Session 16 Investment Returns III Wrapping up Loose Ends

In this session 16 of a 36 session corporate finance class, I'd like to talk about refinements on measuring investment returns. I'd like to start off by talking about how to bring in different currencies into your investment analysis, how to deal with side costs and side benefits, and how to deal with the uncertainty that you will face when you're done with doing your analysis. So we're now on the final piece of the investment returns puzzle. In the last two sessions, we've talked first about accounting measures of returns. In the last session, we've talked about incremental cash flow returns that are time-weighted. In this session, I want to wrap up some loose ends, and there are quite a few loose ends we want to talk about here. Things like how do you deal with different currencies? What do you do about uncertainty in your investment analysis? Because every number you use is an assumption. So let's start with going back to the dollar cash flows that I estimated for the Disney theme park. Remember, this is real Disney, and I admitted upfront that I was cheating by giving you all the cash flows in U.S. dollars. I admitted that the cash flows are going to be nominal REIs, but I did both my cash flows and my discount rate in U.S. dollar terms because I assumed that you as a decision maker in Disney were probably more comfortable with dollar-based numbers. So the net present value that we got using U.S. dollar cash flows and a U.S. dollar cost to capital of 8.46% was 3,296 million, right? So let me ask you a question. If I'd done the entire analysis in nominal REIs, do you think I should have got a different net present value for this theme park? Think about it. If your answer is yes, we're in big trouble because that effectively means that the conclusions we draw for a theme park, the value of a theme park, or the value of any business will be a function of what currency we choose to do it in. In other words, you could have a project that looks good in U.S. dollars, looks bad in nominal REIs, goes back to looking good in euros, and back to bad using pesos. That makes no sense. So if you do this right, you would like to see the same net present value using different currencies, right? You think no chance of that happening? Well, let me take the Disney theme park and see what would happen if I did everything in nominal REI terms. What does that mean? Well, I have to convert my cash flows from U.S. dollars to nominal REIs. So here's how I did it. I know what my cash flows are in U.S. dollars, at least I've estimated them in U.S. dollars. To convert them into nominal REIs, I need an expected exchange rate, right? Here's what I cannot do. I cannot take today's exchange rate, which was at the time of the assessment, 2.35 REIs per dollar, and use them all the way through. I have to get expected exchange rates for the future. Here's how I get the expected exchange rates. I use what's called a very old principle in foreign exchange. And what it effectively allows me to do is take the differential inflation rate, which is 9% in REIs and 2% in dollars, numbers we've used already, and get an expected exchange rate going forward every year for the next 10 years. And notice, because the inflation rate in nominal REIs is much higher than the inflation rate in dollars, I expect the REI to depreciate over time. And it's that depreciated exchange rate that gives me my nominal REI cash flows. In effect, what I'm allowing my nominal REI cash flows to do is grow at a faster rate because there's a higher inflation rate in nominal REIs. This is good, right? You have higher cash flows. Here's the other shoe waiting to drop. Remember that 8.46% cost of capital? That was a U.S. dollar cost of capital. To convert it into a nominal REI cost of capital, I use a technique we've used already. Take the differential inflation rate, 9% in REI, 2% in dollars, and scale up the cost of capital, which gives me a 15.91% cost of capital, which is the discount rate I'm going to use on the much higher cash flows. See what's going to happen? What one hand gives you, the other hand will take away. My cash flows are higher. So is my discount rate. The net present value that I get using that discount rate is 7,745 million REIs. If I convert it back at the existing exchange rate or the prevailing exchange rate of 2.35, I get 3,296 million, exactly the same net present value that I got using U.S. dollar cash flows and a U.S. dollar cost of capital. To me, this is incredibly good news because here's what I read out of it. When you're going to do a project analysis, pick one currency. I don't care which one it is and do it right. Don't try to do five analyses and five different currencies and do them all incorrectly. It doesn't help. So if you're consistent, currency should not matter, either in a capital budgeting exercise or an evaluation, because the inflation feeds into your cash flows, it should also feed into your discount rate. Now let's talk about uncertainty. If you think about how I put together the numbers for this theme park, I made a bunch of assumptions about revenues, about tax rates, about expenses, and every single one of these assumptions can be wrong, either because of firm-specific reasons. I underestimate or overestimate the number of tourists coming to the theme park. My expenses are higher or lower, or because of macroeconomic factors. My tax rates could vary. The government could, you know, the country risk in Brazil could pop up. So there are lots of uncertainties associated with this or any investment analysis, and one of the classic problems analysts face once they're done with the number crunching is what to do about the uncertainty. So here are your choices. The first is you can wait for the uncertainty to be resolved. You know how long you'll be waiting? Basically you'll never invest, because there's always going to be new sources of uncertainty. The second is you can be conservative and say, I will never invest in something if there's uncertainty, in which case you're going to be stuck in cash for the rest of your life. The third is you might be one of those special personalities that doesn't get phased by uncertainty. You're okay, you can ignore it. Or the fourth is maybe you come up with something else. I think most of us would look for at least a comfort zone for dealing with uncertainty, and I'm going to suggest three ways of dealing