Managerial Myopia, Earnings Guidance, and Investment*

This study investigates the real effects of management communication, specifically forecasts or earnings guidance, on investment. Managers can signal strength their projects through accuracy in guidance. leads less accurate managers to distort investments; equilibrium investment strategy involves over-investment when exceed forecast and under-investment fall short. Moreover, we find that are pessimistic forecasts, which helps explain corresponding well-documented empirical regularity. downward bias increases likelihood manipulation but decreases loss from distortion. Interestingly, induced by guidance mitigate classic problem for a myopic manager with unobservable Earnings therefore be value-increasing managerial myopia is severe. Myopie managériale, résultats prévisionnels et investissement Cette étude examine les effets réels de la communication plus précisément des prévisions ou prévisionnels, sur l'investissement. Les gestionnaires peuvent signaler force leurs projets par précision prévisionnels. Cela conduit qui sont moins précis à créer distorsions en matière d'investissements; stratégie d'investissement équilibrée implique un surinvestissement lorsque se situent au-dessus prévisions, sous-investissement au-dessous. De plus, auteurs constatent que pessimistes dans ce permet d'expliquer régularité empirique concordante déjà bien documentée. Ce biais baisse augmente probabilité investissements, mais diminue perte réelle due distorsion. Il est intéressant noter le induit contribue atténuer problème classique pour gestionnaire myope dont l'investissement inobservable. donc favoriser croissance valeur myopie managériale sévère. pervasive among US public firms. A widely held view practitioners, regulators, academics this ubiquitous practice have deleterious firm For example, CFA Institute 2006 position paper recommends firms “end providing quarterly guidance” (Krehmeyer al. 2006, 2). Similar recommendations made Fuller Jensen (2002), Chamber Commerce (2007), McKinsey consulting (Hsieh 2006). In survey executives at firms, Graham (2005) similar practitioners: “Many CFOs deplore culture giving meeting beating guided number. They argue such inhibits thinking about long-term growth, and, instead, puts too much focus targets” (42). strong evidence would sacrifice economic value order meet targets, while Bhojraj (2009) Terry (2017) document consistent taking distortive actions benchmarks.1 Survey responses noted also reveal place little reliance accruals other accounting suggesting primary distortion tool targets operational spending. Despite prevailing views regulators practitioners light evidence, there theoretical research implications The goal provide underpinnings interdependency between decisions presence myopia. finding literature that, market, who has myopic, short-run price concerns inefficiently under-invests (Bebchuk Stole 1993). contrast, distortions under light, shareholders our results an explanation number regularities offer novel testable help guide future research. model, issues firm's then makes decision prior announcement. either informed, receives perfect earnings, uninformed, noisy earnings. manager's informedness her private information. Following forecast, privately observes realized chooses level, where return not until following (final) period. We assume marginal productivity higher informed manager. captures familiar notion better knowledge operations industry conditions more accurately predict returns investments adept identifying profitable (Goodman 2013).2 While market does observe net amount devoted announced end first second final period, publicly observed. At each assessment value, determines payoff. To understand role consider benchmark setup without it. case, both types under-invest relative first-best level; here referred as second-best level. Under-investment arises because directly observable announcement; understating investment, (of type) attempts manipulate beliefs upward conveying being priced though she taken Of course, anticipates corrects it pricing firm. arising established Bebchuk (1993). cannot distinguish uninformed allows possibility separation mimicry type. continues invest However, superior information, always able match announcement previously issued As such, no intrinsic incentive forecast. equilibria type truthfully reveals issuing manager, only target altering particular, after will sometimes so cases short, level whereas often over-invests best (and even than best) unexpectedly high.3 strategically facilitate announcing incentives, helpful link By overstating must, expectation, further reduce (even level), brought forecasted number, thereby successfully mimicking cost reflected function; canonical production functions satisfy Inada concave returns, slight induces disproportionate return. entails severe efficiency losses despite clear gain successful mimicry: believes may positive probability, result period price. biasing downward, must (again, expectation) over-invest mimic informedness. Over-investment situation costly best: inefficient, introduces lower departure optimal managers. That is, becomes disproportionately greater same over-investment. generally prefers under-investment. occurs counter myopia-induced inefficiency, “undo” thus inclined under-invest. Consequently, increase chance downward. conservative estimate provides cushion Hence, equilibrium, pessimistic, downward-biased It counter-intuitive overstate inflate accordingly, contrast extant (Stein 1989). Indeed, upward-biased larger short-term gain. overstated requires heavy exposes