18 Issue 63 – January 2009 MARKET TECHNICIAN

Perspectives on Technical Analysis

By Jasmina Hasanhodzic

Introduction therefore never become obsolete is often heard from seasoned

Ever since the emergence of modern finance theory, the followers technicians. But if fear and greed have always been the main
of technical analysis have found themselves marginalized by the drivers of supply and demand in the markets, then technical
financial establishment and held at bay by fortress walls built on analysis that measures supply/demand imbalances should have
misunderstanding and contempt; misunderstanding, since the spread its web of roots deep in the realms of history and widely
technical jargon of head-and-shoulders, wedges, and pennants across civilizations. Historical evidence suggests that this indeed is
does not resonate with minds steeped in physics and statistics; the case. Arguably as old as the markets themselves, technical
contempt, for the fervent adherents of modern finance reject the analysis dates back to the ancient Babylon of the 7th century B.C.
mere possibility of patterns in market prices. Such a segregative This evidence comes from a set of price diaries that survived until
state of affairs hints at the striking dichotomy which permeates the our time,1 where the prices of the same six commodities such as
financial world. On the one hand, academics tend to embrace the barley, dates, sesame, and wool were carefully charted,
efficient markets hypothesis which, roughly speaking, stipulates continuously through centuries, with the purpose of predicting
that markets are unpredictable, in the sense that there is no way to them. In fact, just like contemporary technical analysts do, the
forecast future prices based on previous ones and thereby make a diary keepers would adjust the frequency at which they recorded
profit. The efficient markets hypothesis was shaped in the writings the market quotations according to the level of market volatility.
of Eugene Fama and Paul Samuelson in the 1960s, and despite the
recent proposals of alternative hypotheses, it remains deeply More recently, technical analysis was used in the 17th century at the
ingrained in the culture of academic finance. Amsterdam Exchange where business was conducted in a way that
is remarkably similar to that of current exchanges. In Confusión de
This stands in sharp contrast with technical analysis, which Confusiones (1688), one of the oldest books ever written on stock
attempts to divine market’s direction based on past prices. Often exchange business, Joseph de la Vega provides a vivid account of
and quite controversially this is done by searching with the naked the dynamics of this Exchange and observes that “on this point we
eye for certain patterns in price histories that are believed to are all alike-when the prices rise, we think they will run away from
embody the prime mover of all market action: crowd psychology. us,” suggesting that people at that time were naturally inclined to
For example, a sequence of successively higher highs and lower make inferences about future prices based on past ones.
lows which traces the so-called “triangle bottom” pattern in the
price data is usually interpreted by technicians as a manifestation From the 18th century, the evidence that technical analysis was
of strengthening confidence punctuated by subsiding terror, and used in Japan comes from the writings of the legendary trader
hence as the presage of an uptrend. Perhaps surprisingly, Munehisa Homma, who in his book The Fountain of Gold describes
technicians in some cases do appear to make consistent profits how market psychology can help predict prices; for example,
that cannot be attributed to chance alone. “when all are bearish,” he suggests that “there is a cause for prices
to rise.”2 Homma also documents the “candlesticks” charting
In the light of this dichotomy, the rift between academics and technique, which is used to this day, especially in Japan.
technicians is not quick to heal, but may require the two
communities to take a careful look at the other. In this article we Skipping one century ahead, to late imperial China, the similarities
present our insights from one exploration of technical analysis, between Chinese merchant manuals and present-day technical
spanning past, present, and future. First, to place technical analysis analysis texts manifest themselves in the emphasis of both on the
in context, we conduct a historical study of its evolution through cyclicality of markets and the role it plays in market timing. The
time and across cultures. Second, we deal with the current lack of statement from Essential Business, one such manual, that “when
formalisation of technical analysis: there is no complete agreement goods become extremely expensive, then they must become
on what technical analysis is, with many texts only partially inexpensive again”3 is as fully understandable now in the context
overlapping on the methodologies to be employed. Consequently, of trend reversals as it was then.
