Browsing by Subject "Econometrics"

Causal inference methods are commonly used across domains to aid high-stakes decision-making. The validity of causal studies often relies on strong assumptions that might not be realistic in high-stakes scenarios. Inferences based on incorrect assumptions frequently result in sub-optimal decisions with high penalties and long-term consequences. Unlike prediction or machine learning methods, it is particularly challenging to evaluate the performance of causal methods using just the observed data because the ground truth causal effects are missing for all units. My research presents frameworks to enable validation of causal inference methods in one of the following three ways: (i) auditing the estimation procedure by a domain expert, (ii) studying the performance using synthetic data, and (iii) using placebo tests to identify biases. This work enables decision-makers to reason about the validity of the estimation procedure by thinking carefully about the underlying assumptions. Our Learning-to-Match framework is an auditable-and-accurate approach that learns an optimal distance metric for estimating heterogeneous treatment effects. We augment Learning-to-Match framework with pharmacological mechanistic knowledge to study the long-term effects of untreated seizure-like brain activities in critically ill patients. Here, the auditability of the estimator allowed neurologists to qualitatively validate the analysis via a chart-review. We also propose Credence, a synthetic data based framework to validate causal inference methods. Credence simulates data that is stochastically indistinguishable from the observed data while allowing for user-designed treatment effects and selection biases. We demonstrate Credence's ability to accurately assess the relative performance of causal estimation techniques in an extensive simulation study and two real-world data applications. We also discuss an approach to combines experimental and observational studies. Our approach provides a principled approach to test for the violations of no-unobserved confounder assumption and estimate treatment effects under this violation.

This dissertation discusses model and forecast comparison, calibration, and combination from a foundational perspective. For nearly five decades, the field of forecast combination has grown exponentially. Its practicality and effectiveness in important real world problems concerning forecasting, uncertainty, and decisions propels this. Ample research-- theoretical and empirical-- into new methods and justifications have been produced. However, its foundations-- the philosophical/theoretical underpinnings on which methods and strategies are built upon-- have been unexplored in recent literature. Bayesian predictive synthesis (BPS) defines a coherent theoretical basis for combining multiple forecast densities, whether from models, individuals, or other sources, and generalizes existing forecast pooling and Bayesian model mixing methods. By understanding the underlying foundation that defines the combination of forecasts, multiple extensions are revealed, resulting in significant advances in the understanding and efficacy of the methods for decision making in multiple fields.

The extensions discussed in this dissertation are into the temporal domain. Many important decision problems are time series, including policy decisions in macroeconomics and investment decisions in finance, where decisions are sequentially updated over time. Time series extensions of BPS are implicit dynamic latent factor models, allowing adaptation to time-varying biases, mis-calibration, and dependencies among models or forecasters. Multiple studies using different data and different decision problems are presented, demonstrating the effectiveness of dynamic BPS, in terms of forecast accuracy and improved decision making, and highlighting the unique insight it provides.

This dissertation asks whether frequency misspecification of a New Keynesian model

results in temporal aggregation bias of the Calvo parameter. First, when a

New Keynesian model is estimated at a quarterly frequency while the true

data generating process is the same but at a monthly frequency, the Calvo

parameter is upward biased and hence implies longer average price duration.

This suggests estimating a New Keynesian model at a monthly frequency may

yield different results. However, due to mixed frequency datasets in macro

time series recorded at quarterly and monthly intervals, an estimation

methodology is not straightforward. To accommodate mixed frequency datasets,

this paper proposes a data augmentation method borrowed from Bayesian

estimation literature by extending MCMC algorithm with

"Rao-Blackwellization" of the posterior density. Compared to two alternative

estimation methods in context of Bayesian estimation of DSGE models, this

augmentation method delivers lower root mean squared errors for parameters

of interest in New Keynesian model. Lastly, a medium scale New Keynesian

model is brought to the actual data, and the benchmark estimation, i.e. the

data augmentation method, finds that the average price duration implied by

the monthly model is 5 months while that by the quarterly model is 20.7

months.

I provide new evidence on the properties of the temporal structure of risk, which answers whether more distant claims to macroeconomic growth are more or less risky than near-term claims. In the first chapter, I use replication and no-arbitrage to estimate within-firm variation in equity expected returns across horizons. I demonstrate that a low dimensional set of returns and state variables, all characteristics of liquid, exchange-traded equity securities, provide a close replication of claims to firm capital gains at different horizons. Calculating returns from the no-arbitrage prices of these claims, I show that the term structure of risk premia is unconditionally upward-sloping for commonly used test assets like the market and book-to-market sorted portfolios. In joint work in the second chapter, we use traded equity dividend strips from U.S., Europe, and Japan from 2004-2017 to study the slope of the term structure of equity dividend risk premia. In the data, our robust finding is that the term structure of dividend risk premia (growth rates) is positively (negatively) sloped in expansions and negatively (positively) sloped in recessions. We develop a consumption-based regime switching model which matches these robust data-features and the historical probabilities of recession and expansion regimes. The unconditional population term structure of dividend risk premia in the regime-switching model, as in standard asset pricing models (habits and long-run risks), is increasing with maturity. In sum, our analysis shows that the empirical evidence in dividend strips is consistent with a positively sloped term structure of dividend risk-premia as implied by standard asset pricing models.