What’s wrong with NPV in portfolio management?

Flightmap BubbleNet Present Value: indicator of value creation

If you would ask me “What is wrong with Net Present Value?”,  the easy answer would be: “Not much”. It is one of the best indicators of potential value creation. It nicely summarizes the consequences of an investment decision in terms of costs and benefits over time. If you are not sure what Net Present Value is, have a look at the Wikipedia page here.

Of course, NPV is a financial metric, and probably one of the best. Better than just revenue or margin, because it looks at investment costs and timing as well. Since NPV focuses on absolute value creation, it is arguably better than its relative sister Internal Rate of Return (IRR). Even with the technical issues of IIR fixed (see Modified IRR or MIRR), IRR-based project selection easily focuses on short-term quick-win projects only, since these have the largest relative value creation.

The importance of financial value assessment

A more fundamental issue with NPV is that it counts financial value creation based on predictable cash flows only. This means NPV cannot be determined where cash flow impact cannot be forecasted. This holds for investing in strategic, mostly intangible assets such as new market exploration, brand awareness, improved competence, where the financial value is so far downstream or so unpredictable that cash flow forecasts make no sense. This in turn indicates a more strategic value indicator is preferable.

So why not just have a good strategic (non-financial) value indicator? In over 15 years in practice I have seen too many portfolios that looked attractive from a strategic qualitative point of view, but that had severe financial value creation flaws. In a previous post, I already elaborated the financial versus strategic value bubble plot as a solution to make sure both dimensions are addressed (as shown in the bubble plot at the top of this post as well).

Don’t blame NPV if you can’t handle KPI-s

Another main objection that is often mentioned is more an objection to the NPV decision rule than to the NPV itself. The NPV Decision Rule (in its simplest format) states:

NPV Decision Rule: if NPV<=0 don’t do it and if NPV>0 do it

This decision rule is indeed an example of misuse of Key Performance Indicators or metrics. It suggests that decision-making is a “no-brainer”. A few years ago, I found that this misconception about KPI-s is easily unveiled with a simple but powerful model, the “MUD cycle”.

MUD cycle

This MUD cycle tries to capture the essence of good use of metrics or KPI-s: the Measurements lead to improved Understanding that in turns drives better Decisions.

If you “forget” any of the three ingredients, a flawed decision-making process results:

  1. without Understanding (such as the above NPV decision rule), the Measurement’s limitations and context are not incorporated in the Decisions;
  2. without Decisions, the process gets stuck in a never-ending cycle of getting more metrics and running more analyses;
  3. and without Measuring, the Understanding can only be based on intuition, and learning and sharing to improve Decisions is hampered.

Even the dashboard in your car is based on this cycle: it shows you metrics such as fuel level, speed, and warning lights, but it requires Understanding of the context to translate the Measurement values into proper action: is it safe and smart to stop for gas, or to stop to let the engine cool off, or to change route?

In short, I am in favour of including NPV as a metric, but not in favour of the NPV Decision Rule (or any other decision rule that dumbs down decision-making).

Specific challenges of NPV in project portfolio management

With the more generic pro’s and con’s of NPV out-of-the-way, what about the limitations of NPV for portfolio management? This is where I see 3 major limitations:

  1. Dependency of projects across the portfolio
    Due to cannibalization, synergies, and other dependencies between projects, the NPV of a certain project can depend on the other projects in the portfolio. Work by Keisler (in the book “Portfolio Decision Analysis”) and earlier by Bucher and Jung of ETH (“Computer-aided R&D-portfolio valuation”) show that ignoring these effects leads to significant mistakes in decision-making. This in turn suggests that we (also) look at incremental NPV: how much does the portfolio-level NPV change when we add another project.
  2. How to accommodate for risk?
    In simplified NPV calculations, it is recommended to include risk by increasing the discount rate applied to the cash flows with a risk premium. This very crude mechanism does not properly mimic the typical reduction of risk as projects proceed, and it also does not help to manage risk. This is why we recommend to look at risk-adjusted cash flows (and their NPV-s).
    Looking at risk and value as separate indicators is much more meaningful, since that doesn’t hide risk in risk-adjusted NPV, Expected Commercial Value (ECV) or Real Option Value (ROV). A 1% chance of getting 100 Mln is nowhere the same as a 100% chance of getting 1M. Again, it is a portfolio level consideration how much risk you want to take. In line with the MUD cycle, separating the risk from the value increases understanding.
  3. How to make sure NPV-s are comparable?
    In a portfolio of projects, we have to make sure that the NPV calculation (or any other KPI for comparing projects for that matter) is based on the same ground rules. So all cash flow forecasts are based on the same assumptions on the macro-environment, the discount rates are the same, and the discounting is applied to the same date (not to the random moment where the project’s NPV happens to have been calculated).
    On top of that, determining whether the project is good enough cannot be decided in isolation, the total portfolio of selected projects needs to meet the value creation objectives.

The answers to these three questions all point to the same direction: look at the portfolio level, not just at the project. In practice, this means that the portfolio-level NPV is a key measurement, and its decomposition across contributing projects is helpful to understand where the value comes from.

So, do you have Portfolio NPV at the center of your portfolio management process?

2 thoughts on “What’s wrong with NPV in portfolio management?

  1. For a detailed understanding of the IRR and NPV please review the following articles:

    ” New Method to Estimate NPV from the Capital Amortization Schedule and an
    Insight into Why NPV is Not the Appropriate Criterion for Capital Investment
    Decision” – This paper is available in the following link:


    Title: IRR Performs Better than NPV: A Critical Analysis of Cases of
    Multiple IRR and Mutually Exclusive and Independent Investment Projects


    These papers question some of the points discussed as acceptance or rejection criteria.Please let me have any comments.

Please share your thoughts on this blog

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s