Office Hours: Who's Afraid of Performance Reporting?
Kathy Petroni (Michigan State) will be joining us June 10th, 11am ET, to talk about her paper “Comprehensive Income: Who’s Afraid of Performance Reporting?”. The paper, co-authored with Linda Bamber, John Jiang and Isabel Wang, now forthcoming in The Accounting Review, touches on some timely and controversial topics.
For starters, let me say that the primary result is not that surprising to me: managers don’t like putting volatile numbers (such as gains and losses from “available for sale” securities) right into the face of their income statement, especially when they have more equity-based incentives and less job security. Instead, they tuck this information into the Statement of Changes in Shareholders Equity, which in my own research I have learned don’t attract nearly as much attention as the income statement.
But here are some open questions for future research that I expect we will discuss tomorrow:
- First, could the managers be wrong about what investors want? People avoid flying because they think of it as dangerous, even though driving the same number of miles is typically more dangerous. Perhaps managers are avoiding a phantom risk.
- Second, could the FASB and IASB be wrong about what investors want? It isn’t just a matter of standard setters wanting all comprehensive income items shown on the face of the income statement. More generally, the Boards have largely bought into an ‘asset-liability approach’ that treats assets and liabilities as primary, and comprehensive income is just a plug figure equal to the change in net assets. Apparently, managers believe that investors choose an income approach, in which the primary goal of financial reporting is to generate meaningful measures of income–even if it takes putting items on the balance sheet that aren’t assets or liabilities, but just plugs needed to get a smoother, more meaningful income number. If the managers are right, is the Board making a mistake by providing investors with statements that give a more meaningful balance sheet but a more volatile income statement?
There are plenty of other issues to talk about, and I hope you will join us–and perhaps post your own questions in the comment thread below.
Details about office hours are available here.
A maintained assumption of this paper is that reporting comprehensive income in the IS is “more transparent” and “more desirable” from investors’ perspective (otherwise, there is no null hypothesis in this paper). Based on this assumption, the authors argue and find evidence that managers allegedly with more incentives to be less transparent (those with high equity incentives and less job security) will choose to “bury” comprehensive income in the statement of changes in equity. I have two questions regarding the assumption and interpretation of results (with my thoughts) below.
1. Is reporting comprehensive income in IS more transparent and desirable from investors’ perspective?
Given the higher volatility of comprehensive income relative to net income, would it be more desirable to separate the two numbers? Also, if comprehensive income includes transitory and uncontrollable items, and is an incomplete measure of the firm’s performance (the authors’ discussion on p. 6), then investors might prefer “not to see” that number or at least not to see that number in the same place with net income (which might confuse them). The bottom line is that it is unclear whether it is more or less transparent to present OCI on IS relative to SCE. I am aware of at least two studies that suggest that presenting comprehensive income in SCE might be “more transparent” from investors’ perspective.
• Chambers, Linsmeier, Shakespeare, and Sougiannis (2007RAST) find that investors price OCI when it is presented in SCE but not when it is presented in IS, although the differential pricing is not statistically significant.
• Bradshaw, Miller, Serafeim (2009WP) find that firms that use accounting methods that differ from their peers have higher analyst forecast errors, presumably because analysts find it more difficult to process financial information reported in “atypical” ways. In the context of OCI reporting, given 80% firms report OCI in SCE, it turns out that reporting OCI in IS would be “atypical”, which might make it harder for FS users to interpret the number.
2. What can we take away from the results?
Although not stated explicitly, the authors seem to interpret that equity incentives and job security concern motivate managers to reduce financial reporting transparency (opportunistic view). This interpretation is reasonable if we rely on the assumption that reporting OCI in the SCE is less transparent. However, if one believes that it is more transparent to report OCI in the SCE to avoid investors’ confusion (question 1 above), then the results might be interpreted differently: equity incentives and labor market forces encourage managers to present financial statements in ways that enhance transparency (altruistic view). This interpretation is consistent with evidence in Nagar, Nanda, and Wysocki (2003JAE) that higher managerial stock-based incentives are associated with higher quality voluntary disclosures.
While BJPW ask why “location” matters, I still ask the same question after reading their paper.