with uncertainty, ranging from a very old-fashioned way of doing it, which I think is still effective, to much more newfangled ways of dealing with uncertainty. Here's the first one. When you make a big investment, four billion, four and a half billion, five billion, one of the big questions you often face is, in addition to all the traditional risks, when will I make my money back? In other words, once you make your initial investment back, you feel a lot better, right? That's called payback. Essentially you're looking at the number of years it will take you to make your initial investment back. In the case of the Disney theme park, for instance, it'll take about 10 years to make your money back if you just add up the nominal cash flows. Basically, all I'm doing is keeping track of my accumulated cash flow. Between year 10 and year 11, I make my money back. If you want to make your money back with the return you should earn, given the risk you've taken on, it'll take a little longer, about 17 years. You're saying, what do I do with that? Many companies actually have payback constraints. In other words, they will not take an investment if it takes more than X number of years to pay back, 5, 7, 10. You're saying it's arbitrary? It might be, but it's a way in which you can bring in the fact that you're uncertain about the numbers and the more uncertain you are, the tighter the constraint might become. If you're in a technology business where you're worried about the technology shifting, you might demand that your project start to pay back in less than 3 years, whereas if you feel that the cash flows are more predictable, you might settle for a 6 or an 8 or a 10-year payback. So that's the first approach. It's been around a long time. It doesn't take a lot of doing to come up with paybacks. The second, and this is something that's become much more common, post-Excel spreadsheets, is to ask what-if questions. Before you had spreadsheets, when you actually had to work through numbers by hand, there was a natural constraint on how many what-if questions you could ask because you had to work through the numbers by hand. That effectively meant that you didn't ask too many questions. Now with Excel, it's so easy to ask questions. In fact, it's often far too easy to ask what-if questions and analysts often get carried away. I often see analyses where people punch out 50 what-ifs, 100 what-ifs. They take every input and they ask a what-if question on it. And when you do that, you drown in numbers. Effectively, you undercut why you ask what-if questions. In other words, you lose sight of the ultimate objective, which is you want to make a better decision. So I have a couple of suggestions if you decide to go this what-if route. And every one of you will probably go down this what-if route if you do investment analysis. The first is, less is more. Don't ask what-if questions about every single input. Focus on two or three key inputs and make your what-if questions about those inputs. It'll be much more informative, much more useful. Second, when you ask what-if questions, rather than churning out table after table after table, scenario after scenario after scenario, try to present your what-if findings in as creative a way as you can. A way in which decision makers can look at a graph, a picture, and say, hey, you know what? This is how I can refine my decision. So let me give you an example of a really creative chart. And this is from a book by a guy called Tufti, T-U-F-T-I. And this is a book called The Visual Display of Information. It's one of my favorite books. And in fact, in that book is one of my favorite graphs of all time. It's a graph that looks at Napoleon's ill-fated expedition to invade Russia in the 1800s. So in this graph, you actually see the size of Napoleon's army as the width of that bar at the start. And you see the army marching from Paris. It's a big, hefty army. It takes off across Europe. And it starts marching all the way across Europe into Russia. See the end of that first, the light bar? That's when they get to the outskirts of Moscow. And it's about December 15th. And if you've ever been in Moscow around December 15th, it's very cold. And that's basically what Napoleon found. He said, it's real cold. I want to go back. And that's basically the army marching back, the black bar. And you can see basically the army getting smaller and smaller. And in fact, on this graph, on the other axis, you see the temperature dropping. This is a great graph. You look at that graph, you can see exactly why Napoleon failed in his attempt to invade Russia in the 1800s. Your task is far simpler. All you have to tell me is how my decision will be affected if my market share is bigger or smaller. Be more creative about explaining what that effect is going to be. So you can compute the payback. You can ask what if questions. Here's the third and final way in which you can bring in or deal with uncertainty. One of the big problems we face in capital budgeting is I'm asking you to make a point estimate for a number we feel uncertain. Let me give you an example. The Disney theme park, I asked you what the revenues would be each year, the number of tourists coming in. And part of you were saying, well, I don't know. It could be higher, it could be lower. Same thing with the operating expenses. I asked you what that would be as a percentage of revenues. And you might say, well, it's 60%. But it could be higher, it could be lower. It's different in Rio. And the same thing with the country risk premium. You gave me a country risk premium of 3.95%. But that was based on the facts as they were at the start of this assessment. Or 3% was the base premium. So with every one of these numbers, I asked you to give me a point estimate. And you really wanted to give me a distribution. What if I allowed you to give me a distribution? So here, for instance, I've allowed for distributions in each of these three inputs. On the revenues, for instance, I've allowed the revenues to be 20% less to 20% more than expected. And the distribution is uniform. Basically, what that means is there's an equal chance it could be 20% below my estimates or 20% above. That's called a uniform distribution. For the operating expenses of percentage revenues, I made that a normal distribution centered around the 60%. That's my base case. But because of the uncertainty, I feel it could be as low as 45 or as high as 