Electronic copy available at: http://ssrn.com/abstract=2103562Dynamic risk management: investment, capitalstructure, and hedging in the presence of
nancialfrictionsDiego Amaya, Geneviève Gauthiery, Thomas-Olivier LéautierzxFebruary 10, 2013AbstractThis paper develops a dynamic risk management model to determine a
rms opti-mal risk management strategy. This strategy has two elements:
rst, for low leveragevalues, the
rm fully hedges its operating cash ow exposure, due to the convexity ofits cost of capital. When leverage exceeds a very high threshold, the
rm gambles forresurrection and stops hedging Second, the
rm manages its capital structure throughdividend distributions and investment. When leverage is low, the
rm replaces depre-ciated assets, fully invests in opportunities if they arise, and distribute dividends; allof these together to achieve its optimal capital structure. As leverage increases, the
rm stops paying dividends, while fully investing. After a certain leverage, the
rmalso reduces investment until it stops investing completely. The model predictions areconsistent with empirical observations.JEL classi
cation: G32, C61Keywords: Dynamic programming; risk management; capital structure; hedging.1 IntroductionModern risk management theory is grounded in the observation that information asymmetrybetween managers/insiders and investors/outsiders limits the ability of
rms to raise externalFinance Department, Université du Québec à Montréal (UQAM), Québec, Canada.yDepartment of Management Sciences, HEC Montréal, Québec, Canada.zToulouse School of Economics (IAE, Université de Toulouse 1 Capitole), France.xDiego Amaya would like to thank FQRNT and IFM2 for
nancial support. Geneviève Gauthier wouldlike to thank NSERC and IFM2 for
nancial support. An earlier version of this paper was circulated under thetitle, "Coordinating Capital Structure with Risk Management Policies." We thank seminar participants atthe Annual Conference on Risk Management and Corporate Governance, the Annual Australasian Financeand Banking Conference, and the Midwest Financial Association meetings for their comments on earlierversions of this paper. Any remaining inadequacies are ours alone. Correspondence to: Diego Amaya,E-mail: amaya.diego@uqam.ca.1Electronic copy available at: http://ssrn.com/abstract=2103562funds. In a striking reversal of Modigliani and Miller propositions (1958, 1963), Holmstromand Tirole (2000) and Tirole (2006) argue that not all value-creating
rms or projects are
nanced: pro
table but cash constrained
rms may not be able to re
nance themselvesafter a negative shock to their cash ow, hence may go bankrupt;
rms with insu¢ cientinternal funds may have to forego pro
table investment opportunities, an issue known as theunderinvestment problem.The interaction between costly external
nancing, underinvestment, and risk manage-ment was
rst modeled in a two-period environment by Froot et al. (1993) and Froot andStein (1998). The former considers a
rm facing random cash ows, random investment op-portunities, and convex cost of external
nancing. At the optimum, the
rm fully hedges ifcash ows and investment opportunity are uncorrelated, and reduces its hedging as the corre-lation between both sources of uncertainty increases. The latter introduces capital structureas a risk management device. A marginal increase in equity raises the
rms capacity to pur-sue risky investments. On the other hand, it generates deadweight costs, arising for examplefrom the tax deductibility of interest payments. The optimal equity level balances these twoe¤ects.A more recent literature has examined this issue in multi-period models. Rochet andVilleneuve (2011) develop an in
nite-horizon, continuous-time model, where a constant-size
rm faces exogenous cash ow shocks and stringent
nancial frictions: the
rm is liquidatedas soon as its cash reserve becomes negative. At each instant, the
rm selects its dividendpayment and decides its hedging ratio or insurance coverage for discrete risks. Rochetand Villeneuve (2011) restate the risk management problem as an inventory managementproblem, in which the cash reserve is the state variable, and dividend payment and risktransfer decisions are the control variables. They then show that the
rm pays dividends ifand only if the cash reserve exceeds a threshold, and it fully hedges if the cash reserve isbelow the threshold. In addition, they show that the
rm insures small risks but not largeones.Bolton et al. (2011) extend Rochet and Villeneuve (2011), most notably by includinginvestment and growth, and less stringent
nancial frictions, i.e., re
nancing is possible,albeit costly. They
rst characterize the optimal dividend distribution, investment, andre
nancing policies. As Rochet and Villeneuve (2011), Bolton et al. (2011)
nd that the
