IS YOUR BROKER TAKING MORE

THAN HIS FAIR SHARE?

Keep more of what’s yours with our low cost platform
for professional traders, investors and institutions.

Interactive Brokers Interactive Brokers Interactive Brokers

Commissions are: 1
Financing Rates are: 1 Executions1 are:

64-89% 78% Stocks:

better than
lower than
• Charles Schwab
lower than
• Charles Schwab
23¢ the industry
per 100 shares

• TDAmeritrade
• ETrade
• TDAmeritrade
• ETrade
Options:
20¢
better than
For 100 to 1000 shares and For a margin loan of the industry
1 to 10 option contracts $50,000 to $1,000,000 per contract
Second Half 2013*

Interactive Brokers ibkr.com/iwantmore

LOWER YOUR COSTS TO MAXIMIZE YOUR RETURN

stocks • options • futures • forex • bonds — on over 100 markets worldwide from one account
Interactive Brokers is a member of NYSE, FINRA, SIPC – Lower investment costs will increase your overall return on investment, but lower costs do not
guarantee that your investment will be profitable. Supporting documentation for any claims and statistical information will be provided upon request.
Commissions and Margin Rates as of March 3, 2014. Services vary by firm. [1] For additional information on our commissions and margin rates and
execution quality, see www.interactivebrokers.com/compare. * Based on independent measurements, the Transaction Auditing Group, Inc., (TAG).
05-IB14-756CH747
 Incredibly Pleasingly The
FAST EASY BEST
Voted BEST
Fast installation, faster scans, fastest Our goal at Worden is to turn “too much”
results. Everything about TC2000 is information into manageable chunks.
made for speed. It’s the only software Customize your own layout windows 21 years straight
that can scan any indicator in real-time with automatically maintained Best Stock Software

“
with no programming required. watchlists, consolidated news feeds Under $500
and the industry’s most vibrant charts.
1993-2013
TC2000 is designed for ease of use.

“Awesome gents, simply awesome. Smoking

fast, well thought out.”
— Platinum subscriber, Monte C., Lawrenceville, GA

“I have been browsing through the video

tutorials and want to say that I clearly see why
your product has been voted best for so long. It
is amazing and I wish I’d discovered it sooner.”
— Martha S., San Francisco, CA

“Your charts are a visual dream. So easy to read

and adjust. Thanks.”
— Ms. Bonnie A., Encino, CA

“…you guys are the best. I am a long time user

”
and you have the best charts. No one beats your
candles.”
— Voicemail left by Customer
®

“I’ve been here on and off for 15 years, the

product gets better every year.”
— Dave V., Galveston, TX
Compatible with

BRAND NEW VERSION 12.4

WATCHLISTS | CHARTING | INSTANT REAL-TIME SCANS AND SORTS FREE
ALERTS | NOTES, NEWS, & CHAT | MARKET DATA | SECTORS
14-Day Trial
Visit TC2000.com/TryFree Today!

Worden Brothers, Inc. • PO Box 1139 • Wilmington, North Carolina 28402

Mon-Fri 9:00am - 5:00pm ET, Sat 10:00am - 2:00pm ET • 800-776-4940 • 919-408-0542 • www.Worden.com
Stocks & Commodities V. 27:4 (46-51): Andrew Lo And The Future of Technical Analysis by J. Gopalakrishnan & B. Faber
INTERVIEW

The Chronicler Of The Heretics

Andrew Lo And The Future

Of Technical Analysis

Andrew Lo, who is the Harris & Harris Group Professor of Finance at
the Massachusetts Institute of Technology’s (MIT) Sloan School of
Management and director of MIT’s Laboratory for Financial Engineer-
ing, and founder and chief scientific officer of the AlphaSimplex Group,
a Cambridge, MA–based quantitative investment management com-
pany, is a radical of sorts, an academic who believes in technical
analysis. In his latest work, The Heretics Of Finance (cowritten by
Jasmina Hasanhodzic), more than a dozen notable technicians are
interviewed about their views on technical analysis and how it influences
their trading. Interviews aren’t unusual — but what Andrew Lo wants to
do with the sum of their knowledge is.
STOCKS & COMMODITIES Editor Jayanthi Gopalakrishnan and Staff
Writer Bruce Faber spoke with him via telephone on February 6, 2009.