our second main endeavour is to understand fully what present-
day technical analysis consists of, and to this end we conduct The foundations of the present day technical analysis were laid
interviews with its current leading practitioners. Finally, we down in the early 1900s by Charles Dow. In the January 4th, 1902
consider the future of technical analysis, which in our view will edition of The Wall Street Journal, Dow proposed his famous
move away from ambiguous heuristics towards formally defined definition of a trend as a sequence of successive highs and lows by
indicators and dynamic strategies, capturing the ever-changing saying that “it is a bull period as long as the average of one high
nature of financial markets. In the rest of this article we discuss point exceeds that of previous high points.” However, even this
these three perspectives in turn, each the subject of a forthcoming most recent form of technical analysis has never been formalised,
book co-authored with Andrew Lo. but continues now as always to flourish on the fringes of the
establishment and pass from one generation to the next through
The historical context of technical analysis odd apprenticeships and confabulations, remaining to this day a
That human nature will never change and technical analysis most colorful part of the financial folklore.
 MARKET TECHNICIAN Issue 63 – January 2009 19

Insights from the interviews with leading technicians In their quest to quantify technical analysis, academics have
Given the lack of standardisation of technical analysis, learning turned, too, to the most controversial of its techniques: patterns.
about it from books is a daunting task: the contradictory advice Finding patterns in price charts is a subjective endeavour that
and loose and frequent use of qualifiers such as “often” and relies on the natural smoothing filter of the human eye; a main
“probably” throws the uninitiated into ambivalence and confusion. challenge therefore in quantifying the patterns lies in modelling
Learning about technical analysis from its leading practitioners is the way in which eyes smooth the data they view. The pioneering
more inviting; after all, that is how this intricate craft survived till works by Chang and Osler (1994, 1999) and Lo, Mamaysky, and
our time-not through books, but through the word of mouth. Wang (2000), as well as the recent extension of the latter by the
author (Hasanhodzic, 2007), take on this challenge by smoothing
If a single word could summarise the interviews we conducted, it the data using statistical filtering techniques, such as kernel
would be “diverse,” as it captures both the intriguing beauty of this regression and neural networks. They then develop algorithms to
craft and the troubles that plague it. Spanning a striking variety of identify automatically technical patterns in that data, and finally
styles in their practice of technical analysis, our interviewees range evaluate the information content of the patterns thus found.4
from short-term traders (Raschke, Weinstein) to educators (Murphy, These works, too, find proof of the validity of technical analysis.
Acampora), long-term investors (Desmond, Deemer), artist
technicians (McAvity), highly eclectic technicians (Dudack), historians The observation that human nature never changes, and that
(Shaw), long-term market theme writers (Farrell), and those who consequently technical indicators designed to measure the
insist on being labeled market rather than technical analysts (Birinyi). reflection of human nature in market prices never change either, is
a notable argument in favour of technical analysis, but one that at
Diverse, too, are their interpretations of technical analysis, a telling the same time underscores its main shortcoming: technical analysis
illustration of which is provided by their responses to the question has not kept up with technological advances. Of course, charting
“does the lack of hard and fast rules in technical analysis ever and data collecting have become automated, but most popular
bother you,” which varied from “that’s exactly what bothers me” to patterns and heuristics of today were developed in the pre-
“but there are hard and fast rules in technical analysis.” The computing age when calculating a simple moving average was a
interviewees do agree however on one point, and that is the formidable task. For example, a 10-day moving average became
significance of the role intuition plays in their decision making. As popular because it was easy to compute-one could divide by ten
Murphy puts it, “I’m not sure I could explain to you how I do what I simply by moving the decimal place to the left-not because it was
do. I look at many things in a short interval of time and come to a statistically optimal, and it remains commonly used to this day.
