Debt maturity structure of private lodging firms

ABSTRACT

This study tests the signalling and liquidity risk hypotheses about the choice of debt maturity structure in the context of private lodging firms. We find a positive and significant relationship between the Z-score (as a proxy for firm quality) and the proportion of short-term debt. This finding supports the signalling hypothesis that low-quality firms prefer to issue long-term debt, while high-quality firms issue short-term debt. However, we find no evidence consistent with the liquidity risk hypothesis. We further find that the most influential factor affecting debt maturity is firm size, followed by Z-scores and then firm age. Our findings are robust to alternative estimation specifications and endogeneity concerns. Overall, our study contributes to expanding our understanding of the debt maturity structure in private firms.

KEYWORDS:

• Debt maturity structure
• private firms
• lodging industry
• signalling hypothesis
• liquidity risk hypothesis

JEL CLASSIFICATION:

• G3
• G32

Acknowledgment

We are grateful for helpful comments from two anonymous referees and Mark P. Taylor (the editor).

Disclosure statement

No potential conflict of interest was reported by the author.

Notes

¹ Prior studies have presented the following proxy variables for firm quality: abnormal earnings (Barclay and Smith 1995; Stohs and Mauer 1996; Datta, Iskandar-Datta, and Raman 2005) and Altman’s Z-score (Brockman, Martin, and Unlu 2010; Huang, Tan, and Faff 2016). In this study, we use the latter as a proxy for firm quality because it is more directly associated with debt financing.

² Our empirical analysis focuses ‘only’ on the determinants that influence corporate debt maturity structure choice. However, the debt maturity choice is one of the decisions on corporate financial policy, including the cost of debt, the debt covenants, the choice between bank debt and public debt, and so on. Thus, further research should consider testing the joint determination of corporate financing decisions in the framework of simulation equations.

⁴ We test the validity of the GMM estimation using the Arellano-Bond test for serial correlation and the Sargan-Hansen test of the overidentifying restrictions. The null hypothesis (H₀) of the Arellano-Bond test is that there is no autocorrelation. The null hypothesis (H₀) of the Sargan-Hansen test is that overidentifying restrictions are valid. In the GMM estimation, we use the lagged explanatory variables as instruments. The result from the Sargan-Hansen test indicates that our instruments are valid.