Finding Alpha: The Search for Alpha When Risk and Return Break Down (Wiley Finance) [Hardcover]

Edit: I'm not going to rewrite this review, it's an honest account of what I thought after reading the book. But I've raised the rating to five stars and want to add the comment that I appreciate it more after letting it settle for a year, and after rereading some parts. I still think it's sloppy and doesn't give enough credit to others, but I take back the mean line about not having read the references and not understanding Minksky (I think I must have been a bit mad after wading through the sloppiness). On further reflection, the good points have grown better while the flaws remain the same.

What I like about this book:

* It contains important new ideas that can help any risk-taker with quantitative skills succeed
* It challenges conventional wisdom
* The meat of the book is based on practical experience, not just things that seem right to the author, but things he has tried, and generally with success

What I don't like about this book:

* $95, this should be a $25 list book available for $15 on Amazon and $9.99 Kindle
* Sloppy argument and editing, to the point that some passages are not intelligible (doubly annoying in view of the first criticism)
* Lack of appreciation for other people's thought, which leads to missing useful links

An example of the sloppiness (and there are many) is on page 21, "The key to the portfolio approach is the variance of two random variables is less than the sum of their variance." This makes no sense. He might mean "The variance of THE SUM of two random variables is less than the sum of their varianceS," but this is true only if their covariance is negative, while portfolio theory is more concerned with the positive covariance case. If the variables are uncorrelated, the variance of the sum is equal to the sum of the variances.

He might mean, "The STANDARD DEVIATION of THE SUM of two random variables is less than OR EQUAL TO the sum of their STANDARD DEVIATIONS," which is true, but a stretch from the original. Confusing variance and standard deviation is not something a quant is likely to do. Things like this destroy the value of the book for most people. Readers not confident of their quantitative skills will likely give up and figure the book is too hard (or worse, believe something false). Remaining readers will probably stop after the third or fourth example, figuring the author doesn't understand what he's writing about. Even those who continue on have lost whatever thought the author was trying to express. This stuff is hard enough even when expressed clearly and precisely.

The book's summary of quantitative finance is backed up by lots of references, but I would bet that the author has not read all the references. He doesn't even seem to be familiar with Hyman Minksky's work, and he was a graduate assistant for Minsky. Nassim Taleb is not cited and he has a best-seller that covers some of the same ground.

Unlike most people who dismiss the achievements of academic quantitative finance, the author does not concentrate on how unlikely the assumptions of the models are, instead he deconstructs the equations. But the point of the research is not the equation, it's the identification of the variables, the supporting argument and the empirical evidence. Financial equations are usually trivially simple, but that's not the same as obvious or useless.

For example, the Capital Asset Pricing Model, which is the major target of the book, is a relation between immediate horizon ex ante expected returns of securities. The author assumes that this is a natural and obvious way to think about markets and valuation, then criticizes the equation. But there are lots of other potential ways to frame the issue: prices, values, cash flows, long-term returns, ex post returns and others. And "expected return" only makes sense in a rigorous context (who does the "expecting," and when?). The author scorns rigor, but then uses the concept of expected return in a model with divergent expectations among investors, without discussion of whose beliefs define the expected return, and fails to distinguish that concept clearly from ex post average return and market-clearing return. He would probably call these questions "hair-splitting," as he does to similar objections in the book, but without answers his ideas don't make sense. Nowhere in the book does the author discuss time horizons, everything is stated in terms of a single one-year period. His theory requires that investors pay for risk, but he doesn't consider that risk can be created free (two risk-lovers can flip a coin for money or, more realistically, zero-sum securities can be created with offsetting risk). Why would something that can be created free have a positive price in the market?

However obvious the equations are now, people believed different things in the past, and acted on them. It was only after years of gathering empirical evidence and debating opposing views that the ideas changed practice. More importantly, it is only with this precision and data that the anomalies the author builds his theory on were revealed. No one ever discovered an anomaly without starting from the standard theory, without a standard there are no anomalies. The author implies repeatedly that academic finance ignores or explains away these anomalies when, in fact, they are the focus of research and other people have advanced explanations similar to his, and more radical ones as well. The author has seen far because he stood on the shoulders of giants, but he calls them midgets and tells the world he did it in spite of them.

Okay, with all that, why should you read this book? Because when the author stops his sloppy and foolish top-down theorizing and begins to reason bottom-up from what he knows, he has some brilliant, incisive and useful points. He describes the multiple facets of risk in a new way. It combines ideas from probability theory, behavioral theory and economics, but has a unique structure. It might be a brilliant theoretical advance, I'm not sure, but it is definitely an actionable view. This is risk that makes sense in the trenches. It is pragmatic and will not lead you into silly errors, as almost all other versions of risk can do (of course, some but unfortunately not all of the people who use these other ideas know that, and emphasize that you have to use them with care). This is risk defined for risk-takers.

