Thinking about Business but parents say it's not a "safe" career...

[sakky]

<p>
[quote]
I don't think NPV assumes any of that. It's simply a recognition that resources are more valuable early in the cycle. Either you could invest it, or you had to borrow it, or you had to spend other money that you could have invested instead </p>

<p>

[/quote]
Uh, no, I'm afraid that that is precisely what NPV assumes. See below.</p>

<p>
[quote]
</p>

<p>It's not that anybody actually is saving this much money. NPV is simply a recognition of the increased value of the resource early on

[/quote]
</p>

<p>You are correct - NPV presumes that resources are more valuable early in the cycle. But think of it this way. Let's say that the real interest rate is 10%, meaning that real prices double every 7 years. Was the hamburger I ate 7 years ago really twice as "valuable" to me as the hamburger I eat tonight? Not really. A hamburger is a hamburger. In each case, I "employed" the hamburger to satisfy my hunger on that particular night. On the other hand, a straight NPV calculation would indeed imply that the hamburger I eat tonight is indeed only half as "valuable" as the burger I ate in the past.</p>

<p>Let me put it to you another way. NPV calculations inherently presume that you have a choice about what to do with the excess funds that you generate in earlier periods. But, the truth is, in this context, you don't. For example, according to a straight, bone-dry NPV calculation, if I had simply forgone eating 7 years ago for an entire year, then the funds I would have saved during that year could have been invested such that today I can afford to eat twice as much. But come on, not eating for an entire year is obviously not an option. Similarly, I could say that if I could be homeless for a whole year, then 7 years in the future, I would be able to afford house payments that are twice as high. But again, choosing to be homeless for a whole year is just not a reasonable option for most people. But that is precisely what NPV implies. In other words, while there is such a thing as the 'time value of money', there is no such thing as the 'time value of food or housing'. These are things that you just need, and need immediately. </p>

<p>You can think of it this way. If you want to use accounting terms, you can think of your regular living costs as just a form of "necessary depreciation" to keep your business operating. That would mean that you are to discount your excess *cash flows<a href="as%20UCLAri%20stated">/i</a>. Funds that are going to simply support your life are not excess cash flows. They become part of doing business, and are therefore not available for future investment.</p>

[sakky]

<p>
[quote]
sakky,</p>

<p>There are plenty of things someone who's "good enough" to be a doctor can do other than medicine. Assuming I wanted to get a job in investment, I probably could. That would mean that I'd save myself another 5 or 6 (thanks to post-bac) years of education. Most of my job offers so far have been in the $65K+ range, with good odds of making at least $100K by the time I'm 30 or so. </p>

<p>Medicine is simply not a financially wise decision for me, considering the time value of money, sunk cost, and opportunity cost. There are many others who I'd assume are like me.</p>

<p>Hell, I bet that most doctors could be lawyers and make at least what the average family practitioner makes

[/quote]
</p>

<p>Hey, I'm not saying that being a doctor is a killer financial deal either.</p>

<p>I am simply saying that a straight NPV calculation is not the most realistic way to compare different careers. Like I said, there are certain expenses that you just have to pay no matter what career you embark upon. Again, you can't just decide not to eat for a whole year.</p>

[UCLAri]

<p>It's not the most realistic, but let's face it: we make plenty of assumptions in any financial analysis. How many really good Monte Carlo simulations have you actually seen?</p>

<p>I think what bigred is getting at is that there are plenty of other good, financially sound careers that one can enter that are likely to yield solid long-run incomes. Medicine, despite its prestige and "cocktail party value" is really NOT a place to go make money, no matter how much spine surgeons make. </p>

<p>I'd say that on a per hour basis, doctors do much worse than most of the execs selling them medical hardware. If you really want to make a killing, big pharma isn't a bad place... Though I recently read something from HBS suggesting that generic pressure is making that a more "interesting" field. Five forces, anyone?</p>

<p>Oh well, I'm still going into med anyway. I can't imagine a life in front of Excel and Stata.</p>

[bluedevilmike]

<p>Sakky: I'm trying to explain that NPV is not a representation of end wealth. It's a representation of the fact that the money is worth more early on. There are no assumptions about eating, rent, driving, etc. because those simply aren't part of the model. The money is worth more because it's earlier.</p>

