Against the Gods: The Remarkable Story of Risk

Rating:
Summary: There's no risk in buying this book
Review: Against the Gods: the Remarkable Story of Risk by Peter Bernstein It is so good that I cannot wait to tell everyone about it. This is the story of the study of probability and statistics from early days of dice and cards (gambling) to its present day of insurance and derivatives (gambling). In this story are the lives of a great many gambling addicts who helped make the modern world possible. It reminds me of the old PBS series Connections, where history took so many strange turns to produce modern technology. I heartily recommend it.

Rating:
Summary: Fun with statistics
Review: I happen to enjoy both history and math, so this book about the history of statistics was right up my alley. That sentence alone probably dooms me to perpetual nerd-dom, but that's ok because Bernstein has done a strong job of taking a subject that could be dull and really bringing it to life and infusing it with color.

If the topic of the history and evolution of statistics sounds interesting to you than I believe you've found a winner.

Rating:
Summary: Where Was This Book When I Took Statistics 101?
Review: This book does much to highlight, underscore and re-affirm the important and pivotal role of critical thinking in the measurement, analysis and interpretation of risk. Mr. Bernstein has put together a very readable text that simultaneously is part history, part tutorial, and part meditation on diverse and often inter-related topics of probability, descriptive and inferential statistics, economics and finance. He skillfully shows how a blind reliance on numbers and quantitation, especially as they apply to business, economics and finance can equally inform or mislead, and goes further to demonstrate that the tug of war between gut (our best guesses) and measurement has not been resolved with the introduction of theoretical treatments of individuals and groups, complex mathematical formulae, innovative management products and computers; it has only been taken to a new and higher level of complexity.

Mr. Bernstein's thesis was a simple one: people had to first make an intellectual leap- they had to believe that outcomes could be influenced by their actions and not merely in the 'hands of the Gods' a matter of fickle 'Fate', before they could even begin to think about taming risk, let alone measuring it. From the author's standpoint, risk management demands the uncommon ability to think with numbers. It also requires the ability to live with varying degrees of uncertainty. While people throughout history had the ability to do the latter, they had not, as a group, the ability to do the former until very recently in recorded human history.

Bernstein takes this as a jumping-off point for his exposition. He first details the origin of numbers, something so banal to us today. He then takes the reader to the gaming tables of Europe, and introduces us to the figures who would lay down the tenets of Standard Mathematical Probability. From there, Bernstein introduces us to various historical contributors to the allied fields of probability and statistics- from Fermat and Pascal through a host of Bernoullis (the archetypal Renaissance men) to such enigmatic figures as de Moivre and Gauss, pious men like Bayes, and right up to privileged and misguided social reformers like Galton and Quetelet. Each in turn contributed a crucial piece of the portraits comprising Standard Mathematical Probability and modern day statistics such as the concepts of mathematical expectation, conditional probabilities, sampling and measurement error, the law of large numbers and regression to the mean. Once these elements were in place, the stage was set for the formal development of risk management as a science and a practice, and at this point, Mr. Bernstein introduces us to the economists, who then as today couldn't agree on anything, could always be counted on to provide 'correct' answers for what has happened, and the wrong answer as to what will happen.

After setting the stage by presenting the tools of modern risk management, Mr. Bernstein then turns his attention in the last third of the book to those who have righfully challenged the assumptions underpinning neo-classical economics, modern finance, and Standard Mathematical Probability. This is perhaps the most interesting part of the book, because it is here that the author introduces the human factor in risk, and calls into question the use of mathematical techniques, largely developed over the course of studying a variety of regularly occuring and repeating natural phenomena, to the actions and behavior of that mass of capricious and unpredictable creatures, The Human Ape. His take on this subject has some very interesting implications in the world of finance, given that the Human Ape is, to paraphrase, smarter than the average monkey, and has the capacity to learn from his or her mistakes (which most unfortunately do not), and as a consequence, adapt to changing conditions.

I found the last half of the book to be the most interesting, because here is where the author fully lays out the limitations of the disparate mathematical techniques, especially as they apply to the Great Upright Walking Ape. Throughout the book, all probabilistic and statistical concepts are explained in plain English, and there are absolutely no equations in the text (a definite plus, as this often gets in the way of critical thinking and true understanding). Mr. Bernstein took great care to fully explain the key concepts in precise detail, and offers the reader many accessible sources for further enlightenment throughout the text.

Perhaps the most important messages of the book can be briefly summarized in the following way. We define risk as the chance of loss, uncertainty as something we don't know, and in this day and age, we often behave as if we know all of the risks and there is nothing that can happen that has not happened before. For us, the past is the best indication of the future course of events. Standard Mathematical Probability was developed under the assumptions of total information, independence of outcomes (past and present moves do not affect future moves), and the ability to reduce everything that matters (and many that don't) to a number. In essence, it looks at the same thing done in the same way repeatedly, and as such, takes the characteristics of any situation to be stable and immutable.

