Everyone who is interested in derivatives pricing and risk management should read this article. Professor John Hull's book Options, futures and other derivative is the greatest in Finance.  Read the article you will know how THE BOOK is evolving over the time and becomes the favorite derivatives book among college and practitioners.
---------------------------------------------


Ask any derivatives professional where they first learned about the subject and there’s a good chance they will tell you John Hull’s celebrated textbook, Options, futures and other derivatives.

Heading for its eighth edition early this year, the book has introduced a generation of traders, quants and investors to the pricing and hedging of derivatives. First published in 1987, it has sold several hundred thousand copies worldwide and is virtually unique in selling strongly in both the college and practitioner markets.


For Hull, the Maple professor of derivatives and risk management at the University of Toronto, the success was slow in coming. He remembers the first edition as a slim 300-pager in fairly large type with only 13 chapters – the eighth will have 35. This constant updating and development of the book has been key to its popularity, he thinks – sales started to pick up once he began reacting to demand for particular material.


“It wasn’t an immediate success, but over time as the derivatives market expanded, the book expanded with it and sold more as a result. It’s not like writing a book on something like differential calculus, where the subject matter is static. It’s been quite a lot of work, but I do enjoy updating it – it keeps me up to date,” he says.
The level of influence it now holds surprises even Hull. “I sometimes get calls from banks saying they would like me to mention a certain model in my book because it will be easier to get internal approval for it,” he says.


He keeps copies of each edition in his office, highlighted and covered in post-it notes and addenda to help him respond to the roughly 15 email queries a day he receives. Some point out typos or errors, while others ask for help in understanding a particular point. “These are especially helpful because they point to places where maybe I need to improve the presentation of something. The focus is always pedagogical.”


For Hull, presentation is vitally important. Mathematics is only used when necessary and notation is kept to a minimum. “The temptation is to put subscripts all over the place and list all the dependencies of a function, but I try not to do that because it makes it harder to read,” he says.


The last edition was printed in 2007 – before the failure of Lehman Brothers rocked financial markets. As a result, the latest version is hugely expanded, with a new chapter focused on the crisis. One section examines the performance of value-at-risk tools during the period. Using real-world data, Hull shows that a bank using the industry-standard 500-day historical calculation method for VAR on a typical international equity portfolio at the beginning of October 2008 would not have had enough capital for the coming turmoil. A weighting of the data by market volatility led to a much healthier figure, he found.


Hull has previous experience with data collection and its use in business decisions. After studying mathematics at Cambridge and operational research at Lancaster University, his first job was as an executive for Leicester-based British Shoe Corporation. One area of responsibility was to determine inventory using the rudimentary computing technology of the time.


Each pair of shoes had a computer punch card inside, which was filed by the clerk upon sale. These cards made their way back to Hull at central headquarters, who determined how to allocate stock based on the resulting program. “If you dropped the cards it was game over, of course,” he says.


A stint studying operational research followed – first at the London Business School, and then Cranfield University, where he earned a PhD in 1976 while lecturing on business and finance. But it was a move to Canada in 1981 that was the catalyst for a jump into derivatives. As a professor at York University and then at the University of Toronto, both in Ontario, he has tackled several subject areas.


Although most famous now for the interest rate model he developed with his University of Toronto colleague Alan White, Hull started working in general derivatives pricing, with the aim of improving the Black-Scholes-Merton model. He focused first on expanding the scope of that work from equity options to foreign exchange.


“What Robert Merton’s 1973 paper tells you is how to price and hedge an option on an asset that pays a continuous dividend yield. So the idea was to look at a currency as such an asset, with the local interest rate as the dividend,” Hull explains.


While presenting this work to a group of Royal Bank of Canada traders in the early 1980s, Hull was challenged over the success of a delta-hedging simulation. “One banker, I think he was part of their treasury operation, put up his hand and said ‘it doesn’t work that way. You get good results here because you assume a constant volatility, but it actually moves around.’ And he was right.”


This led to an interest in stochastic volatility models through much of the decade, and the development of a research relationship with White that lasts to this day. Their paper, The effect of a stochastic variance on option pricing, introduced one of the first stochastic volatility models to be successfully implemented in the industry.


At the end of the decade, the duo shifted to fixed income after reading a 1986 paper by Thomas Ho and Sang-Bin Lee, called Term structure movements and pricing interest rate contingent claims. That article was the first to really take seriously the idea of calibrating a model to an entire yield curve and then determining the possible arbitrage-free future dynamics. Hull was intrigued by the idea, which was to dominate his research over the next few years.


The main fruit of the labour was the Hull-White interest rate model, a generalisation of Ho-Lee that contained an additional time-dependent mean-reversion term in its governing stochastic differential equation – a so-called Ornstein-Uhlenbeck process. The additional term was motivated more by a desire to improve Ho-Lee’s calibration performance, rather than a belief that rates move back to an average level – the additional parameter allowed a greater number of curve shapes to fit.


A downside of the model – like Ho and Lee’s – is that as a Gaussian-distributed model, it has a positive probability of predicting negative rates. During the high interest rates of the 1980s and early 1990s, this probability was small – but in today’s low rate environment, it’s likely to be close to 50% when calibrated to the near-zero rates offered by some central banks.


“It’s a drawback, but for years it didn’t matter so much,” says Hull. “It’s really because of the very low-rate environment that it’s become more of an issue. It’s the model’s tractability and the fact it calibrates fairly well that makes it popular.”


The model gives a fairly simple closed-form expression for the short rate and a slightly more complicated bond price formula. Furthermore, long-term average rates are easily seen to converge to the ratio of the mean reversion parameter to the speed of reversion. These simple, functional properties mean Monte Carlo simulation is not required in many situations, and much of the pricing can be done with another of Hull’s favourite tools – the binomial tree.