<p>EDIT: Sorry, missed your post #41.</p>

<p>The bottom line is that you'd be buying a hamburger anyway. Maybe you'd have borrowed money from a bank to do so. Maybe you'd have spent the money out of previous savings, which you could otherwise have invested. But that's a fixed expenditure.</p>

<p>The money you earn, then, is in fact earning returns -- whether in interest that you don't have to pay because you didn't have to take out debt, or because you preserve your previous savings allowing you to invest those. There's an opportunity aspect involved here.</p>

<hr>

<p>There's one solid argument I can think of against the usefulness of an NPV model, but (so far as I can tell), you don't seem to be making it.</p>

<p>The argument would go:
It's better to earn all your income in lump sums rather than in a more spread out fashion (i.e. to be a doctor) because the relative early poverty might cause you to save more early on, thus increasing your net wealth later in life.</p>

<p>The answer, of course, is: just be frugal early on.</p>

[sakky]

<p>
[quote]
Sakky: I'm trying to explain that NPV is not a representation of end wealth. It's a representation of the fact that the money is worth more early on. There are no assumptions about eating, rent, driving, etc. because those simply aren't part of the model. The money is worth more because it's earlier.

[/quote]
</p>

<p>Yeah, but see, you are implicitly conceding the weakness of the model. NPV is not really a representation of "money" per se. Rather, it's a representation of "excess money" - that is, money that does not immediately have to be consumed in order to keep the operation going. </p>

<p>Let me give you an example. Let's say I have a business opportunity in which I have to pay a business license fee every year while I'm in operation (or else the government shuts me down). So if my business investments are going to provide payoffs 10 years in the future (and I get nothing in the interim), then it is entirely proper for me to calculate an NPV that factors in all of those business fees I will be paying from years 1-9 before I get a positive payoff in year 10. And yes, those business fees are probably going to be obtained by borrowing (hence the time value of money of loans from banks).</p>

<p>But - here is the key difference - I can also choose not to pay those fees at all. In other words, I can shut down the business and not pay anything at all. If I find the total NPV of the opportunity to be negative, I will simply shut down the business. Hence, I can opt out of the business. But people don't have that option. People can't "opt out of life".. Barring suicide, people can't just simply decide that they're not going to pay their costs of living. No matter what career you choose, you still have to eat, you still have to house yourself, you still have to pay for your regular life. These are expenses that never go away. They are incurred whether you like it or not. </p>

<p>What that means is that whatever money you make that goes towards your living costs is not earning returns. It is not. That's different from the business case. I could simply out of the business, take the money that I would have spent on those licenses, and invest that money. But I can't just decide not to eat for a whole year and invest that money. That's the fundamental differences. These living costs are incurred whether we like it or not. They are not truly available "for investment:, which is precisely what NPV implies.</p>

[bluedevilmike]

<p>I concede perfectly well that my first paragraph was non-responsive to your post #41. (I hadn't seen it when I first posted.)</p>

<p>I think the opportunity discussion about money -- earning enough to pay for the burger now opens up other funds to invest, or at least to not borrow -- remains.</p>

[sakky]

<p>I don't disagree that the concepts of NPV can be used to provide some level of financial comparison. What I am saying is that real-world choices are more complicated than that. </p>

<p>
[quote]
I think the opportunity discussion about money -- earning enough to pay for the burger now opens up other funds to invest, or at least to not borrow -- remains.

[/quote]
</p>

<p>Even here is yet another prime assumption, if you are going to apply NPV. I don't remember how you calculated your NPV, but I assume you used some fairly high real interest rate (i.e. probably r=7-10%, which is probably derived from the real rates of returns of US equities over the past 50-100 years or so). But should that rate really be used through your entire calculation? I don't think so. The presumption of a constant interest rate within an NPV calculation is that you always have the choice to make such investments that pay that particular rate all the time, anytime you want, and make payouts risklessly. </p>