Standard Mathematical Probability forms the core foundation of risk management, and while it is a fine tool for the gambling tables, it falls apart when applied to the real world, where for a variety of reasons the assumptions are simply untenable. In fact, in the real world, they are ridiculous, bordering on insane. And this is before we even begin to contemplate the human factor....

In sum, while I do not agree with Mr. Bernstein's assessment of the failure to develop the tools of Standard Mathematical Probability before the time of the Renaissance (which I found puzzling, given the ample evidence of very sophisticated mathematical capability among such diverse ancient societies as the Aztecs, Egyptians, Olmec, Maya and Inca), I wholeheartedly agree with his assessment of these tools as they apply to the human element. I thank Mr. Bernstein for making clear all of those concepts that too many Professors in Statistics 101 (intentionally) obfuscate, and I highly recommend this book, as well as Benjamin Graham's The Intelligent Investor, to anyone wishing to invest with both safety and success.










Rating:
Summary: Keynes is severely misinterpreted;Mandelbrot never appears
Review: Bernstein does a good job overall in discussing the history of decision theory,its interaction with developments in probability and statistics,and the resulting use in portfolio-risk management of the mean-variance-standard deviation model,the capital asset pricing model(CAPM),and the pricing of options(puts and calls)by use of the Black-Scholes model.All three of these models are totally dependent,in terms of the accuracy and reliability of their forecasts, on the underlying probability distribution being normally distributed.Benoit Mandelbrot has,over a fifty year period,presented a massive amount of empirical evidence that calls into question the reliance of the above mentioned models on the normal probability distribution.Bernstein commits a shocking error of both ommission and commission in failing to discuss or even mention Mandelbrot or his work a single time.This ommission costs Bernstein one star.Bernstein loses a second star for his highly inaccurate assessment of the work of John Maynard Keynes in his A Treatise on Probability(1921;TP).Bernstein claims that"Unlike Knight,Keynes does not distinguish categorically between risk and uncertainty;in less precise fashion,he contrasts what is definable from what is undefinable when we contemplate the future".(Bernstein,1996,225).Bernstein appears not to have read beyond chapter 3 of the TP.In chapter 26 of the TP, after having laid the groundwork in chapter 6,Keynes defines uncertainty in relation to his index w(0<=w<=1), which measures the completeness of the potential relevant available evidence upon which a decision maker is going to base his estimates of probabilities.These probability estimates are primarily interval estimates for Keynes.Bernstein fails to inform the reader of this important fact.Keynes identifies the case of ignorance(complete or total uncertainty)with a w equal to 0.In this case no probability of any kind can be estimated.Keynes identifies a w equal to 1 as specifying a complete information set.If w equals 1 a single unique probability distribution can be used to deal with risk.Keynes deals with the case of partial knowledge(incomplete evidence)by specifying a value for w that lies somewhere between 0 and 1.Keynes was not scornful of the law of large numbers,as claimed by Bernstein(p.226).Keynes regarded the frequency interpretation of probability as a valid and sound theory with limited applicability.It was not general because it required that all the relevant evidence had to be composed of statistical evidence alone.Keynes did not heap scorn on the arithmetical mean,as claimed by Bernstein(p.226).Again,there are other measures of central tendency besides the mean,such as the mode,median,and geometrical mean.What Keynes did object to was the use of the arithmetical mean and the normal probability distribution without any empirical support.Yes,Keynes would object to the overuse and misuse of the normal probability distribution in financial and portfolio analysis precisely because of the vast amount of empirical evidence arrayed against its use.One need only read any book or paper written by Mandelbrot,as well as the work of an ever increasing number of researchers who have followed in his path.Finally,nowhere in the TP did Keynes declare that"Perception of probability,weight,and risk are all highly dependent on judgment" and"the basis of our degrees of belief is part of our human outfit".(Bernstein,p.227).Bernstein claims that Keynes made these statements on page 5 of the TP(See his footnote 29 to chapter 13 of his book).This is simply incorrect.There are a number of other errors relating to Keynes's 1937 Quarterly Journal of Economics(QJE) article and the General Theory(1936) with respect to Keynes's use of the word uncertainty.The quotation taken from the QJE article and cited on page 229 by Bernstein specifies that uncertainty is a range.This range goes from little or no uncertainty to mild or moderate uncertainty to complete or total uncertainty("We simply do not know!")Perhaps Bernstein could correct the above mentioned errors and ommissions in a second edition.

