Do Firms Engage in Risk-Shifting? Empirical Evidence

Erik Gilje is Assistant Professor of Finance at The Wharton School of the University of Pennsylvania. This post is based on his recent paper.

How does corporate investment risk-taking change when a firm has high leverage or approaches distress? In high-leverage states of the world, equity holders benefit from successful outcomes of high-risk projects, while losses from unsuccessful outcomes are borne by debt holders. This asymmetry between who receives the gains and losses from a project could make it optimal for equity holders to maximize the amount of risk a firm undertakes when leverage is high. This hypothesized increased risk-taking in a firm’s investments, referred to as risk-shifting or asset substitution, could result in an overall cost to the firm (Jensen and Meckling (1976)).

Concerns about the size, prevalence, and mitigation of these costs have been the focus of substantial theoretical work. [1] However, there is little empirical evidence on the size or pervasiveness of changes in investment risk-taking when a firm approaches financial distress. The empirical challenges are twofold. First, obtaining a measure of the riskiness of a firm’s overall capital expenditures is challenging in most settings. Second, financial distress is not randomly assigned to firms. To the extent investment policies and financial policies are jointly determined, or are driven by an omitted variable, obtaining clean identification of the effect of excessive leverage on risk-taking is problematic. The contribution of this paper is to provide empirical advancements on both these fronts. First, I focus on a setting where a firm’s investment risk-taking is clearly defined by measures of investment risk from Securities and Exchange Commission (SEC) disclosures. Second, I use quasi-random shocks to leverage to identify the effect of an increase in leverage and distress on investment risk-taking.

I use a setting in which investments can be categorized into two different types of activities, one that is high risk and one that is low risk. To do this, I focus on the oil and gas industry, where exploratory projects (high risk) are nearly six times more likely to result in an unproductive project than development projects (low risk). [2] Moreover, these categories have clear definitions outlined by the Financial Accounting Standards Board (FASB) and are disclosed in SEC filings. Therefore, there is a standardization in these measures across firms and over time that is typically unavailable in other settings. I construct a data set from hand-collected data on investment risks from the 10-Ks of 184 firms in the oil and gas industry. Using these risk disclosures, I test how the proportion of high-risk investment to total investment changes as leverage increases and firms approach distress.

Contrary to what risk-shifting theory would predict, I find that firms reduce risk-taking as they approach distress. I find similar results in both a natural experiment setting and firm-level panel regressions. In firm-level panel regressions, I find that a one-standard-deviation increase in leverage reduces the proportion of a firm’s high-risk investment to total investment by 8.5% relative to the mean level of firm risk-taking. I also find that the proportion of high-risk investment to total investment is reduced by 21.6% for firm-years in which leverage is in the top quartile of the sample. Furthermore, this risk-reducing behavior also occurs in the years prior to declaring bankruptcy. To mitigate simultaneity and omitted variable endogeneity concerns, my main identification strategy relies on a natural experiment to test how risk-taking changes with leverage during two significant commodity-based negative leverage shocks in 1998 and 2008. Using a difference-in-differences approach, I find that treatment firms reduce investment risk-taking relative to control firms.

Why might firms reduce risk-taking in distress? Firms could have risk-mitigating incentives that outweigh risk-shifting incentives. In the natural experiment setting, I find that risk reduction is most prevalent in firms with shorter maturity debt and bank debt. This suggests that debt composition is important for firm risk-taking in distress, and provides support for risk-reducing incentives and monitoring linked to banking relationships.

To further explore the role of banks in risk-reduction, I assess the role of bank financial covenants. Banks in my sample do not place explicit covenants on exploration activity; however, financial covenants may allow the bank to exert some indirect control on firm risk-taking activity. To test for this, I hand-collect covenant data from credit agreements of firms in my quasi-natural experiment. I find that risk reduction behavior is most prevalent among firms that have stricter financial covenants and more financial covenants prior to the shock. This result provides new empirical evidence that financial covenants may allow banks to exert indirect influence to reduce debt-equity agency conflicts that have not or cannot be explicitly contracted on, such as risk-shifting.

Whether firms engage in risk-shifting has been an open empirical question. Lack of data and adequate measures of risk, and the endogeneity of leverage and risk-taking have meant this question has not been able to be addressed directly. I use a setting which has quasi-random shocks to leverage and objective measures of investment risk, from SEC disclosures, to test whether firms engage in risk-shifting. I find that firms reduce risk, rather than increase risk, when leverage is high and when they get close to distress.

Prior theoretical literature outlines several reasons for why firms may have incentives to reduce risk-taking in distress. Firms likely have incentives to ensure that they have a good reputation to ensure access to debt markets (Diamond (1989)), which can affect their ability to pursue future positive NPV projects Almeida et al. (2011). I am able to highlight channels linked to debt composition and financial covenants as being important for risk reduction in times of distress. I show that these mechanisms are important for mitigating risk-shifting, and serve to mitigate debt-equity agency conflicts that may not be explicitly contracted on. The evidence in this paper suggests that risk-mitigation incentives and monitoring by banks outweigh risk-shifting incentives in investment decision making for the average firm.

The full paper is available for download here.


Almeida, H., Campello, M., Weisbach, M. S., 2011. Corporate financial and investment policies when future financing is not frictionless. Journal of Corporate Finance 17, 675–693.

Barnea, A., Haugen, R. A., Senbet, L. W., 1980. A rationale for debt maturity structure and call provisions in the agency theoretic framework. Journal of Finance 35, 1223–1234.

Diamond, D. W., 1989. Reputation acquisition in debt markets. Journal of Political Economy 97, 828–862.

Green, R. C., 1984. Investment incentives, debt, and warrants. Journal of Financial Economics 13, 115–136.

Jensen, M. C., Meckling, W. H., 1976. Theory of the firm: Managerial behavior, agency costs, and ownership structure. Journal of Financial Economics 3, 305–360.

John, T. A., John, K., 1993. Top-management compensation and capital structure. Journal of Finance 48, 949–974.

Smith, C. W., Warner, J. B., 1979. On financial contracting an analysis of bond covenants. Journal of Financial Economics 7, 117–161.


[1] Existing theoretical work related to the size and mitigation of risk-shifting includes: Smith and Warner (1979) (covenants), Green (1984) (convertible debt), Barnea et al. (1980) (debt maturity), and John and John (1993) (managerial compensation).
(go back)

[2] The firms in my sample drilled a total of 12,574 exploratory wells, of which 3,326 were unsuccessful (26.4%), and drilled 88,277 development wells, of which 3,809 were unsuccessful (4.3%). Additionally, in comparing reserve additions from discoveries relative to exploration capital expenditures, in 27% of all firm-years, firms failed to add reserves through discoveries that exceeded their exploration spending.
(go back)

Both comments and trackbacks are currently closed.