There is a growing literature (e.g. Bloom, 2009) studying negative macroeconomic impacts of heightened uncertainty. One channel through which uncertainty can discourage economic activity, and investment in particular, is via real option effect, where if firms face irreversibility or fixed cost in investment, higher uncertainty will encourage them to postpone investment, or exercise their "option to wait".

My question is: which of the two forms of nonconvex adjustment costs (fixed cost vs. irreversibility) is more important for this mechanism, and how do they differ in their effects? Is there any good reference explaining the issue, preferably in discrete-time setting? (I'm aware there's large literature on real options in continuous time, but so far I'm more familiar with discrete-time macroeconomic models).

  • 2
    $\begingroup$ Do you mean a fixed cost of production or a fixed cost of adjustment? FWIW, my reading of adjustment costs is that convex adjustment costs get you lots of small adjustments while non-convex adjustment costs get you infrequent larger ones. But a fixed (non-proportional) cost of adjustment means that large firms will adjust much more than small firms while irreversibility would seem to reward many small projects instead. $\endgroup$
    – BKay
    Dec 17 '14 at 13:09
  • $\begingroup$ Yes, I meant that the firm pays fixed cost if it wants to adjust its capital stock, and irreversibility meaning that resale price of capital (when investment is negative) is less than 1, or zero if disinvestment is impossible altogether. $\endgroup$
    – ivansml
    Dec 17 '14 at 16:26
  • $\begingroup$ Optimal stopping problems are always prettier, and often more informative, in continuous time, is there a reason you insist on the discrete time setting? $\endgroup$
    – Michael
    Dec 21 '14 at 1:21
  • $\begingroup$ @Michael , I don't really insist, it's just that my human capital is higher in discrete time setting :) But I guess I should just make the effort (I was planning to read Dixit & Pindyck book sometime...) $\endgroup$
    – ivansml
    Dec 21 '14 at 3:58
  • $\begingroup$ I recommend Dixit's seminal JPE paper also. Fixed cost is the center of the story there. $\endgroup$
    – Michael
    Dec 21 '14 at 21:28

There exists a post on capital adjustment costs, which gives a good overview of the different types of capital adjustment costs.


However, ultimately the post summarizes Cooper, R. W., & Haltiwanger, J. C. (2006). On the nature of capital adjustment costs. The Review of Economic Studies, 73(3), 611-633.

Regarding your question, I guess as long as there are any sunk costs associated to investment you destroy future option, regardless the nature of these costs. Consequently the nature of capital adjustment cost should not really change the way uncertainty affect investment through real option effects.


I am not familiar with real options in macroeconomics but in investment theory. In general, real option analysis is a concept to evaluate investments under uncertainty and irreversibility. Hence, if your fixed cost are reversible, you can simply use net-present-value or to consider uncertainty, you can also work with an investment tree. However, in the case where your investment is irreversible, that is, if the project stops, the investment costs are sunk, real options are the right approach. Hence, the question is not about fixed or irreversible cost, but rather whether your fixed costs are reversible or irreversible.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.