Over the last years we have witnessed a growing interest in project portfolio management, as well from a strategic perspective (do the right projects) as from an operational perspective (do the projects right).
Many organizations are currently deploying or improving the use of software tools for that purpose, experimenting with decision methods like weighted scoring or pairwise comparison and the Analytical Hierarchy Process (AHP) to prioritize projects in their portfolios.
The search for business drivers or selection criteria with an aim to combine them into a strategic metric that allows ranking of opportunities still presents many obstacles, and forces senior management to get better aligned in their decision-making process.
Conventional cost-benefit analysis using net present value (NPV), internal rate of return (IRR), or other financial indicators has been present for a long time, and is still very present in many investment or portfolio governance processes. More advanced and complex theories like real options remained in the background. Improved knowledge and insight in managing uncertainty or risk must allow decision-makers to adopt over time more advanced methods in the evaluation and selection of the projects in their portfolio, whilst consulting organizations must seek ways to improve accessibility to these methods for their customers.
What forms the basis of real options analysis? Contrary to the conventional cost-benefit analysis, a business strategy is much more like a series of options than a series of static cash flows, and therefore we should compare a project opportunity (business case) with one or more (call) options.
Thought leaders who actively cultivate their portfolio of opportunities by pushing the options towards higher profitability and certainty, could hereby turn originally uncertain business cases into success.
This document will elaborate on the basic principles of option valuation techniques applied on project portfolios in comparison to the conventional financial evaluation techniques, and will discuss the managerial and leadership prerequisites for longer-term acceptance and adoption in organizations.
PMI's Pulse of the Profession 2012 identified some areas of focus (PMI, 2012). One of them was “Tight economic conditions will continue to force the issue of good portfolio management.” The message was reinforced recently in the 2013 edition, where PMI's CEO Mark A. Langley made a clear statement on high performers: they focus on maturing their portfolio management processes, on improving organizational agility to allow flexibility and quick response and they track benefit realization (PMI, 2013).
I have observed a substantially growing interest in Portfolio Management over the last five years, on the one hand in the strategic part, which is linked to the selection of the right projects as well in the operational part, where efficient resource allocation to project portfolios is considered as a complex and challenging process.
One of the most important questions in Strategic Portfolio Management is: “Which projects should receive priority in our organization?” Several answers can be given:
- The projects with the highest strategic value
- The projects with the highest financial return
- The projects with the lowest risk
Strategic value is most likely different in each organization. The strategic value of a project can be defined as the degree to which the project is important or useful in relation to the organization's strategy (mission, goals,…). So the next step is to quantify the strategic value.
The most obvious quantitative approach would be to use weighted scoring models to determine the strategic value of each project in the project portfolio, by giving strategic criteria (business drivers) a weight (expressing importance) and by scoring the contribution of each project to that strategic criterion. Many project portfolio management (PPM) software tools allow for that today.
Take a simple example with three projects A, B, and C in our portfolio (see Exhibit 1). We have identified three strategic criteria, and each criterion has received a weight (we could have used pairwise comparison and AHP to calculate the weight). The non-normalized strategic values are now calculated and give project C the highest score.
We have to find the right scale (e.g., 1-3-5-7-9) for the importance of the different criteria (weight) and for the contribution of each project to a criterion. Thomas Saaty's Analytical Hierarchy Process contests such an approach and claims the need for pairwise comparison of business drivers and pairwise comparison of projects for each business driver. No doubt that complexity of his model is much higher, and therefore acceptance and adoption more difficult.
But what about the classic NPV? How do we include this financial metric in the selection process? Portfolio managers might suggest including the NPV as selection criterion. They forget however that the project with the highest strategic value is not de facto the one to give the highest priority, as we still need to build our efficient frontier: this means selecting the project that has the highest strategic value per unit of cost. In our example the highest absolute strategic value is given to project C, however project B has the best ratio per unit of cost.
Finally, one could ask the question: “Suppose project A gives you the highest financial benefit per unit of cost over the next three years. Would you go for project A instead of project C?” Probably project A has the highest NPV. The benefit metric could also have been included in the selection process.
Let us assume that project C is a strategic project as it enables many new other projects in a later stage, but the benefits are relatively weak, as the uncertainty is high and related to market evolution and customer acceptance. Which project now gets highest priority?
This simple example proves the complexity of determining an objective, acceptable and quantitative project selection model in portfolio management.
Mapping Projects to Options—The Basis
Let us bring our investment opportunity in the world of financial options in an attempt to transform it to a “real option”…For those who are not familiar with options in the financial world, here a simple definition: a call option gives the investor the right (but not the obligation) to buy an underlying asset within a specified time period.