<p>But this is simply not true. Equity returns vary all the time. Sometimes (like this year), equities will do gangbusters. Other times (i.e. 2001), equitiies will go in the toilet. Sure, over the long run, one can say that equity returns are around 7-10%. But as Keynes once famously said: "In the long run, we're all dead". </p>

<p>Specifically, my living costs are basically constant, but my equity investment returns are not. I can't just simply decide to not pay for housing or pay for food just because my mutual fund has a bad year. I can't just tell my landlord that because markets did poorly in one year, I don't have to pay him, but that I'll pay him later (with interest) once the markets inevitably turn around. If I don't pay him when the rent is due, he is going to evict me. He wants the rent right now, regardless of the state of the markets. </p>

<p>In other words, my investments are risky, but my expenses are constant. NPV calculations inherently presume perfect capital and insurance marketes such that I could perfectly hedge all of my risk away and hence transform my investment payouts into a stream of constant payments. But of course such perfect markets don't exist. There is no way that I can get access to an equity-rate of return without exposing myself to equity-style risk. </p>

<p>Hence, let's take it back to the example at hand. You have person A who became a doctor and hence assumed a boatload of debt to do so. But that debt interest is constant. He knows exactly what "return" he would get by paying down that debt. Hence, there is no risk. On the other hand, you have person B who takes a regular job and has some extra funds to invest - but he has to invest in risky securities (i.e. equities). Hence, person A has the benefit of a "riskless" interest rate on investments that person B does not have. </p>

<p>Keep in mind that person A, once he's a fully fledged doctor, also has the choice to invest in the same risky securities that person B does. For example, he could just pay the absolute minimum every month on his loans to avoid default and take the rest of his excess income and invest in stocks. On the other hand, he can use his excess income to pay down his debt. Hence, person A's "investment portfolio choices" are at least as good as person B's, because A's is a superset of B's. {For the same reason, mortgages are also a great investment because you can always choose to prepay your mortgage and thus lock in a risk-free return.} </p>

<p>A more fair way to calculate NPV is to incorporate a "risk-premium" for person B's interest rate. But the salient point is that the investment risks that each person is faced with are quite different.</p>

[bluedevilmike]

<p>For the record, I used 7%, which struck me as excessively conservative. It's too far past my bedtime to think about risk adjustments right now.</p>

<p>Among other things, it's not clear to me that risk adjustments are necessary in a long-term window, since we're ultimately talking about a function which is related to lifetime earnings. The $1000 I owe on my rent in a bad month could either be paid for by a job or with debt, but the opportunity cost is still relative to a $1000 which will be considered over a lifetime.</p>

<p>But maybe in the morning, when I'm thinking more clearly, I'll realize that you are in fact correct.</p>

[sakky]

<p>
[quote]
For the record, I used 7%, which struck me as excessively conservative. It's too far past my bedtime to think about risk adjustments right now.

[/quote]
</p>

<p>I personally think 7% is about right (on a real basis, not a nominal basis). </p>

<p>
[quote]
Among other things, it's not clear to me that risk adjustments are necessary in a long-term window, since we're ultimately talking about a function which is related to lifetime earnings. The $1000 I owe on my rent in a bad month could either be paid for by a job or with debt, but the opportunity cost is still relative to a $1000 which will be considered over a lifetime

[/quote]
</p>

<p>But again, that gets down to the problems of NPV when you're talking about people's lives. Again, NPV calculations presume that you have no existing financial commitments that need to be met right now. For example, the fact that I need money to pay for food so that I can eat today doesn't really depend on my lifetime earnings. The restaurant wants to be paid right now. I can't just give the restaurant a bunch of IOU's that say that while I don't have money right now, I will have lots of money over my lifetime, so they should give me a hamburger right now for which I can pay them later.</p>

<p>Now, I know what you're thinking. You're thinking that, theoretically, I could be able to borrow the funds I need so that I can eat right now, and hence shift future earnings to present dollars. However, that presumes perfect capital markets, which do not exist.</p>

