Douglas A. Schroeder

An Empirical-Investigation Of The Effect Of Quarterly Earnings Announcement Timing On Stock Returns

This research examines both the association between quarterly announcement timing (early or late) and the type of news (good or bad) reported, and the relationship between stock returns and timing around the earnings announcement date. Recent research on announcement timing (Givoly and Palmon [1982], Patell and Wolfson [1982], Kross [1981], and Whittred [1980]) has provided evidence that delayed announcements of annual earnings more often convey bad news (i.e., lower than expected earnings) than do early announcements. However, we know of no study which has reported evidence of the same phenomenon for quarterly earnings. Furthermore, there is a limited amount of evidence regarding the reaction of market participants to announcement timing. While three studies (Givoly and Palmon [1982], Kross [1982], and Chambers and Penman [1984]) find that early (late) announcements are associated with higher (lower) abnormal returns or high (low) stock return variability, relative to late (early) announcments, only Kross [1982] controlled for the sign of the earnings forecast error and none controlled for the magnitude of the earnings forecast error. It is well accepted that stock returns are associated with the sign of the earnings forecast error (EFE). Since announcement timing is also

Inferring Transactions from Financial Statements

In this paper, we embed the double entry accounting structure in a simple belief revision (estimation) problem. We ask the following question: Presented with a set of financial statements (and priors), what is the reader’s “best guess” of the underlying transactions that generated these statements? Two properties of accounting information facilitate a particularly simple closed form solution to this estimation problem. First, accounting information is the outcome of a linear aggregation process. Second, the aggregation rule is double entry.

Reconciling Financial Information at Varied Levels of Aggregation

Financial statements summarize a firm’s fiscal position using only a limited number of accounts. Readers often interpret financial statements in conjunction with other information, some of which may be aggregated in a different way (or not at all). This paper exploits properties of the double-entry accounting system to provide a systematic approach to reconciling diverse financial data. The key is the ability to represent the double-entry system by network flows and, thereby, access well-recognized network optimization techniques. Two specific uses are investigated: the reconciliation of audit evidence with management-prepared financial statements, and the creation of transaction-level financial ratios.

Discrete choice models

Choice models attempt to analyze decision marker’s preferences amongst alternatives. We’ll primarily address the binary case to simplify the illustrations though in principle any number of discrete choices can be analyzed. A key is choices are mutually exclusive and exhaustive. This framing exercise impacts the interpretation of the data.

Bayesian treatment effects

We continue with the selection setting discussed in the previous three chapters and apply Bayesian analysis. Bayesian augmentation of the kind proposed by Albert and Chib [1993] in the probit setting (see chapter 5) can be extended to selection analysis of treatment effects (Li, Poirier, and Tobias [2004]). An advantage of the approach is treatment effect distributions can be identified by bounding the unidentified parameter. As counterfactuals are not observed, the correlation between outcome errors is unidentified. However, Poirier and Tobias [2003] and Li, Poirier, and Tobias [2004] suggest using the positive definiteness of the variancecovariance matrix (for the selection equation error and the outcome equations’ errors) to bound the unidentified parameter.

Repeated-sampling inference

Much of the discussion regarding econometric analysis of endogenous relations centers around identification issues. In this chapter we review the complementary matter of inference. Exchangeability or symmetric dependence and de Finetti’s theorem lie at the heart of most (perhaps all) statistical inference. A simple binomial example illustrates.

Marginal treatment effects

In this chapter, we review policy evaluation and Heckman and Vytlacil’s [2005, 2007a] (HV) strategy for linking marginal treatment effects to other average treatment effects including policy-relevant treatment effects. Recent innovations in the treatment effects literature including dynamic and general equilibrium considerations are mentioned briefly but in-depth study of these matters is not pursued. HV’s marginal treatment effects strategy is applied to the regulated report precision setting introduced in chapter 2, discussed in chapter 10, and continued in the next chapter. This analysis highlights the relative importance of probability distribution assignment to unobservables and quality of instruments.

Treatment effects: IV

In this chapter we continue the discussion of treatment effects but replace ignorable treatment strategies in favor of instrumental variables and exclusion restrictions. Intuitively, instrumental variables are a standard econometric response to omitted, correlated variables so why not employ them to identify and estimate treatment effects. That is, we look for instruments that are highly related to the selection or treatment choice but unrelated to outcome. This is a bit more subtle than standard linear IV because of the counterfactual issue. The key is that exclusion restrictions allow identification of the counterfactuals as an individual’s probability

Treatment effects: ignorability

First, we describe a prototypical selection setting. Then, we identify some typical average treatment effects followed by a review of various identification conditions assuming ignorable treatment (sometimes called selection on observables). Ignorable treatment approaches are the simplest to implement but pose the strongest conditions for the data. That is, when the data don’t satisfy the conditions it makes it more likely that inferences regarding properties of theDGPare erroneous.

Overview of endogeneity

As discussed in chapter 2, managers actively make production-investment, choices. These choices are intertwined and far from innocuous. Design of accounting (like other information systems) is highly dependent on the implications and responses to accounting information in combination with other information. As these decisions are interrelated, their analysis is inherently endogenous (Demski [2004]). Endogeneity presents substantial challenges for econometric analysis. The behavior of unobservable (to the analyst) components and omitted, correlated variables are continuing themes.