Risk-sharing and international portfolio decisions – intport

In the last two decades, financial globalization has led to a surge in cross-border capital flows, a large increase in countries gross foreign asset and liability positions, together with increasing global imbalances. Hence, understanding how private investors (and Sovereign Wealth Funds) structure their international asset portfolios has recently become a critical macro policy issue. This is especially true in the wake of the current crisis as we are expecting big changes in savings positions and large shifts in investor portfolios. This feature has generated a renewed attention by researchers to incorporate non-trivial portfolio choice models into standard dynamic model of the world economy. Indeed, while the real side of globalization has previously attracted a lot of attention, the financial aspect of globalization has been mostly neglected until recently. Dynamic models of the international macro-economy used by Central Banks or International Organizations do not incorporate international portfolio decisions. This is an important issue as cross-border financial flows in turn might endogenously affect the real side of the economy, in particular the way shocks are transmitted internationally and the degree of synchronization of business cycles. The main objective of this research is to incorporate such portfolio decisions in an otherwise standard macro-model and confront the model predictions to the data.

Such a contribution would allow one to better analyze the adjustment of external global imbalances, the joint dynamics of exchange rates and asset prices, as well as the international transmission of shocks. This is particularly relevant regarding the recent macroeconomic developments since international financial linkages have been playing a key role in propagating globally the US financial shock.

An important aspect of the modelling strategy will be to extend existing research in the field along three main dimensions.

First, I will allow for a broader class of assets to be traded internationally and in particular incorporate cross-border debt positions in existing models (Task 1). Most of the literature has been focusing on equity positions but in the data cross-border debt positions in different currencies are large and generate significant wealth transfers across countries following movements in exchange rates. The empirical implications of the model will be tested using a new constructed dataset on cross-border investment positions for a large sample of countries across various types of assets (equities, public and private debt, bank loans).

Second, I will introduce financial market incompleteness by allowing limited participation in asset markets within countries (Task 2). These restrictions in the access to international financial markets to some households will bring existing models closer to the data as most people do not have the possibility of insuring themselves internationally against bad shocks. An empirical test of the model will involve extending the standard tests of international risk sharing to allow for heterogeneous agents by using household level surveys on consumption and income. In particular, tests of international risk-sharing will be performed by looking at the cross-country consumption growth of participants versus non-participants.

Third, I plan to provide new methodologies to solve these models even in a presence of large macroeconomic shocks, large risk premia on financial assets and/or incomplete financial markets (Task 3). Indeed most macro models rely on ‘local’ approximations around a deterministic steady-state and might not be accurate in such contexts. This last issue is particularly relevant in the context of the current financial crisis.