risk devastating losses. gives up target, associated willing accept. belief unbiased Nevertheless, enjoy premium, minimizing distorting welfare hinted above, welfare-increasing if counteract improves mitigates sufficiently high. When large, cares strongly price, stronger target. Therefore, embeds excessively invests, counteracts Guidance brings expected closer resulting low, large departs deteriorating efficiency. Our show how value-destroying captured Stein (1989) (1993), beneficial shareholders, conventional wisdom practitioners. These imply uniform rule regarding applied strong, board benefit additionally perspective ex ante, observing informedness, preferences align stage. learned type, case. An strictly guidance: benefits signaling premium type's mimicry. Conversely, issue difficult Without however, pooled preference withholding pronounced heightened uncertainty. addition model predictions. include predictions variation pessimism, issuance. With respect predicts should significant incidence magnitude industries where: (i) salient; (ii) ability relatively stronger; (iii) uncertainty over earnings; (iv) correlation forecasting returns. expect (e.g., CEO turnover) likely discuss these section 5. present contributes studying myopia, spearheaded Narayanan (1985) (1989), built upon Fischer Verrecchia (2000), Guttman (2006), Edmans (2016), P. Liang Cianciaruso (2017), Marinovic Varas (2019), others. complements showing related information release. considers setting soft hard observable, Relatedly, (2016) trade-off release disclosure precise induce financial differs studies (i.e., forecast). Trueman (1986) examines high-ability receive earlier convey informational advantage timing (1986), differ signals rather allow distorts decisions. Kanodia Lee (1998) observed market. excessive discipline technology, over- under-investment, well partially informative any uninformative (1998). (2013) Y. reporting reputation considerations. costless management) probabilistic perfectly detected. Beyer mean variance process generates reported error Aghamolla (2022) likewise misreporting incentives concerned multi-firm setting. aforementioned papers bias. doing so, interdependence strategy. Ramakrishnan Wen biases probabilistically discretion control Kim multidimensional ability. connection arise influence decisions.4 structured follows. Section 2 outlines 3 4 presents results. 5 explores comparative statics discusses 6 extensions concludes. All proofs some technical details relegated supporting online Appendices B, additional presented C, D, E.5 consists three stages: stage, g, receiving incur direct g. quality depends probability ?, whereby eI, 1 ? case learns ? eU, true eU = + ?. beginning next signal, distribution noise ? 0 symmetric, continuously differentiable, single-peaked density function f(?), f?(?) ?f?(??) < all > 0. random variables ?, mutually independent. Note realization ei (i I, U) announced—the affect ultimate (discussed shortly). hence refer pre-investment keep tractable, eI uniformly distributed line ? $$ \left(-\infty, \infty \right) .6 stage (investment), realization, k ? 0.7 variable. yields gross RI(k) RU(k) Ri(k) standard conditions: continuous \left[0,\infty Ri(0) 0; \left(0,\infty , times concave, first-order derivative R i ? {R}_i^{\prime }(k) convex; goes ? k\to . I U {R}_I^{\prime }(k)>{R}_U^{\prime 0).8 deducted first-period normalized k. note specification, encompasses commonly used quadratic ?ix x2/2 scale x ?I ?U see Appendix D), Cobb–Douglas ?ik1 ?. addition, impose condition eliminate pathological situations (see A). third k, verifiable manipulated.9 unobserved individual components finalized risk-neutral prices firm, based publicized denoted P1(a, g). discount two periods g) expectation Ri(k).10 pure gI(eI), gU(?), kI(eI, g), kU(eU, ?, two, gI(eI) specify choices types, respectively. latter correspond Each knows time decision. Bayesian linear structure discussed above. Formally, (gI, gU, kI, kU) P2(a, R) that: linearity satisfied (equations (1-4)), gI, kU solves optimization (problems (5)–(7)), Ri(ki)) conditional investigate types. start straightforward remainder comprises main analysis section. informed. lack transparency interest give rise inefficient behavior. suboptimal action correctly adjusts given rational expectations. path, * {k}_I^{\ast } efficient (first-best) choice. Lemma 1.In invests e g {k}_I\left({e}_I,{g}_I\left({e}_I\right)\right)={k}_I^{\ast earning uniquely determined ? \left(1-\alpha \right){R}_I^{\prime}\left({k}_I^{\ast}\right)=1 establishes choice constant independently path gI(eI)). regard FB {R}_I^{\prime}\left({k}_I^{\mathrm{FB}}\right)=1 scaling back been previous benchmark, momentarily assumptions remain same. nature drives rationally behavior updating beliefs. guidance; k={k}_I^{\ast similarly choosing {k}_U^{\ast \right){R}_I^{\prime}\left({k}_U^{\ast}\right)=1 possible. determining sum market's posterior return: P ? {P}_1(a)=a+\left[\lambda {R}_I\left({k}_I^{\ast}\right)+\left(1-\lambda \right){R}_U\left({k}_U^{\ast}\right)\right] (See C details.) Proposition exists unique equilibrium. choose finds engage payoff M(eU;?, exceeds non-mimicry one VU(eU). VU(eU) closed interval ¯ \left[\underline{e},\overline{e}\right] mimics matches falls within interval.14 {e}_U>\overline{e} point unprofitable.15 Similarly, {e}_U<\underline{e} requisite manager.16 summarize proposition. 2.In ? ? g\in \mathrm{\mathbb{R}} \left[\underline{e}(g),\overline{e}(g)\right] \underline{e}(g)\ge Furthermore, strategies specified above (Lemma 2) determine Upon assigns event {e}_I={k}_I^{\ast }+g ( ={g}_I^{-1}(g) ) remaining uninfor