Financial Leverage, Information Quality, and Efficiency*

We examine information quality and financial leverage when an entrepreneur needs financing to undertake a risky project his effort input affects the project's outcome. show that interact play active roles in both investment decisions. Our analysis shows positive association between optimal quality—when is low (high), (high) optimal. This because with leverage, already motivated by large share of outcome exert effort, high not efficient as precise bad signal discourages entrepreneur's effort. In contrast, thus less residual cash flows, optimal, good encourages study highlights joint effect on overall efficiency through firms' inputs well defining efficiency. Levier financier, qualité de l'information et efficience Les auteurs examinent la le levier financier lorsqu'un besoin financement pour entreprendre un projet risqué que son affecte résultat du projet. Ils montrent interagissent jouer des rôles actifs dans les décisions d'investissement d'effort. Leur analyse montre une entre optimale l'information—lorsque est faible (élevé), (haute) optimale. En effet, lorsque faible, l'entrepreneur déjà motivé, par importante part résultat, à fournir efforts, haute n'est pas efficiente, car mauvais précis entrave l'effort l'entrepreneur. revanche, élevé donc moins motivé ses flux trésorerie résiduels, optimale, bon encourage Cette étude met en évidence l'effet conjoint sur l'efficience globale travers entreprises ainsi définition l'investissement. this study, we effects accounting decisions, how influences choices. Information are important they impact real decisions such investments, contracting, inputs; however, their those remain under-explored. addition, there no clear empirical evidence leverage.1 model setting which external from representative creditor competitive debt market project. For convenience, refer “he” “she.” The can be either or bad, directly observable; public reveals type noise. chosen ex ante entrepreneur. After observing signal, if decides pursue project, he seeks funds creditor. If willing provide funds, she requests repayment break even. Once undertaken, may further productive improve flow. quality. That is, low, whereas high. When case, induces request larger repayment, discouraging exerting smaller also analyze prior probability very favorable, effectively commit choosing quality, However, commitment device longer works, even lowest cannot induce conditional signal. Hence perfect moderate, intermediate level unfavorable. efficiency, represented expected payoff, more sensitive poor. With poor creditor's requested barely affected noisy signal; rather, it determined level. Regardless realization, induced shirk high, makes leverage. helps distinguish projects, determines her only based but Thus, our setting, A should always one still worthy exerts regarding role these two particular, decision jointly while (which leverage). determine quality: so higher has distinguishing projects improves efficiency; intermediate, interacts shaping decision, negative impacts twofold: hand, types exerted; other reveal discourage decreases would lead net present value (NPV). Numerous studies have been conducted systems firms financially constrained. Many considered constraints background context information, focus often designing system (Jiang Yang 2021; Goex Wagenhofer 2009, 2010; Bertomeu Cheynel 2015). These focused biases disclosures (e.g., conservative vs. liberal accounting, fair historical cost) typically assumed unrestricted information-structure does question cutoff thresholds reporting, general (i.e., accuracy signal) presence without unlimited choices structure. contributes literature previous shown, various settings, limiting supply information—for example, lower aggregation, allowing manipulation—can compensating for inefficiency arising lack commitment, contracting problems, difficulties motivating acquisition, etc.2 Although stream large, most do Among consider debt, Burkhardt Strausz (2009) although transparent reduces asymmetry increases liquidation assets, worsen asset substitution problem debts. Deng al. (2019) constrained firm. They find driving forces—a concavity effect. Neither forces exists study. all NPVs regardless types, moot