Numbers generally convey a sense of certainty – especially in accounting and finance. However, this perceived precision also makes it easier to use them in misleading ways. An extensive literature in political science and mass communications discusses how numbers, taken out of context, can bolster weak arguments. For instance, mortality rates are often cited to argue for certain foreign and public policies because they elicit strong feelings. However, mortality rates are complex estimates with large margins of error.
In financial disclosure, firms can use numbers to mislead and create a false impression of a firm’s outlook – potentially to the detriment of investors. In a new paper, we investigate this phenomenon. Our study is motivated by the increasing prominence and use of unconventional numbers in firm disclosures. Numbers from annual reports are widely scrutinized and even partially anticipated before their official release. Therefore, investors increasingly search for alternative numbers (i.e., non-financial data, alternative ratios, and statistics) to evaluate a firm’s outlook. Studies have found that the increased quantity of numerical data in financial disclosure can be beneficial for a variety of stakeholders. However, multiple studies also show that the flexibility of conference calls can be strategically used to deceive and manage the message. Therefore, we posit that, while numbers in earnings conference calls are generally provided for the benefit of investors, there is also an element of strategy and deception at play.
We examine the use of numbers in earnings conference calls and focus on capturing instances where numbers lack enough context. By omitting the proper context, a speaker only implicitly suggests that the numbers are appropriate, whereas stating the wrong context is tantamount to lying. Thus, rather than the accuracy of the context itself, we focus on and measure the proportion of numbers shared relative to the context offered in firms’ earnings calls. We hypothesize that using numbers without a full and proper context in earnings conference calls generates short-lived positive outcomes (i.e., numbers generally convey a sense of certainty) with predictably negative consequences for the future. Using a sample of over 39,000 conference call transcripts, we find that, when firms present numbers in earnings conference calls with limited context, they experience positive cumulative abnormal returns in the three-day window around the dates of the calls. Specifically, one fewer context-relevant word for every number provided corresponds to a 1.6 percent increase in cumulative abnormal return over the three-day window. However, for the 60-day window after the earnings conference calls, these same firms experience negative cumulative abnormal returns. Specifically, one fewer context-relevant word for every number provided corresponds to a 0.7 percent decrease in cumulative abnormal return over the 60-day window following the call. These results are consistent with our argument that managers use numbers out of context in conference calls to boost investors’ perceptions of the reports beyond their actual merit. While investors are initially misled by this strategy, the initial effect is reversed over a longer horizon.
In addition, we find that, when managers use numbers with limited context in conference calls, analysts revise their forecasts upwards and ask more positive questions during the calls. These results are consistent with limited-context numbers in conference calls generating short-term benefits by conveying a more certain outlook for the firms to key stakeholders. Finally, we document that presenting numbers without enough context is associated with a decrease in future earnings and a higher risk of restatements, which suggests that doing so in conference calls signals other long-term consequences for firms. That is, we analyze the implications for future firm performance and outcomes, suggesting that managers, knowing the problems ahead for the firm, are likely using more numbers with limited context when they are aware of (future) negative information.
In additional analyses, we find that, when managers don’t provide proper context for numbers, they also give favorable analysts a bigger spotlight in the Q&A. This corroborates the use of undercontextualized numbers to intentionally intervene in the financial reporting process. Furthermore, we provide evidence that this intervention benefits managers. We document that, after managers use numbers with limited context in conference calls, they are more likely to sell their shares in the firm in the period immediately after the calls. If managers are aware that this strategy is short-lived, and have private information about the future firm underperformance, they have incentives to take advantage before the longer-term reversal.
As the business environment evolves, firms have an increasing variety of ways to disclose information, including press releases, earnings calls, blogs, and social media, giving them more freedom and discretion in how they present numeric or qualitative data. Our study suggests that numerical data in qualitative disclosure is not always useful or accurate. Furthermore, our study validates concerns that have been raised in other disciplines about the perils of our hyper-numeric world and applies them to the study of financial disclosure. The proliferation of the internet has fundamentally altered the supply and demand of quantifiable data. Numbers change the way we participate in politics, appreciate sports, and engage in relationships. However, the concern is that our ability to produce numbers has advanced ahead of our ability to comprehend them.
Our study shows how numbers can be used to influence investors’ evaluation of firms’ profitability, position, and risk. The over-reliance on numbers is subtly encoded into social norms and cultures over time, even when some of these numbers are questionable. Therefore, our study raises the possibility that the propagation of numbers and the associated lack of oversight in certain financial disclosures might eventually change the way capital markets operate and are regulated.
This post comes to us from professors Kris Allee at the University of Arkansas and Chuong Do at the University of Nevada – Reno and from Huy Do, a PhD candidate at the City University of New York – Baruch College. It is based on their recent article, “Contextualization of Numbers in Earnings Conference Calls,” available here.