For one example, and there are many, more than the examples of sloppiness, the book discusses Sharpe ratios on page 245-55 in a quantitatively reasonable way. This is very rare. Or a gem from page 177, "The list of good alpha ideas is highly parochial in practice because a good idea is an improvement on the state of the art, which is peculiar to a specific art." If this seems obvious to you, you are probably a risk-taker, and will find lots of good quantitative analysis in this book. If this does not seem obvious but you got as far as page 177, you are probably a quant who needs risk-taking explained in blunt terms. Both of you will find great value in this book.

The most important single insight is that alpha is not a discovery, but an invention. It's not something you get rich just for finding, it's something you seek because it's an opportunity to invent your way to wealth. Alpha is not a block of gold lying in the street, it's noticing that in some niche, you're a sighted person in the valley of the blind. You are good at telling the difference between good and bad positions of a certain kind, and not enough other people are equally good or better to remove the profit opportunities. Finding this alpha is only step one, you have to work at mining it. Alpha shows you how to mine in a good place (for you), not a played out shaft.

This idea is integrated into a reasonably complete financial theory. The foundation seems solid but, as described above, its superstructure is jumbled and ugly. Also, it has only been subjected to limited testing, so you try it at your own risk. It could change radically or even evaporate as people other than the author attempt to apply it with different skill sets and opportunities. For all of that, this is essentially reading for quantitatively-inclined risk takers.

Eric Falkenstein's book Finding Alpha is a thought provoking book that focuses on the most important issues for a financial practitioner - risk and return. He integrates a variety of research and data sources, some of which are his own, into an unexpected, yet coherent worldview, namely that risk and return are not generally positively correlated.

Falkenstein reaches four basic conclusions:
For most assets, the rate of return is unrelated to its volatility.
Really safe assets have below average returns
Really volatile assets have below average returns
Investors are overconfident, creating excess costs and position concentration.

He argues that this is because:
People are more envious than greedy (i.e. they are more concerned with their position relative to others than in their own absolute wealth.)
People are willing to accept risks in return for the hope (often unfounded) of outsized gains, in investing and other aspects of life.
People save some money in supersafe assets to avoid the chance of losing everything.

Chapter 1 asserts that the common belief that risk begets return (as assumed by most economists and the CAPM) is wrong. He argues that "risk tolerance [in financial markets] is not like physical courage, the ability to withstand a physical pain, but rather like intellectual courage, the ability to withstand ridicule. He argues that finance needs a paradigm shift from CAPM, which doesn't even approximately fit the data, but which dominates the conceptual frameworks of academia and finance practitioners.

Chapter 2 presents the history and derivation of CAPM and later generalizations such as APT, and points out that all share the same basic assumptions about risk aversion.

Chapter 3 presents the history of how academics have bolted, (like Ptolemy) "epicycles" on to CAPM, to explain why it fails to explain the data. For example, why do lower beta stocks outperform high beta stocks (controlling for size), when CAPM's key conclusion is that the opposite will occur?

Chapter 4 provides numerous examples of how higher risks within and between asset classes have not yielded higher historical returns. Most of these examples are within rather than across asset classes. For example, higher leverage stocks underperforming lower leverage companies, the inability of beta to explain fund returns, junk bonds underperforming investment grade bonds, gambling long-shots underperforming favorites, 20 year bonds underperforming 10 year bonds.

Chapter 5 presents data on the behavior of investors that conflicts with the assumptions of CAPM and MPT. The author argues that the underlying utility function (diminishing marginal utility of absolute wealth) employed by economists to explain risk premiums are to blame, and steers the reader to research suggesting that happiness is related to relative status rather than absolute levels of wealth.

Chapter 6 asks "Is the equity risk premium zero?"
The author answers that after adjusting for the use of algebraic returns, survivorship bias, the possibility of large losses, one-time valuation-expansion benefits, tax effects, and transaction costs, a retail investor probably doesn't earn higher returns from equities than risk-free investments.

I think this chapter is a bit weak in that the equity risk premium is a question about the asset class, rather than the typical practice of retail investors. It is worth considering the equity risk premium as a theoretical rather than a practical premium, which can easily be lost to unwise practices. The evidence in the book against within-asset-class risk-return relationships appears to be stronger than across-asset-class risk-return inversion (i.e. that cash would outperform bonds, and bonds would outperform equities). Nevertheless, there is a reasonable argument to be made that that even for fairly long time horizons, one should not assume that historic risk premiums achieved by equities will hold for a given country, and that practitioners should instead make a fundamental assessment of likely future returns.

Chapter 7 checks back in with the CAPM, and notes that despite all of the data and investor behaviors at odds with it, the model is still taught, its assumptions are assumed, and its thought leaders are still given prizes, primarily because the model appeals to philosophically Rationalist (not rational) intellectuals.