inefficiency. examined (Bharath 2008; Li Donelson 2017). sparse research explicit them. provides theoretical predictions future research. An requires capital I, it, borrows I − creditor, where own personal wealth. use size loan relative total investment, β ≡ / $$ \beta \equiv \left(I-A\right)/I , represent successful, generates flow, z = X; otherwise, fails zero 0.3 Without any entrepreneur, success pG pB one. unobservable outsiders; belief θ. assume XpB < XpG > I; yields NPV good. Conditional S, option abandon pursues DS, S ∈ {g, b}, obtain X DS respectively. Otherwise, zero. whether unobservable, costly (e 1), Δp, incurs cost, c. 0), change cost. p B + Δ c X\left({p}_B+\Delta p\right)>I+c . e 1, improved profitable yield NPV. Finally, {0, X}, realized, receive respective payoffs. Notice breaks ante, denoted Π, represents model. timeline illustrated Figure 1. First, first-best benchmark setting. scenario, undertakes turn into positive-NPV scenario achieved wealth sufficient fund need financing), invest signal.5 Obviously, plays benchmark. Lemma 1.In benchmark, role. To must understand interactions two. first equilibrium assuming q, given. By backward induction, start date 3, given b}. if, sufficiently low; c/Δp. Intuitively, work hard keep realized flow after repaying similar Holmstrom Tirole (1997) (2006): too much On 2, even, conjecture about decision. Since its type, anticipates will three cases depending describe them following proposition Table Proposition 1.There thresholds, T θ G T\left({\theta}_S\right)\equiv \left[{\theta}_S\left({p}_G-{p}_B\right)+{p}_B+\Delta p\right]\left(X-c/\Delta p\right)/I, Always-effort: T(θb), 1 Effort-only-with-g: T(θb) T(θg), g, undertaken 1; b, 0 q ¯ q<\overline{q} θu. Never-effort: q>\overline{q} θu; .All proofs Appendix. Always-effort Effort-only-with-g T(θg) θu: otherwise: foregone Never-effort Notes: PG 0.5, PB 0.25, Δp 0.2, 0.3. threshold, T(θS), below supplies 1) S. posterior beliefs good, θS, threshold. θb θg, T(θg).6 return (always-effort case). achieve first-best. Dg, enough him. Db becomes jeopardizes incentive (effort-only-with-g Specifically, strong (θ θu) ( ), overturn project; thus, abandons highly likely bad. never (never-effort case), takes away effort's loses incentive. strong, happens (i) accurate confirms (ii) From readily seen weakly β, increases, moves always-effort case (in first-best) effort-only-with-g eventually reaches never-effort case. Corollary 1.Π β. now each described 2.(i) independent q. 1/2, (iii) likelihood obtaining Higher increase decrease 1/2), weak chance results Second, cases. cases, arises abandoning undertaking supplying As improving neither nor depends probability, θ, therefore leads decision; likely. trade-off countervailing effects. 2 \theta <\left[X\left({p}_B+\Delta p\right)-I-c\right]/\left[X\left({p}_G+{p}_B+2\Delta p\right)-2I-2c\right] dominating, \left[X\left({p}_B+\Delta p\right)-2I-2c\right]<\theta <1/2 possible scenarios. One θu depend relatively implies likely, q>\overline{q}, θu, dominates, affect precise, invests weak. better ones, twofold scarce. signal), accurately identified, generate result importance considering endogenous Moreover, structure (leverage). institutions, airline companies, manufacturing companies extremely capital-intensive ratios (DeAngelo Stulz 2015; Maverick imply higher-quality environment operating damage interaction preceding analysis, showed levels separate ranges. separating ranges summarized proposition: 3.With parameter shrink, range expands q). Recall threshold observes indicates stringent condition ask could converted Meanwhile, above less-stringent reason stronger, asks motivates 3 disappear cost combined Importantly, differently (as indicated 2), 3). numerical example It highest, drops Panel extreme merge vanishes. situation, achieves than merged widest. once beyond dramatically “falls off cliff” down say 0.4, increasingly wider, become narrower (T(θb) grow apart), shown panels C 2. form plateau area plummet 1). D illustrates vanishes, end up plateau. reaching lowest, loss. insensitive summarize remark: Remark 1.When (higher), (less) mostly levels. level, Therefore, standards imposed economy, achieving smaller, remains companies' economy environment, vital, crucial avoid drop worst ready solve