Price: $150 ($75 for students and academics).
Having been introduced to my craft in an illegally photocopied manuscript of Option Markets, written by Mark Rubinstein and John Cox, I was thrilled to get my hands on Rubinstein’s new book. The previous one gave the first generation of quantitative options traders a starting point, a matter for which we owe the authors boundless gratitude.
Rubinstein was among the pioneers of the first paradigm of option theory. His book propounded such unfamiliar concepts as "gamma strategies" at the time when the other major book read in dealing rooms was Lawrence McMillan’s Options as a Strategic Investment. Option Markets was the first work that popularized probabilistic and scientific methods in options, helping inaugurate the derivatives revolution – but not without some pseudoscience to go with the science. It was later undeservedly supplanted by Hull’s Option Futures and Other Derivatives, a volume far less insightful but more structured and pedagogical.
In addition, I have other specific reasons to owe Mark Rubinstein a measure of gratitude, having been long volatility during the stock market crash of 1987 and, through thick and thin, ever since. Most people believe the Leland O’Brien Rubinstein portfolio insurance program played a dominant and catalytic role in the crash . The crash was the event that taught me to disrespect anyone who claims an ability to "measure" risks.
It would not be exact to call these lines a book review since this is not quite a book, but rather a set of multi-modality package class notes, called Derivatives: A PowerPlus Picture Book. The package richly integrates 382 pages of spiral-bound text with 342 multimedia PowerPoint slides, an extensive WinHelp glossary, several additional documents, abundant software (the Rubinstein Option Calculator), and worked examples. Purchasers have the right to download future versions and additional material from Rubinstein’s web site.
I start this review with a comment on the book’s medium, given its originality. It is truly innovative, which makes it difficult to judge with full fairness; since all this depends on tastes and habits. I tried the three different modalities-- the PowerPoint slides, the WinHelp glossary and the spiral-bound book – and comfortably settled on the book (by power of habit). I would have preferred to have the book portion in conventional form rather than spiral-bound given that I used it as the dominant part of the package. Although I occasionally teach derivatives (albeit to option professionals and motivated financial engineering students rather than MBAs), I found no use for the PowerPoint slides –although other instructors might have a different experience and would save precious time by integrating the enclosed slides into their own presentations. As to the software, I found it quite extensive and pedagogical.
I toyed with this package during trading hours, with one eye on my trading screen and the other on the picture book with the WinHelp tool. I found it to be quite an attractive dictionary-style reference. –My judgment is therefore similar to the one I have about on-line encyclopedias: they may be useful to many, but some people may prefer the feel of plain old traditional hardcover volumes. One additional feature of the book: it can be entertaining, with a variety of sounds popping up unexpectedly, and animated figures erupting on the screen (I suspect them to be caricatures of Mark Rubinstein). Derivatives even includes a sampling of Mark Rubinstein’s musical tastes --Grieg and Mozart.
One of the benefits of such a medium is that the buyer is not merely purchasing one edition, but all the subsequent ones. This allows the owner to end up with a typo-free version; the author can also instantly (and un-academically) change his mind on a subject (I have a few suggestions for him later in this piece.). Rubinstein, unhappy with his previous publishing experiences (many technical authors attest to similar disgruntlement), decided to take matters into his own hands. Who needs distribution when the Internet is right there, one upload away? One worry I would have about this publishing model is piracy, since some of the files, namely, the slides, can be easily copied.
The book sections are organized in 8 parts: 1) Introduction, 2) Forwards and Futures, 3) Introduction to Options, 4) Binomial Trees (i.e. discrete time), 5) Black Scholes (i.e., in continuous time), 6) Volatility (realized and implied), 7) Dynamic Strategies, 8) Annotated Bibliography.
I will next discuss the positives. First, Rubinstein, has characteristically invested a great deal of attention in the details. The book is tight, with definitions and descriptions as clear as one would expect from someone who taught derivatives to an entire generation of students. Second, the book is truly pedagogical, even engaging in places. Flipping through the slides provided me with some measure of entertainment, and even made me laugh a few times. I enjoyed the slides in which a binomial tree unfolds forward with a whizzing sound to show the evolution of an asset price, then rolls backward as the call is priced, with a finale greeted with loud applause.
