present work fills an important gap in the literature, since it links the incomplete markets models with fixed trading limits and the complete markets models with the option of default, allowing for a complete characterization of the no-default endogenous trading constraints and for a better understanding of the endogenous portfolio choice problems in the absence of full commitment. Further, while our methodological contribution will enhance the understanding on the numerical techniques used to solve problems with default, it is very likely to be of use to other researchers in the field. Preliminary results on the characterization of the endogenous trading limits for the case in which a finite number of households can trade in shares of aggregate capital already shed light on important implications of the present work that can have a considerable impact on the different strands of literature studying risk sharing in infinite horizon economies. As expected, the endogenous limits do vary to a great extent with the aggregate state of the economy. In particular, when aggregate capital is scarce, and the incentives to default are relatively low, limits on individual capital shares are relatively loose. On the other hand, as the economy accumulates capital, they become tighter. Apart from this, the endogenous limits computed using a preliminary but reasonable calibration allow for short-selling of the stock in the magnitude of ten to fifteen percent of aggregate capital in the stationary distribution, being surprisingly close to the zero fixed limits usually imposed in the literature with fixed trading limits. While this is good news for the models studying steady state behaviour, it suggests that it can be dangerous to impose fixed limits during transition to the steady state, since at earlier stages of development and a relatively low capital they can vary to a great extent. In this case, the different risk sharing opportunities resulting from the presence of endogenous limits also lead to important implications concerning asset wealth and consumption inequality, which are sensitive of the particular stage on the transition path.