rm optimally distributes dividends if and only if its cash reserve (as a percentage of itssize) exceeds a given threshold. They also prove that the optimal investment policy is to setthe marginal cost of adjusting physical capital equal to the ratio of the marginal Tobinsq over the marginal cost of
nancing, a departure from the Modigliani and Miller rule,which is to equalize the marginal cost of physical capital to the marginal Tobins q. Finally,they determine the optimal hedging policy that balances the marginal bene
ts and costs ofhedging.While these two articles are signi
cant conceptual contributions, they do not capture es-sential real-world features. First, corporate taxes and leverage are absent from the analysis,even though they play an important role in corporate decision making, as illustrated for ex-ample by Graham and Rogers (2002) econometric analysis of
rmsdeterminants of hedging.Second, as observed in Graham and Harvey (2001)s survey,
nancial executives and man-agers at large, publicly traded
rms use the Net Present Value (NPV ) of the free cash ows,2Electronic copy available at: http://ssrn.com/abstract=2103562discounted at the Weighted Average Cost of Capital (WACC), to make capital budgetingdecisions, while managers in these articles maximize the NPV of dividends. Finally, at leasta fraction of investment opportunities appears to be stochastic, as was modeled by Froot etal. (1993). Firms growth is shaped by the availability of investment opportunities as wellas by real frictions in adding capital.We are not aware of any model derived from micro-foundations that incorporates theseessential features. Furthermore, micro-founded models rely on information asymmetry, whoseparameters are by nature di¢ cult to estimate, hence these modelspredictions are di¢ cultto test empirically (Graham and Harvey (2001) and Graham and Rogers (2002)).Therefore, to provide a closer representation of reality, Léautier et al. (2007) proposea reduced-form model that aims to represent what
rms actually do: managers of largepublicly traded corporations make capital budgeting and hedging decisions to maximize theNPV of the free cash ows (consistent with Graham and Harvey (2001)s
nding) in a multi-period environment (as in Rochet and Villeneuve (2011) and Bolton et al. (2011)), facinguncertainty about both future cash ows and future investment opportunities (as in Frootet al. (1993)). Financial frictions are incorporated through the expected return required byinvestors, which is assumed to be convex in the
rms leverage ratio, that replaces the cashreserve as the state variable (Graham and Harvey (2001), Pettit (2007), pp. 110-111 and141-159, and Cohen (2003)).This article builds upon Léautier et al. (2007), and incorporates two additional elements:dividends distribution and the possibility of bankruptcy. This representation of managerialdecision-making cannot be derived from micro-foundations. Yet, it provides valuable insightssince the analysispredictions can be compared against actual
rmsbehavior, as capturedby previous empirical studies.We
rst determine analytically the optimal risk management strategy, i.e., the mix ofhedging, dividend distribution, re
nancing, and investment policies. Second, we illustrate theoptimal strategy for a "representative" industrial
rm, using estimates of the main modelparameters. Finally, we show that the models predictions are consistent with empiricalobservations. Our main result is the optimal risk management strategy, which is surprisinglysimple.First, dividend distribution and investment jointly follow four regimes (Proposition 1).For low leverage, the
rmenjoys full
nancial exibility: it fully
nances its investment needs,and distributes dividends to reach its optimal leverage ratio. For intermediate leverage,the
rm faces
nancial tightness: it still fully
nances its investment needs, but no longerdistributes dividends, as leverage increases from one period to the next. For higher leverage,the
rm faces a
nancial constraint: it is no longer able to fully
nance its investment needs.The portion it
nances is determined to reach a target leverage, after which the marginalvalue of investing becomes negative. Finally, for high leverage, the
rm faces
nancialhardship: it is no longer able to
nance any of its investment needs, not even depreciation.Second, full hedging is optimal unless leverage gets higher than some threshold, in whichcase gambling for resurrection becomes optimal (Proposition 2). These results di¤er fromRochet and Villeneuve (2011) and Bolton et al. (2011), who
nd that, when the
rms cashreserve (or cash-to-capital ratio in Bolton et al. (2011)) is high enough, the
rm becomes riskneutral, and, since hedging is costly, stops hedging. In our model, the tax shield drives the3concavity of the value function, hence the optimality of full hedging. By choosing leverageas the state variable, we are able to capture the tax shield from debt, a real e¤
đang được dịch, vui lòng đợi..