<p>Think of it this way. If perfect capital markets really did exist such that you really could transform total lifetime earnings into a stream of steady and smooth payments over your lifetime (through perfect lending tools and whatnot), then investment firms should also be able to use these tools to smooth their earnings. Wall Street doesn't like volatility either. If a mutual fund really could transform an instrument that returns 7% a year but in a highly variable manner for an instrument that returns the same 7% a year, but smoothly and steadily, you know that that fund would strongly prefer the latter. So why don't all equity funds (or at least, all equity index funds) just do that? Why put up with volatility if you don't have to?</p>

<p>I think the answer is that they do have to put up with it. Specifically, there is a quite strong correlation between risk and reward. The greater financial returns you want, the more market-correlated risk you have to bear (which is a direct outcome of CAPM). In other words, if a fund wants to smooth its earnings and hence reduce its risk, then it must forfeit some returns.</p>

<p>So, what does that segue into financial markets have to do with what we're talking about? Simply that it's highly unrealistic to expect that one's lifetime earnings can be neatly transformed into a stream of steady and riskless payments at a reasonable interest rate, and especially not at a 7% real interest rate. If financial firms can't perform that transformation with their own instruments, then people obviously can't do that with their own lifetime earnings. NPV calculations inherently presume that you can always transmute payouts at will back and forth through time, all at the same interest rate. In the real world, this is of course impossibe.</p>

<p>We also have a significant problem of information in the market. I can't just go into a bank and get a huge loan at a low (i.e. 7% real) interest rate. They're going to want to know what I am planning to do with that money, what my occupation is, and basically assess whether I am really likely to pay them back. Hence, what my lifetime earnings are is irrelevent, what matters is what I can convince the bank what my lifetime earnings are going to be. This is a problem that is riddled with asymmetric information. Banks are rightfully scared of the ex-post information problem that maybe I am going to take their loan and squander it, and then declare personal bankruptcy and hence never pay them back. Or maybe I'm just planning to take the money and run off to live in luxury in some poor foreign country where they'll never see me again. Banks are also scared of the ex-ante adverse selection problem that maybe only highly irresponsible people are the ones applying for the loan in the first place (because highly frugal and responsible people don't need loans). Again, whether I am this kind of person is information that only I know. The bank doesn't know. I can tell the bank that I am not that kind of person, but of course the problem is getting the bank to believe me. </p>

<p>What a doctor/med-student possesses that I don't is a highly credible market signal that they are in fact a highly responsible and dedicated person. I think it's safe to say that practically nobody actually graduates from med-school who isn't quite responsible and dedicated. That's why banks are happy to lend such people at relatively reasonable interest rates the 6-figure uncollateralized loans necessary to get a medical education, but aren't willing to loan a person like me anything like that kind of money without collateral. {For example, the only way a person like me could get a large loan like that from a bank at a reasonable interest rate is for a home mortgage, but then the house would serve as collateral.} </p>

<p>Think about what that means. In theory, a guy like me may generate more lifetime NPV than a doctor would. But banks are willing to provide more debt financing at better terms to that doctor than to me. What's wrong with this picture? Are banks just being dumb? Actually, I would say that nothing is wrong with the picture and that banks are not dumb. Rather, it speaks to the unrealistic assumptions of the NPV model. The NPV model presumes perfect capital markets with perfect information, such that everybody truly can transmute payments frictionlessly back and forth through time. In the real world, this is just not so.</p>

[JP_Omnipotence]

<p>Undergrad business can be a little risky, but if you get an undergrad degree in Engineering you'll have something to fall back on before you get your MBA (only about 20% of harvard's mba students were undergrad business) and something to use to get a job in the rare care your mba fails you. And hey, who knows? That engineering job might help you get a promotion to a top business position in your company.</p>

[bluedevilmike]

<p>Something still feels funny, but I'm unable to articulate my feeling. Therefore: to the best of my ability to explain it, NPV is a flawed metric because it usually involves assuming that everybody is able to borrow or loan money at a standard rate.</p>

<p>If those things differ -- say, if a doctor is able to invest more intelligently than an auto mechanic, or if a lawyer is given debt at a lower rate -- then NPV calculations will be thrown off.</p>

<p>To the best of my estimation, the difference in loans shouldn't make too much of a difference. The professions that have to take out loans -- mostly medicine and law -- are getting pretty good financing on their debt anyway. I'm not familiar with the numbers, but I believe sakky above mentioned that 7% was pretty good, so then it appears that the overall assumptions are unchanged.</p>

<p>As for differences in investment skill, let's just say that I think people who do it as a profession are going to be a little better than people who spend all day learning anatomy. If anything this will widen the gap.</p>

[sakky]

<p>
[quote]
Something still feels funny, but I'm unable to articulate my feeling. Therefore: to the best of my ability to explain it, NPV is a flawed metric because it usually involves assuming that everybody is able to borrow or loan money at a standard rate.

