Investing Through Turbulence: Capital Expenditure, Leverage, and the Persistent Performance of Hotel and Restaurant Firms

Abstract

This study investigates how capital expenditure (CAPEX) intensity and leverage jointly relate to performance and risk in publicly listed U.S. hotel and restaurant firms. Using panel data for 67 corporations (23 hotels, 44 restaurants) from 2013 to 2023, the analysis models lagged CAPEX intensity and leverage alongside firm size, sales growth, and operating cash flow in firm-level regressions and portfolio-level time-series analyses. The sample consists predominantly (78%) of asset-heavy property owners rather than asset-light franchisors, with important implications for interpretation. Results reveal that higher CAPEX intensity is positively associated with subsequent accounting performance (return on assets, earnings before interest and taxes [EBIT]/total assets, liquidity ratios, and market-to-book), with associations substantially stronger for asset-heavy hotel corporations than restaurant corporations or asset-light franchisors. Leverage exhibits a negative relationship with accounting performance, particularly pronounced in the capital-intensive asset-heavy hotel sector. Portfolio-level autoregressive models demonstrate strong persistence in operating performance, indicating that firms sustaining higher investment intensity tend to remain in favorable performance regimes across years. Simultaneously, Generalized Autoregressive Conditional Heteroskedasticity models reveal pronounced and persistent volatility clustering in risk-adjusted equity returns, suggesting that intensive CAPEX programs are accompanied by extended periods of heightened uncertainty for investors. The findings underscore fundamental tensions in hospitality finance: while sustained investment is associated with long-term value creation, it must be balanced against leverage constraints and managed investor expectations around short-term volatility during turbulent periods such as the COVID-19 shock. These results should be interpreted as conditional associations rather than causal effects, given inherent endogeneity concerns in observational investment data.