Masters of Finance: Robert C. Merton

when I was a kid growing up I had one skill I had was I was able to take long lists of numbers and add them up pretty quickly and so that's why they all the the employment things that you took in high school so I was going to be an accountant or an engineer because they said whether you rather take you know the dog out for a walk or add up a large column of numbers I said large numbers well that may have been a great skill to have in the nineteenth century but of course with a six dollar calculator you can do it faster and more accurately than I could when I showed up not only was Harold Freeman that on the admissions committee he was the graduate student advisor and I I knew I knew no economics so I was kind of in line especially when remember from the what Paul told every school turned me down except MIT gave me a full fellowship so I was very grateful and I was gonna do whatever they told me so I had done what the book told me and I written economic history and I was gonna take micro and macro and and I went in with my my card and Harold Freeman set back he was white hair just kind of the statistician isn't a nobody he said he says you take that you're gonna be so bored to me by the end of the term he says why don't you go over and take call sent us as mathematical economics of course okay and that's how I go to this course to begin with and I learned my economics backwards because I learned foundations first and I'd like took Bob bishops micro with all the charts and but that's how I found out and I did a growth model actually my fruit not my first published paper because it was published later but my first economics paper was done in Paul's class now it was a growth model with endogenous population growth it was any class so I was interested in a temple optimization for a mathematical point of view and it seemed like an interesting thing then when we got involved and I found out Loren and convertible bond pricing was okay too then I got involved in the uncertainty so you see I so then I thought would it be neat if you can combine uncertainty with intertemporal optimization but the only techniques that I had ever learned anything about were the calculus of variations and then contracted the maximum principle and and so I really didn't know how to do it so I started exploring to find ways to do it when I was a graduate students was not in the mainstream of academic economic research and if I can locate for it I was a graduate student in the Economics Department no course taught in the Economics Department what the modigliani-miller theorem even though Modigliani was here the mean-variance model was taught as a simple example because it was neither its validity was not consistent with expected utility theory and so forth and so it took just just to locate it for you that that's you know the the the the degree of separation from it so when I did my continuous training model what I was imposing on the on the model was the idea that the length of time between trades was short but it turned out that under the sort of prototypical type processes were the distributions of returns remain the same each period and the sample past would continuous those few other things but standard prototype things that led you to portfolio rules which were identical to the mean variance one of the contributions I thought was that it reconciled what was a very neat theory that everybody wanted to use but we're from an academic rigorous point of view you know was not being used within economics because it was felt that the assumptions under which were held were totally inconsistent with reality not just special cases they were just not reality if they couldn't happen you can't have negative prices and you can't have if your distributions to be big enough you can't have quadratic without having truncations and so so so I thought that was one piece that reconciled and that was in there and then from that it allowed me to look at richer models and which you got insights that were about behavior that would never show up in a single period model and they came out of then that was also in this paper I had been doing work in parallel capital structure on pricing warrants and convertibles and so forth and then when we came to the option price of all the Fischer black and Myron Scholes were working on when as I mentioned earlier having gone through all the interview process for a recognized apartness and gone through all the work with the interviews and even getting a couple of job offers at the end Franco came in and said would you like to have a job here and things were going very well here I liked it here and so I had to go meet all the faculty members and one of them was this relatively recent assistant professor appointment of Myron Scholes and after I took the offer and joined I got spent time on Myron and in the interim before actually officially joining but accepting the job we talked about a number of things and eventually found out that area that he was working on in the area of option price saying when I was have been working on we shared that they showed me what they had done which was to take a dynamic strategy trading and at each point in time take a position in the stock in such a fashion to offset the bait or systematic risk of the of the option through the stock so as to create a portfolio with zero beta and once you have that we all know from the equilibrium theory of the spectra turns equal risk free rate and he closed the model and this was actually a fairly Myron and others Mike Jensen and and Fischer a number of people had devised this kind of approach empirically to setting up 0 beta portfolio so that thinking of about it in that way was was I guess that's part of I don't I never was quite sure who contributed which the two between Myron and and Fischer black on that but maybe it doesn't really matter let's say together that's how they did it and so when I went to look at it I course looked at it in the context of the models that I've been doing so I did a continuous version of it and when I did that what happened as I showed in the limit then when you actually did continuously it wasn't just that the systematic risk disappear but all the risk disappeared and then it was pretty clear from that that that meant that there was a dynamic trading strategy that could literally replicate the exact payoff so you could find it binary strategy that when he followed it not an expectation but the actual realization would be identical when fishing Myron published their paper they put both forms of the derivation in their paper Fischer told me and I think it's in public eyes today it's perhaps at the end of his life he died in 1995 that of the two he preferred the capital asset pricing model version because he felt that continuous trading wasn't really you know it wasn't feasible you could understand maybe it's just protections that III thought the other one was that was one interesting once you sort of saw the problem and looked at this thing it was pretty it came pretty quickly that you you know the structure of other securities whether the convertible bonds bonds that might default you know the corporate bonds bonds with warrants attached any of these sorts of things that they were all they all come from you know could be analyzed in the same way and so one of the implications of this was that you