Third, Rubinstein has his feet on earth. Derivatives is rich in details about factual trading matters, something rarely found in textbooks. For instance, there is a slide showing how to cheat one’s customers (illegally), with a list of the major old tricks, such as cherry picking (where brokers allocate the winning trades to themselves or, say, someone like Hillary Clinton), chumming (where market makers trade against each other to create the illusion of activity), and ghosting (where market makers collude to push a price in a given direction), something I have not seen in options books before. Note that Rubinstein’s options culture is equities-oriented; he has the language of the less sophisticated CBOE practitioners rather than that of the more scientific financial product professional.
Fourth, the book is, unexpectedly, not too immersed in the financial theory, which generally glorifies Black-Scholes-Merton and the overrated continuous finance approach. Rubinstein presents in the annotated bibliography a history of derivatives thought which gives some well deserved justice to the various forgotten contributors to the option pricing formulas. To me the Black-Scholes-Merton result is not a "breakthrough" invention a la Pasteur, but an apt economic argument that allows option pricing within neoclassical economics. The true inventor was Bachelier. I may start calling it the Black-Scholes-Merton improvement of the Bachelier-Kolmogorov-Boness formula. Concepts like complete and incomplete markets and the first, second, and third theorems of financial economics have little applied relevance and do not have the odor of sanctity outside of the ivory tower; Rubinstein keeps them somewhat in the background. Still, as a scientific trader, I would have preferred to have Rubinstein treat the subject of options outside of the religious backdrop of the heavily theoretical neoclassical economics.
Fifth, Rubinstein provides a clear discussion of the assumptions behind financial theory, such as the drawbacks of using mean-variance as a criterion for efficiency in the presence of skewness. He presents a clear presentation of the attributes of distributions and shows how skewness and kurtosis can translate into shapes on the option volatility curve.
To conclude the positive points, Rubinstein leaves me with the impression that conventional option theory is an object that he understands intimately, for he discusses it with the simplicity and directness of a master. I have to bemoan my experiences with academics, traders and former academic-quants on the subject, which are often marred by theorem-dropping nonsense and complicated economic arguments which reveal no true comprehension of the generator behind the concept. I have discussed options and derivatives with hundreds of successful derivatives people throughout my 14-year career, and have only felt this effortless mastery with a handful of old veterans such as Marty O’Connell (with options in general), Bruno Dupire (with option mathematics), Howie Savery (with barrier and digital options), and Varun Gosain (with credit products). Rubinstein, as we say on Wall Street about someone who knows options, can connect the dots--a very rare attribute in academia. However, I bemoan that he does not seem as familiar with market dynamics as I would have wished. I wish Rubinstein had a little more scientific realism in his approach.
On the negative side, there are a few minor points. These are merely annoying and, at times, infuriating to a practitioner of the no-nonsense variety. The most irritating one is as follows. There is a tendency by some finance academics to consider that any knowledge not published by their designated journals does not exist. While such a practice can be justified in some disciplines --say, complex analysis or mathematical anthropology-- it is less tolerable in financial derivatives, where practitioners of the non-tenure seeking variety are often years ahead of academia. Not only do they deal with products and techniques of a higher relevance and sophistication than the average derivatives academic, but they are subjected to greater scrutiny from the market. It is high time for academics to understand that there is a difference between the "practitioner" who reads Tom DeMark’s unrigorous and non-testable discourses on technical analysis and the columns of Futures Magazine on one hand, and the financial product professional subjected to all manner of scientific rigor on the other.
Rubinstein provides an illustration of the practice of finance professors to stick their names, and those of their colleagues, on a collection of results that are entirely trivial in nature and well-understood in practice, but that nobody bothered to publish before because of their triviality. For example, he writes, "Mark Garman, while he was an active professor at Berkeley, developed a way [c. 1992] to use binomial trees to compute the risk-neutral expected life of an American option. As in inventor of this concept [emphasis mine], he had every right to name it and called it fugit." Well, I recall that this "invention" was similar to something option professionals commonly used in the 1980s, when American options were still frequently traded in the currencies –the practice was so widespread that we used it as an interview question. Our method consists in using the ratio of the interest rate sensitivity of an American option priced on a binomial tree to that of the same option of the same strike price on a Black-Scholes model and multiplying it by the nominal maturity. So Garman’s invention was merely the publication of a complicated version of the trick. Incidentally Garman had his name stuck (no doubt against his will, for he is a first-rate scholar) on an option formula (the "Garman-Kohlhagen formula") simply by replacing the dividend in the Black-Scholes formula with the foreign interest rate (a brief footnote).
