Purchase Obligations, Earnings Persistence and Stock Returns
- Author(s): Lee, Kwang June
- Advisor(s): Dechow, Patricia M
- et al.
This dissertation examines whether purchase obligations data disclosed in the MD&A section of 10-K filings are useful in predicting and understanding firm performance. The SEC defines a purchase obligation as an enforceable and legally binding agreement to purchase goods and services in the future. Consequently, by entering into purchase obligations, firms procure resources that will be used as production inputs, but they also have an obligation to make payments to their suppliers. Therefore, purchase obligations have aspects of both assets and liabilities.
When the SEC first introduced a rule requiring the disclosure of purchase obligations, the SEC's main concern was the liability aspect of purchase obligations. Thus, a firm's purchase obligations were viewed by the SEC as having a potential negative impact on the firm's future liquidity. However, the level of a firm's purchase obligations depends on the firm's degree of outsourcing, and the growth in a firm's purchase obligations indicates that the firm expects to use more production resources in the future period. Therefore, an alternative view is that an increase in purchase obligations leads to better future operating performance. Given these conflicting views, this dissertation examines the implications that annual change in a firm's purchase obligations has for its future operating performance.
The first essay provides background information on purchase obligations. We begin by explaining the details of the SEC rule which requires the disclosure of purchase obligations. Then, we describe data collection and sample formation procedure, and provide examples of contractual obligations disclosure. After that, we summarize the types and amount of purchase obligations reported by our sample firms. Also, we document the strong persistence over time in the amount of purchase obligations as well as the short-term nature of purchase obligations. Finally, we discuss the differences between purchase
obligations and on-balance-sheet assets, which lead to the main results of this work.
The second essay examines the implications of annual changes in purchase obligations for future operating performance and asset growth. We predict that firms enter into additional purchase obligations when they expect an increase in demand for their products. Consistent with this prediction, we find annual change in purchase obligation is positively associated with future sales and earnings. The results of DuPont analysis shows that annual change in purchase obligation is positively associated with future asset turnover, but it is insignificantly associated with future profit margin. Additionally, we find that annual change in purchase obligations is positively associated with both contemporaneous and future growth in on-balance-sheet assets, particularly fixed assets. These results suggest that the disclosure of purchase obligations provides useful information to investors in predicting future performance and identifying growth stage.
The last essay examines whether equity investors and analysts fully incorporate the information contained in the disclosure of purchase obligations. First, we find that annual change in purchase obligations is positively associated with both contemporaneous and future annual stock returns. This suggests that although investors appear to appreciate the value-relevance of the information in purchase obligations before 10-Ks are filed, they do not fully incorporate the resulting implications of the information in determining stock prices. Second, we find that analyst forecast errors are positively correlated with annual change in purchase obligations in the periods following the filing of a 10-K. This suggests that analysts fail to fully incorporate the information in purchase obligations when they forecast one-year-ahead earnings. This also suggests that the delayed stock price response to the information in purchase obligations can be at least partially explained by analysts' inability to fully incorporate the information.