Chapter 8 argues that relative utility, rather than diminishing absolute utility, is what people really act upon. A simple example, as well as formulas, are provided to illustrate why investors acting to minimize the variance of relative wealth, (i.e. are more concerned with keeping up with "the Joneses", or the benchmark), creates a condition of no risk premium.

Chapter 9 presents an anthropological explanation of why humans are "inveterate benchmarkers," bringing in Darwinian factors to explain the existence of envy and power-lust, and game theory to justify "reciprocal altruism" (essentially repeated cooperation). The author bizarrely argues that "Envy is necessary for compassion in a developed economy" because one's understanding of a fellow citizen's envy allows one to empathize with their feelings. The chapter's conclusion is less than clear to me, but seems to conclude that while people should be greedy (i.e. pursuing their rational self-interest) if their goal is to maximize absolute wealth, in reality, most people are envious, and working to maximize their status.

In this chapter, the author would have benefitted from exploring the ideas of Ayn Rand on egoism and rational self interest, as well as her psychological exploration of "second handers" in her novel "The Fountainhead." In particular, Ayn Rand illustrates why thinking independently while focusing on one's absolute (not relative) wealth is in one's rational self interest. This is, ultimately, what the author suggests that an investor should do in the pursuit of alpha, but doesn't explain why an individual should value additional wealth over, say, conformity, or trend-following, or the "comfort" of following the herd.
Ultimately, the key thesis of the book rests on the assumption that investors act much like Peter Keating, the leading "second hander" in Ayn Rand's "The Fountainhead," and seek conformity, safety in numbers, and focus more on what others do, think, and own.

Chapter 10 presents the book's general framework of four conclusions, and focuses in this chapter on the idea that risky investments are driven by hope, driving investors towards behaviors more like gambling than return maximization.

Chapter 11 provides some examples of alpha-generating activities, and how they fade over time. Convertible bond "arbitrage," index funds, automated credit scoring, and floor traders are featured. These alpha strategies are really just applications of competitive advantage to finance.

Chapter 12, titled "Alpha Games" goes into the behavior of people, institutions, and investors to acquire, maintain, and divide alpha.

Chapter 13 provides several real-life alpha-creating applications from the author, including minimum variance portfolios, beta arbitrage portfolios, bond investment strategies. These are all fairly simple applications that have delivered superior risk adjusted returns and are consistent with the theoretical assumptions provided by the author. These strategies are mostly institutional-level, and thus not likely implementable for retail investors, yet would not be particularly difficult to implement for institutions. He ends the chapter with some motivational words, encouraging the reader to pursue their efforts to find their signature strengths, employ them to create alpha, and maximize their income and career satisfaction.

Chapter 14 summarizes the book.

Falkenstein's book is likely to strike a chord with experienced investment professionals who think independently, and have seen with their own eyes the many divergences of markets from the scholarly models that dominate the profession. Because his model better fits the historical facts, better matches observations of many investors' behavior, and is reasonably simple, it provides a better approximation of how investment markets function than that of certain Nobel Prize winners.

The book is not perfect. The focus of the book is more about "why risk and return aren't related" than about "finding alpha." Alpha, after all, is a concept borrowed from a theory the author demonstrated to be false. While many ideas are integrated into a comprehensible structure, the framework still seems somewhat ad-hoc. And I am not so pessimistic about investors as to believe that all people are always driven by relative utility. That behavior surely exists, but is far from the only element driving markets, or investors. But it seems a good enough explanation for part of the unexpected deviations of actual practice from standard finance theory.

This book is a good follow up to another good book that had an influence on my professional life, Robert Haugen's "The New Finance", which also demonstrated the failure of beta to explain returns, and provided a mostly institutional explanation for why this might be, including groupthink, benchmark-following, and career-risk aversion.

Falkenstein's book reinforces my view that irrationality and a lack of independent thought are major value destroyers - in investing and in life. This book prepares one to make the rational choice, gather one's intellectual courage, abandon envy (and benchmarks), and instead embrace greed in investing.

The problem with this book is that there is really only three messages and each could have been well covered in a 25 page chapter. The first message is that CAPM, APT and related models do not work in the real world. The author instead spends over a hundred pages dismissing the models and the academia that adhere to them in what borders on hysterical ranting. The second message is that investors holding highly volatile assets are probably not getting enough return to justify holding those assets. The author brings this up in an early chapter and then makes one wait 150 pages to back up his proof on this idea. Finally there is the idea that alpha is usually shrouded in secrecy and cannot be easily discerned. It is insightful discussion but the title is very misleading because this is not a book presenting techniques for looking at portfolios managers or investments (other than the volatily - return discussion). The author tries to save himself with some great insights from his broad financial work experience into how the financial community operates. Very smart guy but you wonder if he would be the life of the cocktail party or the guy who traps you in the corner with a dreadful drawn out story. Unfortunately he seems to have had no help from the editor such that the books seems to be continuously heading off on tangents. While very enjoyable in places, this book was too much work and jumped around too much for the limited return it provided.