Sooner or Later? – Paradoxical Investment Effects of Capital Gains Taxation under Simultaneous Investment and Abandonment Flexibility

Abstract

This paper analyzes the impact of capital gains taxation on investment timing decisions for risky investment projects with entry and exit flexibility under differential tax rates for ordinary income and capital gains. We investigate whether capital gains taxation influences immediate and delayed investments asymmetrically, given the optimal abandonment decision. If capital gains taxation induces a lock-in effect, this effect is anticipated in the investment timing decision. In contrast to prior research, our numerical simulations show that this lock-in effect of capital gains taxation can induce normal as well as paradoxical effects on investment timing under simultaneous entry and exit flexibility. A paradoxical timing effect, i.e., investment accelerated by capital gains taxation, especially emerges for high liquidation proceeds or, more conservative tax accounting, low interest rates, and low volatilities. In these cases, capital gains taxation reduces the value of the option to invest and hereby increases the propensity to invest immediately. As a second paradoxical tax effect, capital gains taxation may favor delayed real investment over financial investment. Facing these results, tax legislators should not use capital gains taxation as a short-term tax policy instrument to influence investors' timing decisions.

Acknowledgments

This paper has benefited from helpful comments by the editors Salvador Carmona and Steven Young and two anonymous referees on earlier versions of this paper. The authors also thank workshop participants at the European Accounting Association Annual Meeting and the Annual Meeting of the German Academic Association for Business Research. Rainer Niemann gratefully acknowledges support by the Austrian Science Fund (FWF, P 22324-G11). Caren Sureth gratefully acknowledges support by the German Research Foundation (DFG SU 501/1-2 and DFG SU 501/4-1).

Notes

Neutral tax systems as a reference concept for analyzing tax effects have been proved under certainty by Brown (1948), Johansson (1969) and Samuelson (1964). Under uncertainty, enriching the real option literature by integrating taxation (e.g., Agliardi, 2001; Alvarez and Koskela, 2008; Gries et al., 2012; Niemann, 1999; Niemann and Sureth, 2004, 2005; Panteghini, 2001, 2004, 2005; Pennings, 2000; and Sureth, 2002) leads to investment rules that consider managerial flexibility, irreversibility and tax effects.

For a recent detailed overview of the empirical literature on capital gains tax effects on asset prices and investment decisions, see Niemann and Sureth (2009).

See Schneider and Sureth (2010); Gries et al. (2012).

See, for example, Dixit and Pindyck (1994, pp. 147 ff.); Trigeorgis (1996, pp. 72 ff.). A very user-friendly guide to the critical determinants of successful application of this approach with several examples is Amran and Kulatilaka (1999).

It can be easily shown that the decision rule for combined investment cost and cash flow uncertainty is similar to the decision rule under cash flow uncertainty only. Combined uncertainty simply requires the change of the numéraire from monetary to project units. See Dixit and Pindyck (1994, pp. 207 ff.).

For a detailed numerical analysis comprising all technicalities in a stylized pre-tax case, see Niemann and Sureth (2009).

We do not distinguish between depreciation recapture and increases in value when computing the capital gains tax base.

For simplicity, we do not take other depreciation schedules such as declining balance depreciation into account.

There is no perfect analogy of variations of the multiple and the degree of accounting conservatism. Different degrees of conservatism do not affect cash flows, whereas different multiples induce different cash flows.

See Dai et al. (2008) for lock-in versus capitalization effects of capital gains taxation. In our model, the parameter captures both the lock-in and the capitalization effect of capital gains taxation. It is an empirical question which effect dominates. In our model, a capitalization effect decreases .

The same would be true for a put option on a stock with a variable strike price that is always equal to the current stock price.

See equations (15) and (16) in Appendix C.

For a detailed description of the remaining objective values at the decision nodes at time and , the value of the option to abandon (put option), and the derivation of the value of flexibility by backward induction, see Appendix D.

For a related simulation on non-depreciable financial investments see Alpert (2010).

The effect of tax-induced complexity can be observed exemplarily in Hundsdoerfer et al. (2008).

See Section 30 of the Austrian Income Tax Code in connection with No. 6654a Income Tax Directives.

See Section 34 (3) German Income Tax Code.

Note that real-world tax systems may be characterized by different tax rates for sale and liquidation proceeds. This implies that taxpayers minimize their tax burden by arranging the facts that determine the tax base. These tax planning strategies are reflected by the optimization calculus in our model. Although sale and liquidation are very much related, tax systems often apply different tax rates for these different ways to quit an investment. Integrating different tax rates in a decision model can complicate the calculus considerably. See, for example, Hundsdoerfer et al. (2008) who show how investment and financing decisions can be optimized simultaneously. Based on simple premises they evaluate an indivisible investment project that is carried out in a corporation and find that the decision problem turns out to be rather complex if different tax rates have to be considered.

Superscripts * indicate optimal decisions. Subscripts _t in the expectations operator E _t [·] indicate the time of taking the expectation.

Additional information

Notes on contributors

Rainer Niemann

Paper accepted by Salvador Carmona.

Caren Sureth

Paper accepted by Salvador Carmona.