could come up with a unified theory for the pricing of the capital structure were firm the right side and to put it in context if you go back and look at the textbooks they taught financing prior to this to locate where things were there was a chapter on pricing equities the prior chapter and preferred stock a chapter on debt and none of them seem to have any relation to one another they were you know was kind of editing up the pieces the firm when you almost added up the pieces rather than recognizing that all the claims on the right side are on one fire against the same set of assets and if they're linked to one another and so they're all contingent so that that element which you could certainly if you're giving credit for I mean it certainly would say money on eemiller this work was fundamentally recognizing that and then and and Fischer and Myron Scholes also were doing the same thing so it's we were all kind of doing all this together Paul had done the work on the American option and shown that it would be early exercise we then in our joint papers show for the models we were using which were utility based so at least they were potential equilibrium models false original papers he would be the first to say was just a positive normative model which just said that the means are constants didn't didn't derive them from an equal in theory and we showed in our in our work with utility based pricing which was consistent with the equilibrium pricing that you wouldn't exercise early unless they'll call unless there were dividends so we we've done that there Paul had early exercise as a mechanical I mean what I'm saying II just as a result of the math that if you have two different means on the warrant and the stock you'll get an early exercise so he was it was the first to really explicitly I think work out the valuation I think where I did on an earlier exercise was to show that it came from in essence whoever the choose the selector was whoever had the choice of exercising the option you would look at the answer that the pattern which would maximize the value because otherwise if it was any other pattern you have a lower value if you sold at that price all I have to do is turn around say aha I was thinking something different and changed the price so the only possible equilibrium was that so I did a bit of work on that too but but Paul's work was the first and that miron I is I guess probably the first time was in 1971 went to Wall Street with the black-scholes model the bad news was that two guys showing up in their 20s talking about partial differential equations and stochastic processes and etc was not something that caught the attention of a lot of people today it's course very different you know you at least can get a hearing from somebody and they don't throw you out but at that time it was totally off the wall and one concrete example of where being involved in practice even early in my career really helped my research was they came to me one day are kind of admiring me one day and they said we're doing a lot of trading of an option in Hong Kong that we aren't quite sure how to value it or what to do with it there's a thing called a down-and-out option which was like a cool option but if the stock price fell below a certain level it was cancelled so you're down in your route alright I said well how do you do that and the way the option pricing model was to arrive it was very clear that you could take any any de Rivoli now called derivatives any security whose prices depends on some other securities price and you could essentially replicate its payoffs by some time their strategy and that all you had to do is be able to characterize what the boundary conditions were that characterized the security so I sort of scratched her head and said oh this looks like a very difficult problem and and of course all it meant was we just had to put the right boundary conditions and so it's like having someone giving you a secret sauce or something that allows you outside like a lego set or something that you can whatever that analogy as you can build and so we solve the the down and out problem and came back with them and said this is now later on when I wrote my paper on option pricing laundry list longer thing I was trying to illustrate how you could use the technology for a lot of different things so I loved one of the things I put in there was the down-and-out option I turned out later on there's a whole industry called exotic options that's evolved of which the down-and-out is it sort of being a prototype I never would have even been aware of its existence let alone something been able to solve the problem and write it up if I hadn't been involved in practice if you imagine you have an asset that's not traded and you say alright can I build from traded securities the best asset the one that comes closest to producing the same returns of successor ok so you're saying it's called a tracking portfolio technical but just say could I build a portfolio what's the best portfolio I could build out of traded securities to replicate or comes close to replicating that's if you could perfectly replicate it then you're done but and what's best mean well you could think of managers having minimum tracking error minimum residual variation between the portfolio and the target if you do that and you used all traded securities to do that then what's going to be the outcome of that tracking error it's the same thing you find from regression theory that the residual or the tracking error between the traded and the non trading thing has to be on Karl it with all traded assets in any real world economy I mean well develop a national economy if I have an instrument whose returns are uncorrelated with every traded asset I mean just think about that from them that's pretty hideous socratic risk that's it's probably hard to imagine that being priced that that has an expected return of other than the risk-free rate and there's called diversifiable whatever you want to do it but just thinking about it and if you buy that then you can close them off because you can say I will the best tracking portfolio I will have residual there's no longer be a exact replication therefore you have some error but that error is uncorrelated with everything so in a sense it's going back to to the black-scholes version of the CAPM where they took out the systemic risk but you're not taking out everything you're taking mine you know you're just taking everything you can get with a trade it's security so it's even more than systemic risk factors it's it's everything out there so it seems to me that that allows you to price it and what you would now do in today's world is package up that residual risk and find places to place that risk where people will pay the highest price for it the require the lowest return and those would typically be insurance entities or entities that are used to dealing with actuarial risks that are very like this or some kinds of specialized hedge funds they're always looking for on correlated risks what the point of the matter is that it seems to me that the idea the fact that it isn't traded somehow makes the model not particularly valid I think is my judgment is is this is far too harsh is it seems to have kept its legs in other words it's it's 30 years later it's still able to address at richer set of problems you know if you if you raise