First, BJPW argue that managers with stronger equity-based incentives are less likely to use performance reporting. I agree that comprehensive income is more volatile, which may in turn hurt stock prices. However, what we are interested in is not the impact of comprehensive income on stock prices but the impact of comprehensive income reporting location on stock prices. The answer to the latter is not clear, at least to me. Indeed, as BJPW point out as well, future research could assess investors’ actual reactions to comprehensive income disclosures. Suppose that performance reporting makes comprehensive income more salient, and that in a hypothetical bad scenario there is a negative market reaction; I want to understand whether such a negative price impact faced by performance reporting firms is temporary. If this is the case, why should managers care the reporting location? Even though some of them may have large equity incentives, they probably cannot cash in those incentives in a short time period, possibly due to vesting period, etc. Let me play devil’s advocate, in an extreme, why don’t those managers prefer performance reporting? Aboody and Kasznik (JAE 2000) find that managers rush (delay) disclosing bad (good) news before the option grant date to lower share prices and in turn to maximize their stock option compensation. Could managers that choose performance reporting also take advantage of the temporary market negative reaction (if there is one)?
Second, BJPW argue that managers with less job security are less likely to use performance reporting. While performance reporting may have negatively impact on investors’ perceptions, possible due to “inattention” (Hirshleifer and Teoh, JAE, 2003), why will the boards who decide managers’ compensation and turnover also be “fooled” in a same way as the equity investors? Don’t they have any private insights?
Thanks to Nam and Li for posting interesting thoughts on the paper. I will try to respond to the issues they raise.
First, we adopt the view that performance reporting is more transparent because this is the perspective that the FASB has on the issue. We also define a transparent presentation as one that is easily recognized or detected, which we believe is consistent with performance reporting given that many experimental studies suggest that investors are much more likely to see comprehensive income and incorporate it into their decisions if it is in a performance statement
But I think Nam’s point is broader than the issue of how to define transparency. I believe that Nam is arguing that an alternative explanation for our results is that given investors’ limited attention investors may prefer that CI be reported less saliently and that managers with high equity incentives and low job security are more concerned about pleasing investors so they report CI less saliently. This is an interesting point that we haven’t considered. We don’t think this is likely however given that we have no evidence that users are demanding CI be reported in a statement of equity. Specifically, based on our review of comment letters all of the opposition to performance reporting is coming from managers not from users. The two user groups that wrote comment letters that addressed the location of comprehensive income (Robert Morris Associates and the Fund for Shareholder Rights) both argued for performance reporting.
We agree with Li that it would be very interesting to study the effects of location choice on stock prices. That is not the intent, however, of our study. We are interested in the determinants of the location decision.
I believe Li is also making the argument that managers may use the location decision to manage short run perceptions of the firm because stock price responses to the location decision may be temporary. Given that the decision is quite sticky and it is difficult for managers to predict comprehensive income, we don’t believe that is likely.
We agree with Li that boards are not likely fooled by the CI reporting location. But we are not necessarily arguing that the CEOs believe that Board members were subject to limited attention which would make them more likely to overlook CI if it is reported in the statement of equity. Rather, all that is necessary is that managers are concerned that more salient performance reporting could hurt stock price, which in turn would reflect poorly on the CEO’s performance (see our footnote 18).
Thanks again for your thoughtful comments.
I think BOD will not be fooled. Sophisticated investors will not be fooled (in the office hour today, Dr. Bloomfield mentioned that his talks with credit rating agencies confirm this). But I don’t think managers want to cheat these guys (assuming that managers put CI in the SCE to “hide” this number – which not everyone agrees). These guys are smart enough to find CI in the SCE, but they are also smart enough to understand that a large component of CI is transitory and uncontrollable by managers. So location does not matter for these guys.
However, there might be naive investors out there that are not sophisticated enough to understand the nature of the volatility in CI. These guys might react “irrationally” to CI if presented in IS (again, we have to assume that hiding CI in the SCE makes it more difficult for naive investors to find the number). If managers believe that these unsophisticated guys can influence stock prices in a period that is long enough (so that they have to sell their shares at low prices or might lose their jobs before the prices recover), then they will try to hide CI somewhere so that naive investors will not react to the number (in the paper’s context, they hide it in the SCE). Of course, anyone who believes in the EMH probably will argue that sophisticated investors will jump in and correct the prices immediately. But it is the managers’ belief that matters, even thought their belief might be wrong (that’s why the authors suggest more studies on the market reaction).
