Temptation and forward-guidance

In the aftermath of the recent financial crisis, several central banks have resorted to "forward guidance" in monetary policy - that is, announcing publicly the future path of the short-term interest rate - to stimulate inflation and economic activity, and therefore move the economy away from the liquidity trap. Standard monetary models predict sizable stimulative effects of forward guidance, which are much in excess of what observed in the data and grow exponentially with the forward guidance horizon. This apparent disconnect between theory and data has been labeled by the literature as the forward guidance puzzle. We introduce temptation and dynamic-self-control preferences, as formalized by Gul and Pesendorfer (Econometrica 2002, 2004), into an otherwise standard New Keynesian model. In our set-up, the representative agent faces a temptation to liquidate his entire financial wealth for the purpose of immediate consumption. Resisting temptation involves cognitive effort (or self-control) and hence some disutility. Optimal behavior therefore trades of the temptation for immediate satisfaction with long-run optimal consumption smoothing. We show that, for suitable parameterizations, dynamic self-control preferences deliver a discounted Euler equation,lower households. sensitivity to real interest rates, and make the Phillips curve less forward-looking. These features combined help solve the puzzle. Moreover, they have stark implications for what concerns the conditions for equilibrium determinacy, the adverse effects of large negative shocks to the real interest rate, and the paradox of volatility.