the ante on what you're they're looking for in terms of a modelling it allows you to do it it's more complex more computationally complex man I have all the data but it has answers and I think that's been a feature of it that is pretty important quarter 6:00 in the morning the phone rang and my first reaction was to say okay if it's if it's anything to leave a message when I get to New York oh I'll pick it up but then I said oh no I'll pick it up and just say sorry I'm running to a plane call me later well I picked it up and when they identified asked me you know there was me then they said we have some interesting news for you and needless to say I never got out to the airport because they it was the head of the the Nobel Foundation who actually delivered the news and so forth at the end of all that formality he then put the chairman of the economics selection committee on who is somebody that knew to verify that this wasn't a hoax so I even changed put it on a proper suit and and went down and there was the world very nice it takes a little while to sink in and just to be considered is as a great honor and the idea of actually winning it it's just nothing to compare it Paul Samuelson said to me and on the day when I received the news he said you know they're not only going to think that you know everything answer to the economics problems but it doesn't stop there they ask you medical everything so there's there's I think they sometimes confused the prize with this a Leonardo da Vinci Renaissance man or woman prize but it's really for a very specific thing and in my case it was for evaluation models for derivatives the world is a very complex place and we need all the good tools we can and what I think the theory that I written down which is I think at the end of the day the neoclassical model will explain asset prices and risk distributions not everyday not in the short run but but you know sort of unconditionally I think that do a good job common sense tells you that it would be a remarkable result that the behavior of asset prices in a world where all choices of stock investments are done directly by individuals versus a world where all stocks are selected by institutional managers operating in groups and and that individuals are only indirectly investing because the behavior of individuals it's different than group or institutional behavior not because the group or institutional behavior also isn't flawed but the idea that there's an invariance principle that if you organize yourself as institutions that the behavioral aberrations will be identical then as you had individuals behaving it's just yeah it's possible I suppose but that would be a great theorem and a theorem I don't think that can be proved if you accept that then the structure the way people are doing tests by mechanically trying to mirror the test that we're done with in a classical model are flawed but that's not a reason to throw it away it's rather to say I think precisely because the institutions matter that if these behavioral phenomenon let's posit that they exist and that they're important that I think there'll be important telling us why we have the institutions we do not why asset prices are that way a lot of people like to play and if they want to play on the stock market on the internet that's their decision but for their serious decision the retirement decisions savings decisions tuition so I think that most people don't want to do that the idea of people doing all this on computers on the cell that they that institutional solutions in which they essentially as you do with your medical and so many other things that the best thing is to invest you get good advice for many interesting good institutional integrated solutions look for solutions not rates of return you're going to see the the creation of integrated solutions which are all sold on the basis of not things like rates of return but on performance of they do things you pay this amount of money your tuition for your children's taking care of in 18 years forget it you pay this amount of money to save this amount for this number of years you will receive a standard enough money in retirement to sustain your stand living that you enjoyed in the latter part of your work life that's what you care about in the classic Markowitz you have the risk and return and your preference function tells where and you don't even have to know your preference on so you can kind of do introspection and say well I can live with this amount of risk for this return versus this point but how much do you weight the say you shift your portfolio towards let's say longer-term bonds to hedge yourself against anticipate changing interest rates or ahead yourself against the shifts in the risk return slope that's less intuitive for those who have technological advantages in the business I'll tell my producers I think that's complexity is a good news because it's a barrier to entry it's a way of earning the building a franchise so that you should embrace complexity not view it as as something negative and finally to the people who are thinking of France as a career I think it's it's just it's it's a wonderful exciting opportunities really believe there's a chance to do well financially by doing good which is to do things that matter and have exciting problems and while we have Wars and depressions and recessions that slow things down you look at from a longer horizon it's absolutely clear that all the developing nations and if you included as China's emerging work have to redo their financial systems it's also cleared the second largest economy by some measure of the world Japan the very old economy has to completely redo its system it's also clear that while we have monetary union in Europe we don't have institutional Union so that's all got to be redone so you in terms of total need and this isn't going to happen in two years or five so the needs out there now the next thing it's exciting is that's just something you can do about it you've got all this technology I talking about swaps derivatives all these contracts SPD's all these models everything that they they're all market proven there's two hundred and seventy trillion notional amount of derivatives in the world so the idea are we going to have to excuse me no longer a question it's a reality it's a little bit like when we phone with a phone system if you were going to build a phone system today as people do in countries that need them you wouldn't look at the United States and say well that's a great phone system let's put some telephone poles up why because it was a great phone system when it was built it isn't worth tearing it down but if you were building a new one you wouldn't build it that way you build it as a digital system and that's the way it's done just so in finance if you've got to rebuild a financial system you don't look in the rearview mirror of how it was done even best practice but rather say how would you do it using existing technology so again repeat I believe long term this huge need second there's huge technology to do something to meet that need it doesn't get much better than that you

Channel: afajof

Views: 16,877

Rating: 4.9338841 out of 5

Keywords: American Finance Association, AFA, American, Finance, Association, Journal of Finance, Journal

Id: cmzON43dWA4

Length: 24min 55sec (1495 seconds)

Published: Sun Sep 15 2013