[/quote]
</p>

<p>
[quote]
To the best of my estimation, the difference in loans shouldn't make too much of a difference. The professions that have to take out loans -- mostly medicine and law -- are getting pretty good financing on their debt anyway. I'm not familiar with the numbers, but I believe sakky above mentioned that 7% was pretty good, so then it appears that the overall assumptions are unchanged

[/quote]
</p>

<p>I think what you just said right there holds the key to understanding what is wrong with NPV. </p>

<p>Doctors and lawyers probably can borrow at a 7% real rate, hence reliably and risklessly moving 'future' money into the present at that rate. But can I move my present money into the future at that same 7% riskless real rate? No, I cannot. I can invest in equities and perhaps obtain that rate on average. But my return would be uncertain, and therefore risky. I wouldn't really be able to plan how much money I would have at any point in the future, because my equity investment would be risky and therefore subject to great variance. Or, I can invest in highly safe securities, i.e. T-bonds, but then I think it's quite clear that I would not be obtaining a 7% real rate. Maybe 2%. {For example, I see that right now, infation is ~2.5%, and the yield on a 10-year Treasury is about 4.3%, for a real rate of 1.8%.}</p>

<p>See, right there, those numbers illustrate the problem. It is far easier to risklessly move money forward in time than to move it backwards in time. In other words, real-world financial markets are chronologically asymmetric. NPV presumes perfect symmetry. In other words, the person who decides not to go to med school, but instead just gets a job and saves/invests his money, is not going to be able to reliably and risklessly defer his money at a 7% real rate. He can do so only by taking on risk, but NPV presumes no financial risk. </p>

<p>
[quote]
As for differences in investment skill, let's just say that I think people who do it as a profession are going to be a little better than people who spend all day learning anatomy. If anything this will widen the gap.

[/quote]
</p>

<p>Well, let me put it to you this way. If it really was so easy to find an investment that really did provide a 7% real return, then why would anybody ever put up with the measly returns on US T-bonds? I believe that US T-bonds are the largest single investment security class in the entire country, and perhaps in the entire world. Why are they so popular, if there really is some other investment out there that is just as safe, and offers more than quadruple the real returns? I would think that surely if these investment professionals were smart, they would have found such an investment opportunity by now, if it really existed. </p>

<p>
[quote]
If those things differ -- say, if a doctor is able to invest more intelligently than an auto mechanic, or if a lawyer is given debt at a lower rate -- then NPV calculations will be thrown off.

[/quote]
</p>

<p>Yes, exactly right. Let me try to explicate further. The NPV method was borne from corporate finance, to calculate the worthiness of projects. One problem with NPV is that it presumes that you can obtain all the funding that you need frictionlessly. For example, let's say that I am running a small business, and I discover a project that is NPV-positive, but that requires a massive initial cash outflow (let's say a trillion dollars for sake of argument). According to NPV theory, the initial payout shouldn't matter at all, as I should still choose the project because it ultimately produces a positive NPV. But of course in the real world, it DOES matter, because before I can even start the project, I need to somehow obtain that trillion dollars in the first place. No company, not even Microsoft, has a trillion dollars in cash just lying around. I would have to go to the financial markets to obtain that financing, yet the fact is, real-world financial markets are not going to just front a small company a trillion dollars at a reasonable interest rate. Heck, they probably wouldn't provide that amount of capital at ANY interest rate. </p>

<p>NPV calculations presume that I can simply "prove" to the markets that my NPV opportunity is so compelling that I really should be able to procure that trillion dollars in financing at a reasonable interest rate. In other words, we have to assume that there are no information problems, and in particular, no asymmetric information problems. In the real world, if I'm a small businessman who claims to have an idea that is worthy of a trillion dollars worth of initial investment, I am going to have a devil of a time getting anybody to believe me. </p>