A
ndrew, what got you inter- in the possibility of technical
ested in technical analysis? analysis having something to
I’ve been interested in say about financial markets.
technical analysis since high And shortly after that, I began In the end we all have the
school. It was a bit of a hobby in terms of doing some direct research on same goal, which is to
looking at various charts and patterns technical indicators. forecast uncertain market
and trying to understand what implica-
tions they might have for the markets. I JG: Many of our readers are
prices. We should be able to
actually did not get a lot of time to devote diehard technical analysts and learn from each other.
to it because as part of my doctoral pro- may not be that familiar with
gram in economics, I was taught early on the Random Walk hypothesis, so can ket prices. So if you can’t predict mar-
that technical analysis was a waste of you explain what it is? ket prices, then you might as well give
time. I was still interested in it, but cer- Sure. It was an idea developed in the up on technical analysis as well as other
tainly that interest was dampened by the 1950s and 1960s theorizing that market forms of active investment management.
training I received as an economist. prices are unpredictable, at least as ap-
plied to the financial markets. From the JG: So what did your findings tell you?
JG: When did that change? economic perspective, the reason why The paper that I wrote with Craig
It wasn’t until I was an assistant pro- is that many individuals are trying to MacKinlay that was published in 1988
fessor at the Wharton School and started forecast them, and the more that people was titled, “Stock Market Prices Do Not
doing original research on testing the try to forecast them and incorporate Follow Random Walks.” That was our
Random Walk hypothesis that it be- their information into market prices, the basic finding. If you believe in the Ran-
came clear to me that, first of all, the less easy it will be to predict and the dom Walk, then there were certain im-
Random Walk was not a good model for more likely the prices will look random. plications that had to hold and that you
stock market prices, and second, if the So the theory of the Random Walk ba- could test statistically. When you did
stock market was not described by Ran- sically says: Since so many people are that, you got a significant rejection us-
dom Walk, then maybe there was some- trying to predict markets, and since in- ing a variety of statistical methods. That
thing to all of these technical patterns I formation is therefore incorporated into was the bottom line in the first paper
had read about. So I got interested again market prices, you cannot predict mar- that we wrote. Then, in a series of 10
Copyright (c) Technical Analysis Inc.
Stocks & Commodities V. 27:4 (46-51): Andrew Lo And The Future of Technical Analysis by J. Gopalakrishnan & B. Faber
INTERVIEW