Rating:
Summary: Trying to Tame the Untamable: RISK
Review: +++++

This book, by Peter Bernstein, author and president of an economic consultant firm, is about those thinkers whose vision showed how to put the future at the service of the present. What exactly was this vision or idea? Answer: The calculation and subsequent quantification of risk. That is, they attempted to manage risk. "The capacity to manage risk, and with it the appetite to take risk and make forward-looking choices, are key elements...that drive the economic system forward."

"[T]he stock market and the bond market are natural laboratories for the study of risk because they lend themselves so readily to quantification." Thus there will be some discussion of stocks and bonds.

This easy-to-read book has five parts--

Part 1, two chapters (up to the year 1200):

Tells us how the modern conception of risk is rooted in the Hindu-Arabic numbering system that reached the West seven to eight hundred years ago.

Part 2, three chapters (between the years 1200 and 1700):

Describes how a serious study of risk began during the Renaissance period, "when people broke loose from the constraints of the past and subjected long-held beliefs to open challenge." By the mid-1600s, the "theory of probability" was discovered, a theory which is "the mathematical heart of the concept of risk."

Part 3, six chapters (between the years 1700 and 1900):

By this time, mathematicians transformed probability theory from a "gambler's toy" into a major instrument for organizing, interpreting, and applying information. Quantitative techniques of risk management gradually emerged that "have helped trigger the tempo of modern times." During this time you get the development of such things as:

Life expectancy tables; life annuities; the Law of Large Numbers; statistical sampling; the normal or "bell" curve; standard deviation; the Law of Averages; the groundwork for modern principles of investment management; and regression to the mean.

Part 4, four chapters (between the years 1900 and 1960):

At this point "uncertainty, and its handmaiden luck" are discussed. The concepts encountered in this part never occurred to the mathematicians and philosophers of the past since they were too busy establishing the laws of probability to "tackle the mysteries of uncertainty." Included in this part is a good discussion of game theory and how to measure risk when investing in securities. All theories discussed in this part are called "models of rationality."

Part 5, four chapters:

This part continues discussing the concept of uncertainty. Here there are examples of how the results of the models of rationality depart from what actually occurs in reality. Included is more discussion of securities such as stocks and bonds, and new discussions of derivatives (such as stock options) and volatility.

At the end of this book, Bernstein sums up his entire book:

"The central theme of this whole story is that the quantitative achievements of the [historical figures] we have met [in this book] shaped the trajectory of progress over the past [centuries]. In engineering, medicine, science, finance, business, and even in government, decisions that touch everyone's life are made in accordance with disciplined procedures that far outperform the seat-of-the-pants methods of the past. Many catastrophic errors of judgment are thus either avoided, or else their consequences are muted."

Along the way, Bernstein passes along little tidbits of wisdom. For example, "you cannot expect to make large profits without taking the risk of large losses."

One of my favorite sections in this book is when the author does an in-depth statistical analysis of the stock market by using the results of five hundred stocks from January 1926 through to December 1995. His analysis leads to a conclusion that most rational people are intuitively aware of: "The stock market is a risky place!"

Be aware that this book is not an investment cookbook that gives the reader step-by-step instructions of how to invest in the stock market. Also, this book is not exclusively about investment risk. As well, this book is not a textbook.

Throughout the book are pictures, tables, charts, and graphs to aid in the discussion. I found these helpful.

The only weakness I found was that some explanations especially numerical explanations were not clear and were too wordy. At these points, diagrams would have benefited the explanations, making them less wordy, and easier to follow.

In conclusion, this book details the fascinating story of risk. You'll find "that [this] story helps define what it means to be a human being!!"

**** ½

(first published 1996; acknowledgements; introduction; 19 chapters; main narrative of 330 pages; (foot)notes; bibliography, name and subject index)

+++++


Rating:
Summary: History Buffs: Here you go!
Review: Against the Gods draws you through a vast time span. Peter Bernstein begins with the conception of the Arabic numbeting system, up through present time super speed computers. Although, the history found in this book is interesting, the title leads you to believe it is all about investment risks, however it is more of a history text book than a manual. This book is a story of theories and how they developed. You will learn quite a bit about ancient times and how things evolved into the way that they are now, but do not expect any great help or advice on how to deal with risks in the investment world. Once you get into this book, Bernstein's writing sytle draws you in. The book is interesting enough, Bernstein's knowledge of hisotry is astounding. History Buffs: here ya go!

Rating:
Summary: Disappointing - Not really about the stock market
Review: Bernstein is a true historian that writes in a very easy to read and engaging style. I had a hard time putting this book down. He takes you through the rise of mankind's understanding of risk from the intervention of gods to gambling with bones to game theory and beyond. This will also get you a quick diploma in the history of mathmatics along the way. Berstein also writes The Power of Gold: The History of an Obsession which anyone who likes Against the Gods will also want to read (or anyone who wants to learn about the history of Gold and money).