<p>Now of course, you're probably thinking that a trillion dollars is a ridiculous amount of money. Of course it is, but I choose it to illustrate a point. The fact is, financial markets are riddled with information problems. That's why financiers demand to have corporate control mechanisms whenever they provide financing. For example, if a company applies for a loan of any amount, the bank will usually institute debt covenants to make sure that you don't just abscond with the money. If you are floating a bond, then the bond prospectus will state what restrictions you are placing on yoursef to provide higher confidence that bondowners get their money back (i.e. the collateral they can claim, their claim priority, etc.) If you procure capital through an equity offering, then new equity owners will usually demand voting power and other shareholder rights. None of these tenets would be necessary if not for the information problems inherent in financial markets. </p>

<p>Hence, put another way, again, let's consider person X (who has been admitted to med-school) and person Y (who has not been admitted to med-school). X will take on, say, $200,000 of debt at a 7% real rate. But person Y does not have this financing "opportunity". He can't just go to the bank and ask for $200,000 at a 7% real rate, not without any collateral (and person X wasn't offering any collateral). Y will probably have to pay a much higher real rate to get that kind of loan, or, more likely, won't even be able to get the money at all. In other words, X has a financial "opportunity" that Y simply does not have. </p>

<p>Just like if I'm a small company and I discover a NPV-positive project that requires a billion dollar initial payout, I can't realistically do it, because I can't get the funding. Microsoft has billions of dollars in cash lying around, so they can do it. But I can't. Hence, Microsoft's opportunity set is larger than mine.</p>

[UCLAri]

<p>sakky,</p>

<p>You do a lot more finance than me, so I'm curious: what would you say is the best alternative to NPV? IRR?</p>

[sakky]

<p>IRR ain't that great either, because it suffers from the same many of the same problems that NPV does, and in particular, assumes that future cash flows are riskless (in that you know what future cash flows will be with certainty, or at least within a highly bounded set). Yet the truth of the matter is that future cash flows of any business investment are not only inherently highly uncertain, but also tend to be serially correlated (for example, a business project that produces an early cash flow that is much higher than you had anticipated it would be will probably continue to produce unexpectedly high cash flows for the lifetime of the entire project.) IRR and NPV also presume a level of omniscience that business leaders (or anybody else for that matter) simply do not have. Let's face it. Most companies don't really know exactly how successful a project will be, or whether it will even be successful at all. As a case in point, I highly highly doubt that anybody, including, Bill Gates and Microsoft management, really knew that the MS-DOS/MSWindows and MSOffice franchises were going to be the blowout successes that they became. {Gates probably had an inkling that these were important projects, but I doubt that even he truly knew just how momentously profitable these projects would turn out to be.} </p>

<p>I think a more realistic tool would have to be something that incorporates risk and the ability to adjust in the face of risk. {For example, if early financial results in MSWindows were unpromising, then Microsoft would have probably killed the project lest it continue to sink more money into a probable failure.} It would also have to incorporate some factor that accounts for market imperfections, i.e. financial imperfections. Something like real options adjusted for the true interest rates (forward and backwards) that you can use. Yet even the technique of real options is problematic because it presumes that the stochastic nature of your 'options' follows conforms to certain statistical parameters that probably don't hold in any actual 'option' that you face in real life.</p>

[UCLAri]

<p>What do you think of the recent Monte Carlo work? To be honest, I always felt that NPV was only as good as your information, but I'll be damned if the finance profs ever would concede that.</p>

<p>This is by no means my area of interest, but since I'm being force-fed the stuff, I might as well TRY to understand it...</p>

[sakky]

<p>Even Monte Carlo work has its limits. The notion that Monte Carlo (or any simulation) accounts for information limitations is a mirage - as the probability distributions that you input into your simulation is itself subject to information limitations. After all, how do you really know that some probability distribution really is normal or uniform or exponential, or whatever it is that you're inputting? Sure, you can look at historical trends, but as has been said by portfolio managers countless times: past trends are no guarantee of future returns.</p>