Let’s start by saying that this blog was NOT paid by the publisher or the author of this book to make this review. It was a mere consequence of my search engine knowing more or less what the writer is in the mood of reading at the right time of the week. And for the same reasons now there is one too many mason jars on my kitchen counter, but that is a completely different story.
The first time the I heard about Mandelbrot, I was a teenage girl in the process of finding her identity. So instead of doing this, this or this, the “thing” to do was having desktop backgrounds that no one else had. And the answer was a program that generated crude graphics using different types of fractals among which was the Mandelbrot set. Many hours and desktop backgrounds later it all went into the “stuff to remember for later” part of my long term memory, and 20 years later while attending a course on evolutionary computing things started clicking together. And 5 years after that course, my search engine shows me this book. “Mandelbrot eh? like the power-law and the terrains for games? what is that guy doing with markets?”. Unfortunately, he is dead. Since 2010. But the book was a good read.
A main goal of the book is to show you another way of thinking about randomness in finance. Especially, about the assumptions behind most financial methods. In fact it does show you another way of thinking about randomness in everything, but that is part of the charm. And the nicest thing about this book is HOW it takes you to that other point of view.
The book is divided in three parts, titled (respectively) “The Old Way”, “The New Way” and “The Way Ahead”. Part 1 and Part 2, the writer pulls you into the biographical and social context of the people behind most significant pillars of modern finance. The historical recaps felt very vivid. It was very easy to relate yourself to the people making those discoveries. So you could understand why things were made the way they were currently. And then, thru the whole book, you had that feeling in the stomach that something was about to happen. You get glances of the place where the book wants to take you, like mirages in the desert. I particularly enjoyed a metaphor of his in which the distribution of the location of the lost arrows of a blindfolded archer was used to compare the mild randomness that could be expressed with a Gaussian distribution vs the wild randomness that could be expressed with Cauchy.
Part 3 has only two chapters, and they can be read without the previous parts: “Ten Heresies of Finance”, in which he summarizes many points that have been scattered as breadcrumbs thru the previous sections, and “In the lab”, in which he presents the work of other people with similar observations as his.
In my opinion, the corner-stone message of this book is that markets behave like turbulent processes with bursts, pauses and fractally scaled parts, rather than gaussians, and critical events tend to cluster other small events around them and cause turbulence. Looking at markets as turbulent phenomena has interesting consequences. For instance, big gains and losses concentrate into smaller time slots. The biggest fortunes are made and lost with price variations right before and during such critical events. With this in mind, arbitraging becomes a more significant driving force in the market, so price differences become more interesting than average prices themselves. He also looked at the fact that prices change in leaps, rather than in smooth glides. So timing becomes quite important.
But the former observations also hint at the fact that risk has been underestimated in markets. Since market behaviors have busts and pauses that makes them vary more wildly than gaussians, working with “averages” alone in stock markets is indeed risky and inappropriate. It is more meaningful to work with out-of-the-average values when estimating risk.
And it does not stop there. Another strong statement on this book is that markets everywhere work alike, but endogenous and exogenous factors can make a difference. Bubbles appear as a consequence of interactions and turbulence, and patterns can change from a moment to the other. Markets can exhibit dependencies without correlation. For instance, the fact that the prices went down yesterday does not mean that they will fall today. But we could have the case in which having our prices plummet by “x” percent today will increase the odds of another “x” percent move later. So, we could have a strong dependency, without a correlation. Large changes tend to be followed by more large changes. Volatility clusters. Yet spurious patters appear everywhere, and that is just another consequence of turbulence and wild randomness. Humans tend to look for patterns and may find them even if they are not there.
For reference, the complete title of the book is “The Misbehaviour of Markets: A Fractal view of Risk, Ruin and Reward” by Benoit Mandelbrot and Richard L. Hudson. It has won a Financial Times award for the most innovative book in business and finance published worldwide.
I will now leave you with something to look at. Embrace the turbulence, and see you next post!
The featured image was taken from here.
statsletters.com 