The attributions can be excessive in this book, no doubt a product of the financial economics culture. Listing financial options, names are invariably stuck to the discussions, even if the results are beneath trivial. Rubinstein is guilty of overquoting his financial colleagues, referring to papers that are rarely true contributions to merit quoting, or are often too complicated to be appropriate auxiliary reading list in an introductory book. He drops too many names between slides 38 and 53. I give a few: automobile leases (Miller), taxation of income (Draaisma/Gordon), liquidity (Beck), fishing industry entry licenses (Karpoff), 401(k) deferred-tax saving plans (Stanton), bull and bear floating rate notes (Smith). Nowhere did I find in the book the slightest hint of the existence of a financial products professional capable of learning anything on his own, in a clinical manner, without help from academic papers or precious advice from their local derivative academic. While minor academics may need to glorify each other, a practice the Romans call asinus asinum fricat (donkeys rubbing each other), I find it strange to see Rubinstein engaging in such practice, given his clout.
The last topic on my list is portfolio insurance. Rubinstein provides in part 7 an option theoretic discussion of dynamic strategies, which makes this book different from other option books. I enjoyed his presentation of the distinction between convex and concave dynamic strategies, where the payoff from the stream of purchases and sales in the underlying security can resemble a positive or negative gamma position. In all fairness, he considers some of the weaknesses of portfolio insurance, which I find commendable –but he does not go all the way. He seems to underestimate the evils of execution. Any trader who has been short gamma would know that he is the weak hand, and that a market is like a large movie theater with a narrow door –which explains such underestimation. Although I find it unfair to bring up in a discussion elements foreign to this book, namely, his record of advocacy of portfolio insurance, I have to say that Rubinstein failed to learn from the crash of 1987 the difference between the strong hand and the weak one. A weak hand is someone who is obligated to act in a certain way in response to a market move, like, say, sell after a decline. Such person can be highjacked by the market. "Show me a reluctant dynamic hedger and they will front run him until bankruptcy", I wrote in DS (December-January 1997). Why didn’t he learn that? Because, at heart, he is not a trader. The idea that comes out of this is that one does not have to be as intelligent and insightful as Rubinstein to become a good trader; it suffices to have enough intelligence to learn to operate a telephone, be extremely self-critical, have the ability to learn from one’s losses, and, above all, observe markets with a clinical, as opposed to theoretical, eye.
If dynamic portfolio protection does not work, it flows that the opposite may. In other words, the method may work if one is not obligated to sell dips, but has the option of doing so. Some aspects of dynamic strategies have been extremely successful, particularly when being used by the aggressor rather than the sitting duck—namely inverse portfolio insurance. This is the sort of things legendary traders Paul Tudor Jones and Bruce Kovner might have been doing as teenagers; the concept bears the name of cut your losses and add to your winners or play with the house’s money, not your own. The method has been practiced ever since markets were invented; its math was popularized in Feller’s famous textbook under the concept of gamblers’ ruin.
Another evil of short gamma strategies Rubinstein avoids discussing is the measurement of rare events. Rubinstein provides some testing of portfolio insurance to illustrate its performance and shortcomings. However, he ignores the effect of the 1987 crash on portfolio insurance, although he presents the simulations results for 1982 when "it failed miserably". Nor does Rubinstein present the potent economic arguments against portfolio insurance by Sanford Grossman, although the book boasts being about something called economic intuition rather than the math.
This discussion leaves me with a thought: It is unfortunate that studying the properties of past data does not provide much insights about the incidence of the gaps and crashes: it is tantamount to studying the properties of the past data in Mark Rubinstein’s life and, not seeing him die even once, accepting the null hypothesis of his being immortal with a high confidence level.
To conclude, I am excited to have Derivatives on my hard drive and bookshelf. Overall, this remains by far the most pedagogical and most colorful introductory academic book on options ever written. I am happy to see that another generation of students will be trained by Mark Rubinstein. However, I must add that this is not quite a book for practitioners.