papers that we wrote afterward, we docu- cians do. The series of interviews really picked 10 patterns because we wanted
mented further aspects of this nonran- came about from that line of research. to have a reasonable set to analyze. We
dom behavior of stock prices. We looked picked the most difficult ones to quan-
to see whether there was any kind of JG: Out of the interviews that you did tify simply because we wanted to be
market microstructure effects. Was it with a number of well-respected tech- able to give the benefit of the doubt to
the bid–offer spread that was inducing nicians, I found that each person is the academics and argue that even in the
these predictabilities? Was it the sec- unique in the way they interpret the case of these difficult patterns, like a
tors? Was it the particular time of day markets. Wouldn’t that make it diffi- head & shoulders, to identify that there
for trading? cult to quantify technical analysis? is value in observing their occurrences.
It does. It highlights the challenges of So in that 2000 paper we took these
JG: What was the answer? systematizing technical analysis, and it 10 technical patterns and developed
Little by little, we began to chip away also explains, sociologically, why tech- mathematical descriptions of them so
at this Random Walk hypothesis that nical analysts have not received the that we could automate the process of
most academics had been taught to the kind of respect that the fundamental identifying these patterns. Then we ran
point where we finally concluded, after analysts or quantitative portfolio man- those algorithms on historical data go-
probably a decade of research, that mar- agers have. It’s because technicians ing back about 30 years across a hun-
ket prices definitely have predictable don’t yet have a well-developed com- dred different stocks distributed evenly
components in them, and there are many mon body of knowledge or even com- across five market capitalization
ways of getting at those components, mon lexicon that they can use to discuss quintiles to get a representation sample
including technical analysis, and other their ideas. of stocks.
kinds of mathematical and statistical
techniques. JG: Do you see that changing? JG: What else?
It’s probably something that will We asked a question: Every time you
BF: How did the Random Walk theory change over time, but it’s interesting to saw a head & shoulders pattern, what
become so popular, and so entrenched, see the evolution of this particular field. would that tell you, if anything, about
when it is so obviously flawed? It has definitely not got to the point what that particular stock was going to
I think it was because it was proposed where it can be considered a full-fledged do three days later, or five days later, or
by two very well-known economists, type of scientific discipline the way that seven days later? We were trying to get
Paul Samuelson and Eugene Fama, and portfolio management or portfolio opti- a sense of how much information was
when they tested it in the data back in mization has become pretty system- contained in one of these patterns in
the 1960s and 1970s, it looked like a atized. But it also suggests that each terms of forecasting the underlying se-
good approximation. I think in retro- technical analyst has some insights that curity after the pattern occurred.
spect it was also because the Random they have in common with others.
Walk was culturally tied up to the effi- BF: What were the results?
cient market hypothesis, the idea that JG: What’s an example? It turned out that for these 10 patterns
markets always reflect available infor- The notion of managing your bal- that we looked at — we separated them
mation and that you cannot, systemati- ances, your assets, thinking about stop- into the New York Stock Exchange
cally, generate profitability. It sounded loss settings, and looking at support and (NYSE) stocks versus NASDAQ stocks
like an impelling and counterintuitive resistance levels. A number of things — we found that for the NYSE stocks
idea that took on a life of its own in the are common across these interviews in these patterns didn’t work particularly
academic community. It is really a cul- Heretics that suggest that a systematic well. You know there were a couple of
tural kind of excitement. field of knowledge can be developed. patterns that seemed to have some reli-
That is the hope, to try to come up with able information content in them, but
JG: Are your findings from the re- a common language and a common set what we found is that for the NASDAQ
search you did on the Random Walk of ideas that can be called the science of sample, all of the indicators had tre-
hypothesis what led you to write The technical analysis. mendous statistical significance. It made
Heretics Of Finance? a very big difference for the three-day
Yes, it was part of the research pro- JG: Have you done any work on spe- stock return, whether or not a head &
gram that I started out on many years cific chart patterns or indicators? shoulders pattern had occurred before
ago. The idea was to understand what it I have. With some coauthors, Henry that. So if you knew that a head &
was about market prices that caused Mmaysky and Jiang Wang, about the shoulders occurred, that would be very
predictability. And that led to a number year 2000 — it has been almost 10 valuable information for N ASDAQ
of different research directions, includ- years! — we published a paper where stocks. That was probably the first time
ing my work in behavioral finance, we looked at 10 specific indicators; that there was any real documentation
hedge funds, and proprietary trading, head & shoulders, triangles, rectangles, in academic literature that a truly tech-
and ultimately, looking at what techni- inverted bottoms, things like that. We nical pattern added some value from the
Copyright (c) Technical Analysis Inc.
Stocks & Commodities V. 27:4 (46-51): Andrew Lo And The Future of Technical Analysis by J. Gopalakrishnan & B. Faber
INTERVIEW

