What is Alpha?

Introduction to the Alpha Interface Series

Alpha is a term that originated in the 1970s. It defines the degree to which a trading or investing strategy can outperform the overall market indices. To exhibit alpha you must have an edge that allows you to outperform the average of all other market participants. The term alpha underscores a remark made by many motivational speakers, “If you’re not living on the edge, you’re taking up too much space.” You either must be better informed than most other market participants, or you must use a superior methodology.

This edge is rather rare. Most fund managers underperform the S&P 500 average. Many traders who expect to earn a living by their wits find themselves routinely returning their occasional profits back to the market – with interest and penalties.

Are there market patterns that can be exploited by traders using new methods? The answer is both a frustrating “yes” and a “no.” “Yes,” because there is a constant stream of reports in both the academic and popular literature of new angles and methods in the market that, in actuality or on paper, yield a profit. “No,” because alphas eventually fade, at least for a time, once the trading edge is discovered and applied by enough traders in the market.

As an acknowledgment of this profound truth about the ever-changing markets, I have made a point of using the past, or present perfect, tense when describing the research findings in this book. (Present perfect describes actions that have happened at some point in the past. For example, this researcher has found that… Present perfect is also used to describe changes over time, such as: this research has demonstrated that…) I have avoided describing any of the findings in this book as statements of invariant scientific truths. Although the questions used in the section headings may be posed as if one is seeking a universal truth, I have endeavored to frame the answers always in relationship to the time, the place, and the data from which they were obtained.

I find that financial markets do not operate like giant machines, but rather like great ideas. Because markets are always evolving, I have also largely limited my reporting of empirical findings to studies that were released during the past decade.

Science is a search for invariant principles: those that remain constant even while everything around is changing. As a result of this focus, the original papers cited here were all written in the present tense. This tendency is so strong that most scientific researchers even use the present tense to describe prior research done decades ago. Talented researchers seem to truly believe, as exemplified by this deeply imbued literary convention, that their findings represent glimpses of a universal, eternal reality.

Unfortunately, when it comes to the financial markets, the findings of empirical science are rarely, if ever, invariant. One is best advised to view them as statistical probabilities concerning the particular market conditions under which the observations were made.  Careful traders will test these methods and regularly monitor how they perform.

The result of this dynamic is that there is a constant need for traders to become more creative and more sophisticated (i.e., complex and nuanced) in their approach to the markets. To put it more bluntly, it would not be a good idea to bet the farm on any one of the strategies presented in this book even though each has a measure of scientific validation.

Literature on the details of profitable trading is rare, due to its proprietary nature. However, in the past decade, a surprising number of empirical studies have been published in the academic literature. More than ever before in history, traders have the opportunity to empower their trading with knowledge assembled by hundreds of highly trained, scientific researchers. While this information will always fall short of the elusive “holy grail” of trading, i.e. – a strategy that would always work – it does provide a unique edge that can benefit astute traders. Literally, hundreds of such studies have been published with dozens appearing every month! Yet, they might as well be considered secrets as far as the trading community is concerned.

For the most part, this research has not come to the attention of traders and investors. It has been largely kept in the domain of academic journals for the benefit of other researchers and scholars. In preparing this volume, I was unable to find anything similar in the publications aimed at traders and investors.

The reason for this, I suppose, is simple enough. Researchers are generally not engaged in selling either their methodologies or their finding to the general public. No doubt, much research eventually captures the attention of proprietary trading and investing companies, software developers and trading coaches. Clearly, there are hundreds of hypotheses explored by researchers that can stimulate better trading and investing.

The Alpha Interface series is meant to shine a spotlight on scientific studies that would otherwise remain largely hidden from traders. It is a compendium of empirical results, selected and reported after careful review of a much larger universe of published, academic papers. This thorough review is one of the major advantages for readers who will not have to wade through all of the studies that were (and were not) included.

Because of the scope of the project, the fine details of each study are not covered. Due to the limitations of time and space, and for the convenience of readers, I have reported the primary findings only. Many important mathematical and theoretical nuances, and fine points of methodology are skipped over, or mentioned only briefly, so the focus remains on the heart of the empirical findings.

There is no doubt that the original research in this field is complex. If you were to use a scale from one to ten, with one being a summary of research that anyone with a high school education could read and understand, and ten being the actual research that can take years of careful study to fully understand, then most of the research that I am summarizing started out as an eight or higher.

The original papers typically ran from ten to thirty pages in length. I have attempted to narrow down the conclusions of those papers I decided to include here to just a few paragraphs. No doubt, college level statistics will help you plow through some of these summaries. I have sought out the help of smart, educated people who were not traders to provide feedback. They often reminded me when additional definitions were required. Perhaps, however, this dense material is still too dense. I apologize for that. Please let me know when this is the case. I am very open to making additional clarifications for the benefit of future readers.

Any comments you provide will help the next edition, as well as future books in this series, be clearer and easier to get through. Nonetheless, I think that most traders understand this field is not for the weak hearted and the unmotivated. This is a difficult field and requires focused effort to get and maintain alpha.

