The Holy Grail of Macro-finance

Hang Bai (UConn) and I have just circulated our new working paper titled “Searching for the equity premium” (paper, slides).
 
We view this work as a solid progress report toward the holy grail of macro-finance, which (in our view) is a unified theory of asset prices and business cycles.
 
The persistence of the Mehra-Prescott (1985) equity premium puzzle in general equilibrium production economies has given rise to a long-standing dichotomy in macro-finance. Finance specifies “exotic” preferences and exogenous cash flow dynamics to match asset prices but ignore firms (Campbell and Cochrane 1999; Bansal and Yaron 2004; Barro 2006). Macroeconomics analyzes full-fledged dynamic stochastic general equilibrium (DSGE) models but ignore asset prices with primitive preferences (Christiano, Eichenbaum, and Evans 2005; Smets and Wouters 2007).

This macro-finance dichotomy has left many important questions unanswered. What are the microfoundations underlying the exogenously specified, often complicated cash flow dynamics in finance models (Bansal, Kiku, and Yaron 2012; Nakamura, Steinsson, Barro, and Ursua 2013)? What are the essential ingredients in the production side that can endogenize the key elements of cash flow dynamics necessary to explain the equity premium? To what extent do time-varying risk premiums matter quantitatively for macroeconomic dynamics? How large is the welfare cost of business cycles in a general equilibrium production economy that replicates the equity premium?
 
We embed the standard Diamond-Mortensen-Pissarides search model of equilibrium unemployment into a DSGE framework with recursive utility and capital accumulation.

Highlights of our quantitative results include:

• Calibrated to the consumption volatility in the Jorda-Schularick-Taylor macrohistory database, our DSGE model yields a (leverage-adjusted) equity premium of 4.26% per annum, an average interest rate of 1.59%, and a stock market volatility of 11.8%.

• Our model yields strong time series predictability for stock market excess returns and volatility, some predictability for consumption volatility, and weak to no predictability for consumption growth and real interest rate.

• The model features wage inertia with a wage elasticity to labor productivity of 0.256. We provide a fresh estimate of this wage elasticity to be 0.267 in the historical U.S. 1890-2015 sample.

• Wage inertia yields strongly procyclical dynamics of profits, which are sufficient to overcome procyclical investment and vacancy costs to turn dividends procyclical. Time-varying wage inertia also explains the predictability for market excess returns and stock market volatility.

• Investment absorbs a large amount of shocks, making consumption growth and the interest rate largely unpredictable.

• Risk aversion strongly affects quantity dynamics, overturning Tallarini (2000).

• Our model yields downward-sloping term structures of equity return and volatility.

• Our model features a reasonable timing premium of 15.3%, avoiding the pitfalls of some parameterizations of the long-run risks and disaster models.

• The welfare cost is huge, 29.1%, and strongly countercyclical.

 
In all, the DSGE model with recursive utility, search frictions, and capital accumulation is a good start to forming a unified theory of asset prices and business cycles.