This study proposes a novel nonparametric estimation-based approach to solving asset pricing models. Our method is robust to misspecification errors while inheriting a closed-form solution that facilitates ease of implementation. By representing the Euler equation into a well-posed integral equation of the second kind and further transforming it into an equivalent regression model, our estimate is fully identified and free of estimating the transition density of the state dynamics. We establish the large sample properties of the proposed penalized splines estimator under very mild conditions. With the merit of penalized splines, we design a fast data-based algorithm to effectively tune the smoothing parameter. Our approach exhibits superior performance even for a small sample size. For application, using US data from 1947 to 2017, we reinvestigate the return predictability and find that high implied dividend yield, obtained from our misspecification-free approach, significantly predicts lower future cash flows and higher interest rates at short horizons.
Friday, February 28, 2020
3:30pm – 5:00pm
Small reception to follow in Room 426. All are invited to attend.