Readers who are interested in following up on any of the results are encouraged to find and read with care the original research papers. When possible, hyperlinks are included to the original articles (or, at least, their abstracts). However, it is likely readers will discover some of these links to be inactive, as research papers may have been posted online only temporarily. In such instances, I advise copying the article title from the reference section at the end of this book, and pasting it into scholar.google.com.

Certainly, in spite of the wealth of information presented herein, astute readers will realize immediately that there are some limitations to this approach. Academicians, who are often not traders themselves, sometimes report only in-sample findings, or backtest results, without true out-of-sample tests. Other studies fail to account for transaction costs. When possible, I highlight these weaknesses – or exclude those studies altogether. Some papers simply did not provide enough detail for me to be certain that these weaknesses were avoided. Of course, it is generally possible to conduct post-hoc, or over-optimized, studies of market conditions producing results that, while appearing to be highly significant and important, are of almost no real value. Obviously, the best researchers avoid this trap. Some do not.

Conversely, as many researchers appear to be interested in developing a career (or at least a sideline) providing consulting to the large trading and investing houses, it is also very likely that – once preliminary studies are conducted and published – the confirmatory findings are sold and become proprietary and thus unavailable to the research community at large. Even the original research papers do not always provide adequate detail for replication. In some instances, readers who wish to pursue particular strategies will want to directly contact the authors of papers I have cited.

I believe that the main vulnerability rests not on the research papers that are reported, but on those that are not – both positive and negative. On occasion, reports have failed to confirm findings presented in other studies. Because I did not view these failures to replicate as salient, they are not presented in this book.

The tendency of academic journals to publish only positive findings is known in the research community as the “file drawer problem.” It suggests the possibility that the positive findings presented are not truly representative; and that researchers may have file drawers full of unpublished, negative results. Fortunately, there are now repositories for unpublished studies (even with null results) on the internet. I have endeavored to include negative findings as they seemed relevant.

This volume may be justifiably thought of as a compendium of research. Yet, it falls short of being a meta-analysis that would attempt to combine the results of all published (and available unpublished) studies on each topic. That would be a very useful analysis, but it is beyond the scope of the present work.

Another admitted limitation of this presentation is the focus upon empirical research findings, to the near exclusion of theoretical approaches and especially sophisticated mathematical models of the markets. Readers with a special interest in quantification schemes will find some information in the original articles cited here. However, more often than not, highly mathematical papers of a theoretical nature were skipped over completely as they often lack empirical confirmation. Even studies with empirical tests often mystified me regarding their practical value for traders, as their focus was mostly theoretical. Yet, there have been more than enough good scientific studies that made sense to me to include in this book.

Just what are the prospects for successfully implementing the trading strategies presented in The Alpha Interface series? Every trader and investor wishing to acquire an alpha edge will need to figure this out.

One answer has been recently provided by McLean and Pontiff (2013), from the Massachusetts Institute of Technology and Boston College, who examined the impact that publication had on the profitability of 82 characteristics that were identified in 68 different academic studies between 1972 and 2011. They limited their exploration to studies cross-sectional studies using monthly data. These studies were of the type typically used for portfolio analysis in which alpha was typically about 5%. They did not include time-series studies that might be of interest to short-term traders and investors.

They reasoned that, if the reported results were spurious, they should not be replicable at all – even during the out-of-sample period after the conclusion of the original study and before publication. The term “out-of-sample” refers to testing on fresh data that was not used in developing the original hypothesis.

On the other hand, they reasoned that if the reported profitability was associated with risk, profitability should not decline no matter how well publicized were the results. After all, rational investors are not likely to change their strategies, if they believe a well-publicized method is too risky. Finally, they reasoned that if the reported profitability were due to mispricing in the marketplace, they expected to see decay in profitability once the anomaly was reported to the investing public.

They found that there was a decay of about ten percent in profitability of the 82 characteristics between the conclusion of the original study and prior to publication. Statistically, this was not distinguishable from zero. Thus they were able to reject the hypothesis that the original reports were based on statistical errors, data snooping, or other spurious factors. They suggested that the ten percent decay they observed might have been due to some traders and investors learning of the results, and acting on that information, before publication.

At the same time, they observed that the post-publication decline in profitability was about 35%. This decline was largely attributed to the activity of sophisticated traders who became aware of the publication. The market activity of these individuals was sufficient to eliminate about 35% of the mispricing that had originally been reported.

To the extent that this study can be generalized, a reasonable rule of thumb is that readers who choose to implement the trading strategies from The Alpha Interface can expect to achieve results about 65% as profitable as those reported in the original research studies. Even greater success, I believe, will come to those who learn to creatively combine different, independent strategies.

