Investment-Cash Flow Sensitivities Are Not Valid Measures of Financing Constraints
回应Fazzari等人对Kaplan和Zingales的批评，指出前者实际上支持了后者的结论，即投资现金流敏感性不能有效衡量融资约束，并讨论了未解决的问题。
Work by Kaplan and Zingales provides both theoretical arguments and empirical evidence that investment-cash �ow sensitivities are not good indicators of �nancing constraints. Fazzari, Hubbard, and Petersen {this Journal} criticize those �ndings. In this note we explain how the Fazzari et al. criticisms are either very supportive of the claims in earlier work by Kaplan and Zingales or incorrect. We conclude with a discussion of unanswered questions. Fazzari, Hubbard, and Petersen {1988} (hereinafter, FHP {1988}) introduce a methodology to identify the presence of �nancing constraints based on the differential sensitivity of corporate investment to cash �ow. Kaplan and Zingales {1997} (hereinafter, KZ) provide both theoretical arguments and empirical evidence that this differential sensitivity is not a valid measure of �nancing constraints. Fazzari, Hubbard, and Petersen {2000} (hereinafter, FHP {2000}) criticize those �ndings. In this note we explain that the main arguments in FHP {2000} are, in fact, quite supportive of KZ’s main conclusion, while the speci�c criticisms in FHP {2000} are unjusti�ed. I. POINTS OF AGREEMENT FHP {2000} admit that �nancially distressed �rms are likely to have lower investment-cash �ow sensitivities than less �nancially constrained �rms. This is exactly the point that the KZ model makes: investment-cash �ow sensitivities are not necessarily monotonic in the degree of �nancing constraints. The only disagreement FHP {2000} have with KZ is how pervasive the nonmonotonicity result is. But this is ultimately an empirical question. FHP {2000} also recognize that the literature on investmentcash �ow sensitivities has not been based on a solid theoretical foundation. As KZ point out, the practice of (1) splitting the sample according to a measure of �nancing constraints and then