economic perspective. were looking at relatively short term. law of gravity, it doesn’t get annoyed at
These were like 20-day horizons. Our you and change its mind. In the stock
BF: Do you have any idea why the hope was that the paper we published market, you definitely have that kind of
patterns work better on the NASDAQ was really meant to be a proof of con- a feedback, so that is another challenge
than the NYSE? cept for others to start doing more sys- that nobody has taken seriously in the
We do. The NYSE tends to have much tematic research in technical analysis. literature. And it is definitely worth
larger investors — that is, institutional So for example, the question you just pursuing.
investors that trade using a certain dis- asked, long term versus short term, that
cipline, and also those stocks tend to be is actually an easy question to answer. JG: Most of our authors and readers
more informationally efficient because, In the context of our study, all you need are sophisticated technical analysts and
by definition, they are larger stocks that to do is lengthen the horizon of the they tend to be flexible. They may be
are the focus of all of the analysts and all holding periods and then you would be disciplined, but if something isn’t work-
the pension funds and so forth. There able to test that quickly. ing correctly, they have to alter their
are more people looking at them and trading systems.
there is a more active type of informa- JG: That sounds as though you should That is right. That’s another aspect of
tion-gathering. be able to do it. technical analysis research that needs to
But the NASDAQ markets are smaller The problem is that in technical analy- be studied. Technical analysts some-
stocks with more retail investors in- sis there isn’t the same discipline and times act as if it doesn’t matter why a
volved. So you would expect that devotion to empirical testing that there pattern is there, you just have to use it.
NASDAQ stocks may be more driven by is in quantitative and fundamental analy- Over time, we are going to want to dig
investor psychology rather than by these ses. Even a fundamental analyst who is deeper into the analytics to develop the
disciplined, automated algorithms for not a statistician would look at book to underpinnings of behavior. It could well
trading. At least that is our conjecture market as a reasonable factor for stock be that a head & shoulders pattern is
about why that difference exists. selection and he would test it out over a something behavioral that is based upon
20-year period. He would go back into certain kinds of investor psychology.
JG: What did you use to identify those the data and look to see whether book to That is useful to know if it is true, and
10 specific patterns in your paper? market actually yielded stocks that ulti- the particular mechanisms are useful
The algorithm is described in the pa- mately ended up performing better or because they could change over time,
per, but briefly, we identify a head & worse. That is the kind of discipline that particularly in crisis periods like we are
shoulders pattern mathematically as a I think technical analysts may be lack- in now. It may end up that fear and greed
price pattern where you simply have ing. are a lot more important over the last 12
three local maximums — two shoulders I am generalizing, of course, because months than it has been over the last six
and a head — and the middle maximum there are some technical analysts who years. That is the kind of thing that
is bigger than the other two by a certain are extraordinarily disciplined and so- would be useful to take into account if
amount. Then we used something called phisticated from a statistical perspec- we were able to do more research.
“kernel regression” to essentially put a tive — see the recent textbook by Charles
curved, continuous line through a scat- Kirkpatrick and Julie Dahlquist — but JG: You just mentioned the current
ter of points. that does not characterize the entire crisis situation. Do you have any
field as of yet. Our hope is that by taking thoughts on whether, in a situation
JG: What’s that? our approach, we would be able to put where markets are going down as they
It’s kind of an approximation algo- technical analysis on firmer footing. have been, technicians are at an ad-
rithm that evens out the ups and downs vantage when it came to managing
of prices to get a smooth curve. Then JG: One of the things I noticed from their portfolios?
you use standard calculus methods to the people you spoke with in your book Definitely the technical analyst has
identify the optimum. As long as the is, a lot of them mentioned that you an advantage now. These are the kinds
maxima are within a certain distance, have to be flexible in the sense that if of periods that technical analysis was
you have a head & shoulders pattern. So you have something that works, it may designed to capture. Much of technical
it is just coming up with a mathematical only work for a short period of time, analysis is based on investor psychol-
rule that can be used to say, “Whenever the market may change and that par- ogy and behavior as well as supply and
these conditions hold, you’ve got a head ticular system may not be applicable demand considerations. But they are
& shoulders.” any more. Would that be another chal- done in a relatively practical and useful
lenging aspect? kind of a setting. Economists tend to
JG: Did you do any work on whether Absolutely. Unlike physics, where think analytically about supply and de-
these indicators work better in the long when you drop a ball it doesn’t know or mand too. During periods of normal
term or short term? care that it is obeying the law of gravity. markets, that is fine. You will definitely
We didn’t really look at horizons. We And if you tell the ball it is following the be able to say something about market
Copyright (c) Technical Analysis Inc.
Stocks & Commodities V. 27:4 (46-51): Andrew Lo And The Future of Technical Analysis by J. Gopalakrishnan & B. Faber
INTERVIEW