conclusion.” That elusive skill may explain why these practitioners
do not have a problem with sharing their knowledge, the tools Suboptimal parametrisation is only a symptom of a chronic disease
they develop, or the strategies they pioneer. There is no single right afflicting technical analysis: its static nature cannot account for the
way to put it all together. Everyone does it in slightly and ever-changing character of financial markets. In the past, when
sometimes widely different ways. Some technicians operate best in execution was manual and costly, and financial systems far less
complete isolation from the outside world-Linda Raschke, who connected and complex, “static” used to be a prerequisite for
does not watch TV or read The Wall Street Journal, is a prime implementability, now it is a condition for failure. As markets
example-others prefer to complement their technical perspectives evolve and trading strategies become more sophisticated, the
with fundamental, economic, and political ones. need for new, dynamic indicators becomes apparent. Never has
this need been more urgent than now, when in the light of the
The palpable heterogeneity among leading technicians that events of last year’s August, one can no longer deny that hedge
emerges from our interviews yields a potential explanation for the funds have become a paramount force: it is not the feeling of the
lack of impact that technical analysis has had on the broader crowd but the action of a single gigantic hedge fund that
financial community. Without a unified, standardized, and broadly precipitated the crisis continuing to this day. This is echoed by
recognised body of knowledge in which every practicing many of our interviewees, including Walter Deemer who finds
technician must be conversant, it is difficult to see how technical ingenious ways of tracing the hedge-fund activity indirectly, via
analysis can spread. The advent of the Chartered Market certain mutual funds: “I am convinced that these 6 or 7 billion
Technician program by the Market Technicians Association is a step dollars of assets in Rydex funds reflect the general hedge-fund
in the right direction but, as our interviews underscore, there is still trading activity which is the driving force in the market,” he notes.
a considerable amount of art and subjectivity in the practice of
technical analysis. Directly tracking hedge-fund activity is of course what one would
want, but this is problematic if not infeasible for, shrouded in
Towards a science of technical analysis secrecy, hedge funds are notorious for their inaccessibility and
For too long languishing in the murky waters of part art-part infrequent reporting. A possible way to circumvent this problem is
science, technical analysis is finally starting to develop a more jointly proposed by the author (Hasanhodzic and Lo, 2007), and
rigorous approach. Big strides, indeed, have been made towards consists of replicating hedge funds in a transparent and publically
the standardization of technical analysis in recent years, such as by available format. Starting from the empirical observation that a
Aronson (2007) and Kirkpatrick and Dahlquist (2006). The impetus large portion of hedge-fund returns can be obtained by passive
for statistically evaluating technical analysis naturally comes from exposure to certain common risk factors-such as those captured by
academia, with studies yielding evidence of its validity in wide- stocks, bonds, currency, commodity, and credit markets-we show
ranging areas, such as moving averages (Brock, Lakonishok, and how hedge-fund returns can be expressed in terms of returns
LeBaron, 1992), genetic algorithms to discover optimal trading realized in those markets, via a linear regression model. Based on
rules (Neely, Weller, and Dittmar, 1997), and the Dow Theory (Brown, this idea, a number of prominent asset-management firms have
Goetzmann, and Kumar, 1998), to name a few. recently launched mutual funds that aim to replicate hedge-fund
20 Issue 63 – January 2009 MARKET TECHNICIAN

factor exposures. Unlike hedge-fund returns, the returns of these Lo, A. and J. Hasanhodzic, 2009, The Heretics of Finance:
funds are public and updated daily, and can be used as an indicator Conversations with Leading Practitioners of Technical Analysis.
of the aggregate performance of the hedge-fund sector. Moreover, New York, NY: Bloomberg Press.
assets flowing in and out of these mutual funds are publically Lo, A. and J. Hasanhodzic, forthcoming, Tales of Future Past: The
available, which raises the possibility of using them to gauge the Fascinating Story of Technical Analysis. New York, NY: Bloomberg
investors’ sentiment shifts. Press.
Lo, A. and J. Hasanhodzic, forthcoming, A Quantitative Approach to
Conclusion
Technical Analysis. New York, NY: Bloomberg Press.
