Working Papers

The Rise of Negative Earnings and Demand Shifting Investment

Joint with Dalton Rongxuan Zhang

Last updated: May 2026

Abstract We document the rise of negative earnings between 1980 and 2019: a secular increase in the percent of firms reporting losses, both among public firms and in the broader universe of US corporations, and a secular increase in the persistence of losses year-to-year among public firms. This rise has occurred alongside a spreading of the sales and earnings distribution and a recomposition of firm spending away from production costs and traditional investment and towards selling, general and administrative expenses (SG\&A). Motivated by prior literature, we consider SG\&A spending a form of intangible investment and estimate the corresponding ``customer capital'' stock using the perpetual inventory method. We find that the sales elasticity of customer capital has secularly risen since 1980. We rationalize these phenomena with a model of heterogenous firms engaging in \emph{supply and demand shifting investment}. Our model includes a \emph{scale elasticity of demand} determining the relationship between the intensive margin of demand (demand per customer) and the extensive margin of demand (number of customers). We are able to quantitatively match the rise in reported losses and qualitatively match (1) the increased persistence of losses, (2) the spreading of the sales and earning distribution and (3) the recomposition of firm spending with this parameter as the single driver of changes across steady state equilibria. The rise in the scale elasticity associated with the increase in reported losses is in line with the rise in the sales elasticity of customer capital we estimate in the data, and has non-trivial aggregate implications: in our model it lowers GDP by -9.1\% by reallocating labor away from goods and capital production and reallocating demand away from productive firms.

Paper, arXiv, Code

Sector-Specific Substitution and the Effect of Sectoral Shocks

Last updated: May 2026

Abstract How a shock to an individual sector propagates to the prices of other sectors and aggregates to GDP depends on how easily sectoral goods can be substituted in production, which is determined by the intermediate input substitution elasticity. Past estimates of this parameter in the US have been restrictive: they have assumed a common elasticity across industries, and have ignored the use of imports in production. This paper uses a novel empirical strategy to produce new estimates without these restrictions, by exploiting variation in import ratios and in input expenditure shares within industries rather than across industries. I find that sectors differ meaningfully in their ability to substitute inputs in production, and that the uniform estimate of the intermediate input substitution elasticity is biased downwards relative to the median sector-specific estimate. Relative to imposing the uniform elasticity, sector-specific substitution causes domestic prices to rise more in response to oil import shocks and less in response to semiconductor import shocks. It also implies the average GDP response to a sectoral business cycle is 0.35% higher, making sectoral business cycles 17.7% less costly.

Paper, arXiv, Data and code

Pre-PhD

Insurer Competition in the Age of Provider Consolidation

(Undergraduate Thesis, 2019)

Abstract This paper investigates the impact of hospital competition (or lack thereof) on insurer participation in the ACA's individual market. Using public data from CMS, and private data from the American Hospital Association (AHA), I construct the Herfindahl–Hirschman Index (HHI) for hospital and insurer markets at the county-level in 34 of the 36 states using federally facilitated marketplaces, across 2015 and 2016 (hospital HHI is lagged by one year). I fit a linear model on 2063 counties across two years in these states, controlling for county-level covariates and fixed effects for year-"rating area" (a geographic designation created by the ACA, which typically amounts to a collection of counties). I estimate my parameters using OLS. I find higher hospital HHI levels are associated with higher insurer HHI levels at a coefficient of .033, log linearized. I lay the groundwork for further analysis once more years of data are available, contributing to the existing literature by focusing on insurer competition rather than premium price as my primary outcome, leveraging "rating areas" for better model specification, and outlining a novel approach to hospital market HHI construction using hospital "radii" rather than pre-existing geographic bounds.

Paper