Finally, in preparing this volume, I was particularly impressed with the globalization of financial research. Some very bright and creative researchers around the world are considering increasingly numerous strategies to predict market movement. To help convey an idea of how widespread this enterprise has become, I list the academic and national affiliations of the researchers whose work is presented


One comment on “What is Alpha?
  1. Joe John Lang says:


    I have finally “found” the summary of what you have been doing with this “alpha” phenomenon. This is truly fascinating to me, and I will “read on.” IF I “bracket” my dislike of what people uncritically call “THE scientific method” and my dislike of mathematics and quantitative issues, and if I, having always thought of myself as very or totally naive about investing (I think of it as something like a trip to a casino), can actually understand what you’re “fleshing out” in this book, I hopefully can in a positive way grasp the meaning. I do “get” the “alpha” idea immediately and instinctively, and even though I’ve never been a systematic investor in anything, have observed it occurring with people who are investors, even as they don’t notice it.

    I see now why you are doing more than “Thinking Allowed,” as you have been thinking extensively about a reality of life that is extremely important and that too many of us have never been educated about.

    As a psychologist and educator, I devoted myself to a focus on “human experience” rather than “behavior,” as I believe we have to know the meaning of behavior truly to “understand” it. The huge philosophical revolt that came with “the crisis of the European sciences” (“Geisteswissenschaften”) that became the basis for phenomenology (“back to the things — as the present themselves and we can describe them — themselves) and its huge critique of positivism and scientism are what have kept me busy. Via this critique and insistence on paying due respect to the Geisteswissenschaften came about the assertion that because language only has meaning in a context and that the context for research has to be made as explicit as possible rather than left implicit and that the researcher must “suspend” his prior pre-conceptions to as great a degree as possible in order to let human phenomena speak to him, evolved the thinking that led to the idea that new methods need to be found for doing research, particularly with psychological phenomena, the methods invented to serve our inquiry rather than to be straight-jackets into which we mold our perceptions. Hence we need to engage in what the phenomenologists call “the phenomenological suspension” as we “let” new experiences and perceptions speak to us.

    I have found that for psychologists’ interests phenomenological thinking provides a broad and open way of studying them without straight-jacketing them into ideas that may be conventional but also never adequately “checked out” by “asking people” rather than just “observing” them.

    The institutes and psychology departments that have been trying to study phenomena not studied systematically before have constantly been caught up in a “rift” for their professors and students between what they “do” and want to do and how they may study and do research about these things. I found that by teaching phenomenology in the most jargon-less way possible even to undergraduates I got papers and research that were of immensely high quality. The hardest part of the task is educating people in just enough philosophy to handle the basic concepts.

    But people who have been devoted to this study of phenomenology have not always got along well. Because the “bases” haven’t been clearly defined, people have “combined” and “twisted” phenomenology in many ways, particularly Jungians, that have been exotic and semi-poetic rather than rigorous and discursive.

    I believe that whenever someone is trained in a given rationale or mode of research, as you have been (I can see the influence of your psychology undergrad degree in Madison) if he “works it” through enough samples and applications, it will eventually yield ever-more than it started out doing. I get the impression, just from reading your “intro,” that this is what you have accomplished with the Alpha research and publication. My interest in it will be to discern whether or not both the method and the topic (“alpha” edge) can teach me more about this “traditional” “scientific” research paradigm than I have understood to date. The “hurdles” for someone like me trying to read and study what you’ve written are my own avoidance of the “traditional scientific research paradigm” for many years, my avoidance of quantitative issues all my life, and my lack of knowledge of investing.

    While I “challenge” the ideas of “investing” our societies try to “run on” these days, I don’t know enough to be articulate about these. But I also see investing as “the” mode of survival, other than crime, in contemporary society.

    I will keep trying to follow. You are “in the middle” of a “very hot spot” in contemporary society.

    Joe John Lang

Leave a Reply

Your email address will not be published. Required fields are marked *


Book Three: Trading With The News

Learn about a news-based trading system that yielded a back-tested, average annualized, compounded return from 2000 to 2011 of 58.6%.

“Only once you’ve done your homework will you be able to understand how the stock market works and learn to distinguish between news and noise.” Maria Bartiromo, Use The News

Book Two: Technical Analysis

Learn about the "trend recalling" algorithm that yielded researchers a simulated annual return of greater than 400% in multiple tests.

“The scientific method is the only rational way to extract useful knowledge from market data and the only rational approach for determining which technical analysis methods have predictive power.”
David Aronson, Evidence Based Technical Analysis

Book One: Analysts’ Forecasts

Learn the strategy, based on analysts' revised forecasts, that yielded researchers an average of 1.13% - 2.19% profit per trade, for trades lasting one to two days?

Learn how certain analysts' recommendations, following brokerage hosted investment conferences, yielded profits of over 3% during a two-day holding period?

Learn how researchers found an average profitability of 1.78% for two-hour trades following an earnings announcement?

"This set of tools can help both ordinary and professional investors alike to re-think and re-vitalize their stock picking, timing and methods. A young, aspiring Warren Buffet could put this book to good use."
James P. Driscoll, PhD, investor

Statistically Sound Machine Learning for Algorithmic Trading of Financial Instruments by David Aronson (software included)

Evidence-Based Technical Analysis by David Aronson

Archive of Earlier Posts