Although Net Present Value (NPV) modelling is a robust approach to testing the financial basis of a project case, it does have a number of limitations. As a consequence, NPV is not suitable for all projects. Its results should always be considered in the context of the decisions being made. The following are examples of where limitations of the approach may be important.

Projects with objectives that cannot be measured in cash terms

Some projects have no cash benefits. Projects with social, political, strategic or military purposes often fall into this category. Examples would include a government funded project to build a new aircraft carrier or a charity funding the earlier stages of medical research. NPV modelling is not suitable for these types of project.

Some projects have a mixture of cash and non-cash benefits. It might be appropriate to approve such projects with a negative NPV provided they have additional non-financial benefits judged to be sufficiently good value for money to more than make up the difference. 

Long projects with end loaded costs

Discount rate assumptions might be of concern on long-term projects with large end-loaded costs e.g. a mining project that is obliged to return the site to an environmentally acceptable standard. If a high discount factor is used, there could be a valid concern that such costs were being modelled with insufficient prudence.

Project Portfolio considerations

A project with positive NPV might be correctly declined when taking into account other projects in the organisation’s portfolio. For example the organisation may lack sufficient capacity to run the project alongside existing or future projects. Or it may be deemed to be insufficiently well aligned with the organisation’s strategic objectives.

The Net Present Value Rule - treat with caution!

A number of sources identify a “Net Present Value Rule”. Typically, the rule is described as being “Projects with negative NPV should be rejected but projects with positive NPV may be approved”. 

Although this rule is easy to understand, the above examples of NPV modelling limitations illustrate why there many exceptions in which applying the rule would be wrong.

A further issue with the Net Present Value rule is that it assumes that projects have a either positive or negative NPV and is thus based on the use of deterministic NPV models. As identified by the risk modelling pages on this web site, a project that is likely to have a positive NPV, might nevertheless, expose the organisation to a risk of loss. This can be an important factor in the approval of major projects, where the scale of loss on a single project may become significant to the organisation as a whole. 

Limitations on the accuracy of models

No model is ever a perfect representation of reality. NPV models are no exception. The following points are worth remembering:

• Since NPV models are reliant on estimates, perfect accuracy is an unrealistic expectation.
• Any financial model is developed in the context of a framework of simplifying assumptions. As more detail is added the implications the associated assumptions and constraints may go unrecognised.
• The discount rate underpinning any NPV model is itself often derived from a separate model and is thus a model within a model.

When using the results of NPV models, these limitations should be borne in mind. Where NPV forecasts with three significant figures have been cited elsewhere on this site, it is only for the purpose of making the examples easier to follow!

The good news: NPV Models are still Useful!

 "All models are wrong, but some are useful" (George Box)

No forecasting method is perfect. NPV modelling may still be a better approach for supporting decision makers than any of the alternatives. When this is the case, understanding its limitations is an important part of the skill set required to build models that are fit for purpose.