JAMES WESTON: Hi.

Welcome back to "Finance for Non-Finance Professionals."

I'd like to talk in this video about sensitivity analysis, which

is an important component of our capital budgeting tools

that we're talking about in week two, "How to Spend the Firm's Money."

We've talked about a number of capital budgeting tools like NPV,

and internal rate of return.

We've talked about a couple of accounting ratios and payback tool.

And what I'd like to do now is talk about using sensitivity analysis

in order to round out or get a better sense of how those different capital

budgeting metrics hang together.

Everything that we do in all these capital budgeting

tools, whether its net present value or payback, depends on a forecast.

We're going to pretend like we know what it's

going to cost to do the project, which we probably don't, and how much we're

going to sell off the project, which we probably don't know,

and how much we're going to sell five years out from now, which we probably

don't know either.

Well, that can actually start to get depressing,

because all forecasts are wrong.

When the weatherman says it's going to rain with 20% probability,

he's wrong, because it never rains with 20% probability.

It either rains or it doesn't rain.

So once you get the realization of that uncertainty, you go back

and say, well, the forecast was wrong.

So all forecasts are wrong conceptually.

And that's just a concept that you need to be comfortable with.

What's important in terms of making decision under uncertainty

is that we not say, well, because we can't forecast things perfectly,

we should just forget it, and just kind of go with your gut and wing it.

Throw a dart.

That's the wrong thing to do.

Even though we live in an uncertain world

where we never really know what's going to happen,

and all forecasts are essentially wrong from their conception,

that doesn't mean we shouldn't try to make the best forecast possible.

And that's what these capital budgeting tools are sort of doing.

And using arm's length objective and transparent metrics

for making capital budgeting decisions, we're

trying to make the best forecasts that we can.

The fact that we never really know what the answer is going to be

shouldn't prevent us from trying to make the best answer that we can.

So what we're going to talk about in this video

is we had to make a set of assumptions.

What do I think first year, second year, third year sales are going to be?

Where do I think when we finish the project, what's

the closing cost going to come in at?

How much is the project going to cost to do?

We have some sense of it.

We have some forecasts for it.

But we're probably going to be a little bit off, plus or minus.

And what we want to think about now is, well,

how sensitive is our decision to a lot of those assumptions?

And we can do that really easily within the framework of the capital budgeting

tools that we've already talked about.

And we'll think about what are the main value drivers,

and if I tweak those, if I wiggle them around a little bit,

how much does that change the answer?

And that sensitivity analysis, or scenario analysis,

will give me an idea of how comfortable I am with the forecasts

that I've made for the project.

OK, so let's walk through just a really simple example

of what am I going to do here.

I've got in time 0, I'm going to spend $5,000.

What do I get in return for spending that?

I get a whole series of cash flows, $3,200, $2,500, $1,200, $1,200,

$1,200 over the next five years.

If I compute an NPV at 15%, that gives me a net present value of $1,745.

OK, so that's what I expect to happen.

Is that what's going to happen?

Of course not, because when it comes to the realization in year three,

maybe I make $1,100.

Maybe I make $1,700.

I don't really know.

It's just that $1,200 right now is my best guess for what I think

is going to happen.

And so if you feel uncomfortable about that, you probably should.

Because I want to know like, well, that's just a set of guesses.

How close would I be to getting profitable or not profitable

if we wiggled those numbers around a little bit?

And so that's what scenario analysis does for us.

Let's say, hey, let's talk to the sales team.

Let's talk to the research and development people.

Let's talk to the contractors and think about how much might this project cost.

It might be that the project comes in a little over cost.

In a pessimistic scenario, it's going to cost $6,000

to get the project off the ground, whereas it

costs $5,000 in my other two, expected scenario and optimistic scenario.

Maybe sales don't come in as strong as they should.

They're lower in all of my years.

So what I've done is I've sort of dialed down

my forecast in my pessimistic scenario.

And I've made a realistic forecast by talking

to the people involved in the project.

In a good case scenario, if things went better than expected,

what would the cash flows look like?

And I've given sort of an optimistic scenario.

And now I compare these three NPVs.

At 15%, I've got minus $587, $1,745 and $3,665.

OK, so what have I done?

I've pulled, and I've twisted and I've played

with those assumptions a little bit, and said, OK, now I'm

getting a better sense.

Instead of just looking, that $1,745 and making the decision only

on that number, I'm kind of wiggling those numbers around a little bit,

and saying, OK, what would be a good case scenario?

What would be a bad case scenario?

And now, hopefully what's the worst case scenario is I lose $587.

Now I can think about, well, is it worth it to gamble the $5,000 in order

to maybe make even more money?

I mean, that number's not as negative as this number is positive.

And I can start thinking about making those kinds of trade-offs

now that I've kind of fleshed out the project a little bit more.

Sort of articulated the uncertainty by mapping out

a couple of different things that could happen and weighing those off

against each other.

So sensitivity analysis, we could expand lots of different scenarios.

We could do five different scenarios.

We could calculate sensitivities.

How sensitive is the NPV to the discount rate, to the first year cash flows,

to the initial cost?

Spreadsheets make a lot of this easy.

And we'll walk through some of those spreadsheets

together, because all I have to do is change

the input on one of the assumptions, and the rest of the assumptions

kind of flow through the NPV.

And so it's easy to calculate those sensitivities and scenarios

once we build up a spreadsheet model for doing that.

So let's talk about it.

All of our capital budgeting tools depend on a whole bunch of assumptions

that we have to make about what's going to happen in the future.

Forecasting, you might be getting the sense now

from this sort of loose discussion, is more of an art than a science.

All of this really, finance and economics is a social science.

And so because it's involving decision-making under uncertainty,

we never really know exactly what's going to happen.

And so mapping out future cash flows, getting

a sense for how they hang together or how sensitive they

might be to different assumptions can be sometimes

a qualitative, artistic sort of use of financial decision-making.

There's never 100% right answer.

We're never really going to know, except ex post,

whether we should or shouldn't have done the project.

But that shouldn't stop us from trying to make the best forecasts possible

and making arm's length transparent and objective capital budgeting decisions.

We have to understand the limitations of the things

that we're calculating in order to feel comfortable

making good financial decisions.