ñòð. 158 |

20 NPV $5m

%

Enter Market

Present cost

$20m Low Demand

PV $10m

NPV $(10m)

Fig. 12.5 Swimming pool cover project

One hardly even needed a decision tree thanks to the upside and downsideâ€™s

offsetting each other. With no skew, the mid-point outcome was the expected

value. It is worth building one, however, because by doing so we can test the

impact of different assumptions on the overall economic viability of our pro-

ject. For example, we could determine that if the respective probabilities for

low, medium and high demand were 40%, 40%, 20% then the NPV would

fall to $2m. A zero NPV would be the result if the probabilities were 46â…”%,

40%, 13â…“%.

Now, however, let us introduce the idea of a market survey. The cost, in

present value terms is $1.5m. One could simply subtract this from the ini-

tial view of NPV of $5m and claim an NPV of $3.5m. The project would

still appear worthwhile but our valuation would, however, ignore the entire

point of carrying out the survey. This is because we appear to be taking no

notice of what it concludes. We need to make some assumptions about how

the survey might influence our decision-taking. What we must not do at this

stage, however, is assume that the survey can change the likelihood of the

511 Second view: Valuing flexibility

three possible outcomes. This information can only become available after

the survey is complete and we must take care to respect this fact as we make

our assumptions.

The most optimistic assumption which we could make about the survey

is that it would always be right and would warn us if the low-demand scen-

ario was going to come about. If we were sure that if the survey indicated

low demand, then demand would indeed be low, and that if medium or high

demand was indicated then low would not happen, our decision tree would

look like this:

Outcome

High Demand

PV $40m

NPV $18.5m

%

20

Overall NPV

Medium Demand

60%

$5.5m

PV $25m

NPV $3.5m

Enter Market

Present cost

$20m

Survey

Present Cost

20

$1.5m %

Low Demand

Do not Invest

NPV $(1.5m)

Fig. 12.6 Swimming pool cover project with â€˜perfectâ€™ survey

The â€˜perfectâ€™ survey has added $0.5m to value compared with not carrying it

out because the NPV has risen to $5.5m. If we do the survey but then ignore

it we would lower value by $1.5m.

Now let us investigate what the result would be if we believed that the sur-

vey would be right 80% of the time. We will still assume that we will act on

the surveyâ€™s recommendation but now there will be a probability that, with

hindsight, we regret this. We will assume that if the survey indicated low

demand then 80% of the time this would be correct but 20% of the time the

demand would be medium. If high demand is forecast then we will assume

that 80% of the time this is right and 20% of the time demand is medium.

Again, we must be scrupulous about ensuring that we do not change our

overall assumption about the balance between low, medium and high demand.

512 Three views of deeper and broader skills

All that we can do is change the success rate for the surveyâ€™s predictions. This

means, for example, that the survey must forecast medium demand with a

60% probability and that, in our view of reality, the overall percentage of

medium demand must also remain at 60%. This means that once we have

decided upon the accuracy in relation to when the survey indicates high or

low demand we must back-calculate what the probabilities must be for when

the survey indicates medium demand.

The decision tree will look like this:

High Demand

NPV $18.5m

16% probability

Medium Demand

NPV $3.5m

4% probability

h

ig

-H

High Demand

y

%

ve

20

NPV $18.5m

r

Su

4% probability

Overall NPV

Medium Demand

Survey Medium

$4.9m

NPV $3.5m

60% 52% probability

Low Demand

Survey

Su 20 NPV $(11.5m)

%

rve

4% probability

y-

Lo

w

Medium Demand

NPV $(1.5m)

4% probability

ñòð. 158 |