In this article we presented some of our perspectives on technical
analysis spanning past, present, and future. Our historical Lo, A., Mamaysky, H. and J. Wang, 2000, “Foundations of Technical
exploration reveals that technical analysis was a force throughout Analysis: Computational Algorithms, Statistical Inference, and
centuries and across cultures. Our interviews with the present-day Empirical Implementation,” Journal of Finance 55, 1705-1765.
masters of technical analysis help us understand what technical Lufrano, R., 1997, Honorable Merchants: Commerce and Self-
analysis is, and armed with that understanding we can move Cultivation in Late Imperial China. Honolulu, HI: University of
towards the future, which in our view consists of standardisation and Hawaii Press.
development of new indicators to measure fast-changing market Nison, S., 1994, Beyond Candlesticks: New Japanese Charting
environments. Although the fortress walls separating technicians Techniques Revealed. New York, NY: John Wiley & Sons.
from the adherents of modern finance still stand tall, they are not
Neely, C., Weber, P. and R. Dittmar, 1997, “Is Technical Analysis in the
insurmountable, and we hope that the growing volume of in-depth
Foreign Exchange Market Profitable? A Genetic Programming
examinations of technical analysis will awaken the sceptics and open
Approach,” Journal of Financial and Quantitative Analysis 32,
the door for the dialogue between the two communities to begin.
405-426.
Slotsky, A., 1997, The Bourse of Babylon. Bethesda, MD: CDL Press.
1
As documented by Slotsky (1997).
2
As quoted by Nison (1994). Jasmina Hasanhodzic, Ph.D. is a research scientist at AlphaSimplex
3
As quoted by Lufrano (1997). Group, LLC. The views and opinions expressed in this article are those
4
It is important here to distinguish between evaluating of the author only, and do not necessarily represent the views and
“profitability” of technical trading rules and evaluating the opinions of AlphaSimplex Group, MIT, or any of their affiliates and
“information content” of technical analysis; the former necessitates employees. The author makes no representations or warranty, either
the modelling of the trading implementation and risk expressed or implied, as to the accuracy or completeness of the
management, whereas the latter detects supply/demand information contained in this article, nor does she recommend that
imbalances regardless of whether or not one can profitably act on this article serve as the basis for any investment decision-this article is
that information. for information purposes only. The research summarized in this article
was supported by the MIT Presidential Fellowship and the MIT Lab for
References Financial Engineering.
Aaronson, D., 2007, Evidence-Based Technical Analysis. Hoboken, NJ:
John Wiley & Sons.
Brock, W., Lakonishok, J. and B. LeBaron, 1992, “Simple Technical
Trading Rules and the Stochastic Properties of Stock Returns,”
Journal of Finance 47, 1731-1764.
Brown, S., Goetzmann, W. and A. Kumar, 1998, “The Dow Theory:
William Peter Hamilton’s Track Record Re-Considered,” Journal of
Finance 53, 1311-1333.
Chang, K. and C. Osler, 1994, “Evaluating Chart-Based Technical
Analysis: The Head-and-Shoulders Pattern in Foreign Exchange
Markets,” working paper, Federal Reserve Bank of New York.
Chang, K. and C. Osler, 1999, “Methodical Madness: Technical
Analysis and the Irrationality of Exchange-Rate Forecasts,”
Economic Journal 109, 636-661.
De la Vega, J., 1957, Confusión de Confusiones. Cambridge, MA:
Harvard University Printing Office.
Hasanhodzic, J., 2007, “Investments Unwrapped: Demystifying
Technical Analysis and Hedge-Fund Strategies,” Ph.D. thesis,
Massachusetts Institute of Technology.
Hasanhodzic, J. and A. Lo, 2007, “Can Hedge-Fund Returns Be
Replicated?: The Linear Case,” Journal of Investment Management
5, 5-45.
Kirkpatrick, C. and J. Dahlquist, 2006, Technical Analysis: The
Complete Resource for Financial Market Technicians. Upper
Saddle River, NJ: FT Press.