Managerial Investment and Changes in GAAP

In my paper, Managerial Investment and Changes in GAAP: An Internal Consequence of External Reporting, which was recently made publicly available on SSRN, I investigate whether changes in Generally Accepted Accounting Principles (GAAP) affect corporate investment decisions. I hypothesize that the relation between changes in GAAP and investment manifests for at least two non-mutually exclusive reasons. First, I hypothesize that changes in GAAP can affect investment because the numbers reported in financial statements have a direct bearing on contractual outcomes. For example, debt contracts often contain covenants based on numbers reported in financial statements (Leftwich [1983]). Consequently, if a change in GAAP has an unfavorable (favorable) impact on current and future financial statements, and debt covenants are not adjusted to incorporate the changes, the change in GAAP will likely tighten (loosen) covenant slack. As a result, managers may alter their actions to avoid covenant violation. Specifically, since most investments have an uncertain future outcome and some positive probability that the outcome is a loss, they increase the probability of violating covenants in the future by adversely impacting future financial ratios. Consequently, managers might respond to changes in GAAP that adversely affect financial statements by cutting investment in risky assets with the goal of preserving net worth and preventing deterioration of financial ratios.

Second, I propose that changes in GAAP cause managers to collect and process additional information to comply with the new standard, which changes their information set and their subsequent decisions. One of the primary objectives of GAAP is to provide investors with information about firms’ future cash flows to facilitate their investment decisions. Managers, like investors, evaluate investment opportunities by forecasting and discounting the expected future cash flows from investments. Therefore, preparing financial accounting statements for external investors can have a spillover effect on managers’ information sets by requiring managers to assimilate information to comply with GAAP. Although managers have access to more detailed and timely information than that reported in financial statements, managers have limited attention and are unlikely to process all the information available within the firm (Simon [1973]). Consequently, some changes in GAAP that obligate managers to collect and process additional information to implement the new rule can incrementally inform managers about the future cash flow consequences of their decisions. For example, Singh [2001] quoted Ben Neuhausen, a partner in Arthur Andersen’s professional standards group, as saying, “I think some companies were genuinely clueless about how much these benefits were going to cost them over the long haul…once Statement 106 [post-retirement benefits] forced them to measure these obligations, a lot of companies realized that they had offered benefits they could not afford.” I hypothesize that changes in GAAP that inform managers that they overestimated (underestimated) the future cash flows and net present values (NPV) of their investment decisions cause managers to decrease (increase) investment. For example, if the adoption of SFAS 106 (post-retirement benefits) informed managers that they underestimated the cost of employees, this information is likely to cause a downward revision in NPV estimates, turning some previously positive NPV projects into negative NPV projects. Any such revision in managers’ NPV estimates is likely to decrease total investment.

I test whether changes in GAAP affect investment using a sample containing forty–nine mandatory accounting rule changes implemented between 1991 and 2007. I measure investment as capital expenditure and research and development (R&D) expenditure, and I use the cumulative effect of an accounting change to measure the impact an accounting rule change has on firms’ earnings and book equity. My results show that the cumulative effect of an accounting change is positively related to both capital and R&D investments, consistent with my primary prediction that changes in GAAP affect investment decisions.

To test the contracting hypothesis, I examine whether firms with covenants in private debt agreements are more likely to change investment in response to a change in GAAP. I find that changes in GAAP affect both capital and R&D investments in the presence of financial covenants. Additional tests reveal that changes in GAAP affect investment via its effect on covenants only for firms with contracts that are affected by the change in GAAP (i.e., contracts based on floating GAAP). In contrast, when debt contracts explicitly disallow accounting changes from influencing the computation of covenants, I find no relation between changes in GAAP and investment. I also find that changes in GAAP have a larger impact on investment when borrowers are likely to find renegotiating the debt contract costly.

To test the information hypothesis, I exploit the variation in the nature of the changes in GAAP. Specifically, I classify changes in GAAP into two groups based on its likelihood of providing managers with information. For example, SFAS 123R (expensing stock options) required firms to recognize the cost of employee stock option in earnings when the prevailing rule required firms to only disclose the cost. Since such information was already disclosed in financial statements, compliance with SFAS 123R is unlikely to affect managers’ information sets. In contrast, compliance with standards such as SFAS 106 (post-retirement benefits) and SFAS 143 (asset retirement obligation) are likely to affect managers’ information sets because they required managers to collect and process significant amounts of additional information not required previously. The information hypothesis suggests that only cumulative effects arising from changes in GAAP that are likely to inform managers are associated with investment.

Consistent with the information hypothesis, I find that the relation between the cumulative effect of an accounting change and investment is positive and statistically significant only when the change in GAAP is likely to inform managers. This finding suggests that changes in GAAP affect investment by changing managers’ information sets. In additional analyses, I examine whether changes in GAAP led to an improvement in investment efficiency. If a change in GAAP provides managers with incremental information that is relevant for decision making, the change in GAAP is likely to increase investment efficiency by reducing both over- and under- investment. Consistent with this prediction, I find that investment efficiency improves following changes in GAAP that are more likely to inform managers.

The full paper is available for download here.