Kathy Petroni led a thought provoking session on corporate performance reporting. At issue is why managers choose to report comprehensive income in different manners under SFAS 130. Kathy discussed key elements of her paper which suggests that CEOs with strong incentives to maintain their stock price and/or CEOs who have low job security are more likely to put comprehensive income in a statement of shareholder equity rather than in the income statement.
One issue was that if all the participants have complete information, rational expectations would suggest that location of a particular number in the financial statements is unimportant. Kathy admits that the market might not pay attention to the location of comprehensive income. However, what matters for her study is what managers believe the market pays attention; this belief is sufficient to cause managers to alter their financial reporting choices.
Another issue was why some on the FASB and IASB prefer a single statement of income. One observation is that no clear theory exists for deciding which the changes in a company’s net assets should be reported in earnings versus parked in other comprehensive income (OCI). Office hour participants speculated that some standard setters are uncomfortable with reporting OCI items in a fundamentally different manner from other changes in net assets.
Another question regarded whether managers preferred to keep OCI out of the income statement in order to manage earnings (e.g., cherry picking gains from available for sale securities) or to avoid income volatility. This evolved into a discussion of whether the single income statement approach combined persistent and non-persistent income components, which could be detrimental if investors naively attributed the same level of persistence to all income components. For example, Nam Tran’s comment above suggests that perhaps investors want an income statement that is not cluttered with items of OCI, in which case the choice to report OCI in a statement of equity does not appear to be opportunistic. This led to a side discussion of the asset-liability versus income view of financial reporting. I believe that a lot of misconceptions exist regarding what the FASB means by its asset and liability view. Also, as Rob says above, if one takes a strict “make sure reported earnings represents permanent income” view, then managers must have the discretion to book a variety of debits and credits that do not meet most of our definitions of assets or liabilities. Interestingly, many of the commentators who appear to champion the income approach also dislike giving managers a lot of discretion in financial reporting. As I said, I think this “battle” between asset liability view versus income view is fraught with misconceptions and misunderstandings. Perhaps discussions like the one we had today in office hours will help clear up this issue.
The session was very well attended, and I hope everyone enjoyed it as much as I did.
I kind of agree with Nam Tran’s point that the location of CI should not matter to a sophisticated investor, consistent with the theory of rational expectation, although for a naive investor, it may matter. Given that for many of the publicly traded companies today, a large part of their outstanding shares are held by sophisticated investors such as mutual funds, pension funds etc., I wonder why do the managers will think about the ‘perception’ of naive investors and not the sophisticated ones?
Thus, unless we can show that the stock price of a firm does indeed drop if CI is placed in IS instead of SCE, which is not the focus of this paper, we will be tempted to argue for alternative explanations, a few of which have already been mentioned.
I will close this by saying that these alternative explanations do not refute the story of the paper as long as we maintain the assumption that managers ‘perceive’ that stock price will drop if CI is placed in IS, even though the perception may be wrong.
I’d like to thank everyone for their interest in our paper!
As the above comments point out, our hypothesis that managers believed the location of information would affect users’ perceptions is controversial, because it is at odds with the traditional rational economic-based perspective in which many of us have been trained. Nam’s second comment provides an outline of our thinking, but given the additional questions Sanjay poses, perhaps it might be helpful if I try to explain our reasoning in more detail.
There are several reasons we hypothesize that managers likely believed the location where the firm reported comprehensive income could affect investors’ perceptions.
For those who find it hard to accept the story behind our hypotheses, note this is a difference of opinion about the motivation for the paper’s tests, not a controversy about the empirical results themselves. We hope that even skeptics might find it interesting to learn that managers with greater equity-based incentives and less job security are more likely to ignore policymakers’ stated preference by reporting comprehensive income in the statement of equity, even if they don’t buy our reasoning as to why these particular managers do so.