75. And for the country risk premium, while it was centered around 3%, again, I allowed for the fact that Brazil could get riskier over time or safer over time. The premium could be lower or larger. Now you're saying, what do I do with this? Well, what we're going to do is called a Monte Carlo simulation. And here's what's going to happen. I'm going to go pick one outcome out of each of your distributions. Let me explain what I mean by that. I'm going to pick a revenue number. It could be higher than your base case or lower than your base case. I'm going to pick an operating expense number as a revenues. And I'm going to pick a country risk premium. I'm going to compute the net present value. And I'm going to do this again and again and again. In fact, the default on the program that I used to run this simulation called Crystal Ball is 100,000 simulations. And what it does, it runs 100,000 simulations. It takes it about a minute. And it comes back with a distribution of net present value. You say, what do I do with this distribution? The median value is about 3.3 billion, which is very close to what I got in my base case. You say, I didn't learn much by doing this. Well, here's what you learn in addition. If you look at the outer ends of the distribution, the net present values range from minus 1 billion to 8.5 billion. I don't have to do any best case or worst case scenarios. They're built into my simulation. In addition, remember what I'm worried about here, right? That this project could be a bad project. If I count the number of outcomes under zero, it's about 12% of the time this project could be a negative net present value project. So if you're a decision maker, think of how much richer your decision can be now. In addition to giving you the net present value for the project, I can tell you the best case and the worst case. I can give you a standard deviation. I can tell you the percentage of time this project would be a bad project. And you can make a much better decision, I think, based on this whole set of numbers rather than just a single net present value. 25 or 30 years ago, doing a Monte Carlo simulation was out of the reach for most of us because it's expensive and you needed access to a mainframe computer. That's no longer true. If you have a PC or a Mac and you have Microsoft Excel installed on it or some spreadsheet program, you can add on a simulation program for almost nothing or at least very little. So if you get a chance, try a simulation program out. It allows you to face up to uncertainty and put a picture on it. The other issue to face here, of course, is when you look at an investment like this and you face a lot of uncertainty, part of you wants to get rid of the uncertainty. And increasingly, you're given ways of hedging risk. So one of the risks that Disney faces, for instance, on this particular theme park is exchange rate risk. Exchange rate risk in what sense? The cash flows are going to be in REI. You don't know what the exchange rate will be. You can go out and hedge that exchange rate risk. So I have a question. Given that you can buy that hedge, protect yourself against that risk, should you? Of course, if I can do it. Well, let's see what the trade-off is when you think about whether you should do it. This is the picture that I use to make a judgment on whether a company should hedge exchange rate risk. I start with a basic question. What does it cost you to hedge the risk? If your answer, it costs you nothing. If you can hedge this through your operations, for instance, Disney might be able to take other investments in Latin America, which move in another direction when exchange rates move. If you can hedge costlessly, it's a no-brainer. Of course, you should hedge exchange rate risk. If it's costly, if it's expensive to hedge the risk, then the choice becomes a more difficult one. The question you've got to ask yourself is, can your investors hedge this risk? And to hedge this risk, your investors have to be able to see the risk, right? So if they can see the risk and hedge the risk, then ask yourself, what will cost your investors to hedge the risk? If it will cost your investors less to hedge the risk than it costs you, let them hedge the risk. I would argue that 80% of the risks you see hedged in the market right now, companies would be better off passing that risk through to their investors because it would be less expensive for investors to hedge that risk than for companies to do it. So whether you should hedge risk or not doesn't just depend on whether there are products out there to hedge the risk, but whether it makes economic sense for you to hedge that risk. Those are loose ends to tie up. Here's the final piece of the puzzle. When you look at investments, especially investments made by big companies, there are always side costs and side benefits. Side costs in what sense? If Disney opens this real theme park, maybe there will be employees that work for Disney already that might be transferred to this theme park or other resources that Disney owns right now that will find themselves in use in this theme park. That's an opportunity cost. I've got to factor that in. There might be side benefits. By opening this theme park in Rio, Disney might increase its potential movie revenues in Brazil or its merchandise sales in Latin America. Those are side benefits. I don't have an argument against these side costs or side benefits. The only point I would make is if this is the argument you're making, then you have an obligation to bring them into your expected cash flows and incorporate them into your returns. If there are side benefits from this theme park, let's bring them in as additional positive cash flows. There are side costs. Let's build them into the expenses. The return that you should end up with should incorporate all your side costs and side benefits. In summary, when you think about measuring investment returns, here's what you want. You want a cash flow-based measure of return. It should be time-weighted and should have all side costs and side benefits built into it. In these last three sessions, we've laid out how to get from just the numbers to these final measures of returns. Once you get those returns, you can compare them to the hurdle rate that you estimated earlier in the process to make a judgment on whether this investment is a value-creating, value-neutral, or value-destroying investment. Thank you very much.