A simple option can be defined by the description of the asset, an exercise price at the expiration date, and a current value; for example, we could buy a call option of BNP Paribas Fortis today at 3,60 € with an expiration date in December 2013 and an exercise price of 48 € (X), whilst the stock price (S) today is 44 €. What happens in December 2013? The stock price might be lower than 48 €. We will not invest. If the stock price is 50 € and thus higher than 48 € we will buy and have a gain of 50 € – 48 € = 2 € (S – X).
The interest on the delayed expenditure of X can be calculated or X can be changed to the PV(X), using a risk-free rate of return rf. Then we can define a new metric called NPVq = S÷PV(X).
Now we look at uncertainty in our calculations. This can be done by including probabilities in the model. From statistics we know that variance (σ2) is an expression of the spread around the mean. The variance can therefore been seen as an attractive measure of uncertainty. However, it does not include a time dimension. Therefore we speak in terms of variance per period and multiply or variance by the number of periods between now and the expiration date: σ2t. Simplifying it again, we take the square root and define a metric on volatility as σ√t.
The last step in our comparison with the financial world is to use the Black-Scholes model. Both economists invented in the ’70s an equation to simulate the price of an option. Many websites on the Internet allow computation of option prices by simply entering some basic input (S, X, rf, σ and t).
Real options analysis does not ignore conventional DCF calculations; it just goes a step further.
It takes some practice to identify options “hidden” in our project opportunities. They could be identified in the DCF calculations by examining patterns in the cash flows (e.g., an investment after several years). Let's take an example of a project with a second investment in year three (see Exhibit 2).
The conventional approach would lead to a no-go decision, as the NPV (-19,75 K€) is negative (using a risk adjusted discount factor of 12%). Now we apply the real options approach and separate the short-term decision on phase 1 from the longer-term decision in year three.
We apply a risk-free rate of return of 3% on the investment X of 70 K€ and calculate PV(X): 64,05 K€. We also calculate the net present value of S (revenues & expenses): 60,90 K€. Assuming a volatility of 33% we now simulate the value of the call option using the Black Scholes model, resulting in a value of 12,56 K€.
The value of the total project now becomes interesting as the value is positive: -8,78 K€ + 12,56 K€ = 3,78 K€. Or time and uncertainty have given our investment opportunity a greater value, as we can defer the decision to a later stage (until the expiration date).
A Portfolio of Real Options
Strategy includes uncertainty and must allow for adjustment and proactive decision-making over time. Can we see strategies as a series of options instead of a set of static cash flows? We still observe today that decision making around projects is centered on a key milestone in the life cycle (e.g., at the end of the definition phase) using a high level or detailed business case.
At portfolio level we can create “option space” by using the two key metrics (Luehrman, 1998). Our first metric is NPVq or S÷PV(X), an indicator for the profitability of our projects, and one can been seen as a key value creating a left and right side in our space (see Exhibit 3). Our second metric is an indicator on the stability of our assumptions (volatility), and therefore a key indicator of risk. This metric σ√t is equal to 0, when all uncertainty is gone (σ = 0), or when time has run out.
What can be considered as the ideal project? A project with a high NPVq and a very low volatility. How can we achieve low volatility? If assumptions are pretty stable, we should apply a low σ, and if the expiration date is pretty close, we should decide to invest. The first area in our option space is now identified, a green zone with projects that can get a go immediately (invest now)…if we have or can find the resources! At the same time an area is created on the right side where the projects will get a no-go (never invest).
The challenge for the project manager is to propose to the portfolio governance structure of the organization a threshold level for the volatility.
Let us move on and have a look at project A1 and A2 in Exhibit 3: whilst both project have the same NPVq (value), the project A2 has a higher volatility (thus risk).
The diagonal is a simplified border (in reality it is a curve) where the conventional NPV of the project is 0. For both A1 and A2 the value is higher than 0, for project A3 the conventional NPV is < 0. Project A3 will probably end up in the left (red) part when time runs out.
Whilst the conventional approach with NPV gives us only one metric (NPV), the options approach gives us multiple metrics and a broader set of actions put in a time perspective.
Many other dimensions are still not discussed in this paper, as it was the intent to reveal the basics of real options analysis in the project and portfolio management world. Time for some conclusions and recommendations…
Conclusions and Recommendations
If your organization still uses the “intuitive” way of project selection, there is still a long way to go…even when your organization is already applying weighted scoring and/or pairwise comparison. By entering the real options analysis world, I have learned over the last years that pure conventional thinking with NPV does not reflect the agility and dynamic decision-making that is required to align your projects portfolio to today's volatile market conditions.
The complexity of real options and therefore the difficulties that one can expect in achieving acceptance must not stop us to use the concepts in our regular strategic reflection process. I see great value in the inclusion of time and uncertainty in the portfolio selection process, reflected by the volatility metric, whilst our NPV calculations can still be used. The challenge is to identify the options within our business opportunities, and to get key stakeholders interested in the approach….