dynamics using traditional economics. generally accepted technical indicators question is why. From what neurosci-
But during periods of extraordinary may be a reflection of commonly used entists can conjecture, the brain works
stress and market turmoil is where psy- patterns that yielded profitable trades. It’s in a very different manner. It has a
chology takes over, and technical pat- just a conjecture at this point, but we hope different architecture than a computer,
terns actually work better now than they to confirm or reject it in our ongoing and it has organically evolved to engage
do during normal times. Frankly, that is research on the statistical foundations of in pattern-matching. In that sense it is
one of the reasons why I think technical technical analysis. not surprising technical analysis devel-
analysis works better in, say, forex and oped as a distinct field, because it is in
commodity markets than equity mar- BF: Do you trade at all, or do you just fact the most natural act that humans
kets. Equity markets have so much fun- study trading? can engage in when confronted with data
damental information that is relevant to About 10 years ago, I started an asset of the type that we have in the financial
the fortunes of the company, much more management company called markets. Because it is more natural, there
so than a currency or commodity. AlphaSimplex Group, which manages are certain aspects that may never be
Currencies have fundamentals, but a couple of hedge funds, some mutual automated, and maybe should not be
they don’t change day to day. So if you funds, and overlay strategies. So we do until we change the architecture of the
asked what drove changes to currency trade, and we are managing about $600 modern digital computer. That tells us
markets day to day, the answer is going million at this point. more about where technical analysis is
to be mainly investor psychology. That’s going to be more important, and where
why most foreign currency traders swear BF: Do you use black boxes because of it might be less successful.
by technical analysis, whereas most your mathematical background? When it comes to number-crunching,
equity traders don’t. We don’t call them black boxes. We technical analysis has nothing to say
call them glass boxes. We tell our inves- about that. When it comes to pattern-
JG: That’s true. From the people you tors what we do and we give them some matching, which is quite different than
spoke with for your book, did you find sense of how we do it. We are pretty number-crunching, fundamental analy-
that a lot of them used a combination investor-friendly. We like to think of sis has nothing to add. It is all about
of fundamental and technical analy- ourselves as being fairly transparent, technicals. These different approaches
ses, or did most of them use technical even though we do have some technolo- and perspectives have their strengths
analysis just by itself? gies that are fairly hard to replicate. But and weaknesses. What my coauthors
It varied. Some of them were much we are not a black box. We prefer to let and I are going to do next is try to unlock
more agnostic and open about the dif- our investors know what we are doing. those differences, and maybe from that,
ferent tools that they applied to their develop the technical analysis of the
trade. Others were diehard technicians JG: What is next for you after this book, next generation to combine the best of a
who looked at nothing but charts. That now that you have discovered so many variety of these different disciplines.
is another example of the heterogeneity different opinions? In the end we all have the same goal,
that you see in the discipline. I think what we are going to do now which is to forecast uncertain market prices.
is to try to put it all together. We would We should be able to figure out a way to
BF: You said earlier that one of the like to come up with a systematized learn from each other to do that better.
reasons technical analysis isn’t smiled framework for thinking about technical
on quite as much is probably because patterns. We want to develop tools that JG: That is exciting. Thank you very
there isn’t one single thing that every- will allow technicians to develop new much for your time, Andrew.
body can agree on. Is it possible that patterns and analyze old patterns in a
there are just so many ways that you new light, in a systematic way. And I SUGGESTED READING
can make money that so long as you have been working on neurophysiologi- Kirkpatrick, Charles, and Julie Dahlquist
have a discipline and follow it, that you cal and psychological foundations of [2006]. Technical Analysis: The
can probably be successful? technical analysis. There are some very Complete Resource, FT Press.
There may be some truth to that, but interesting insights that can be derived Lo, Andrew W., and Jasmina Hasanhodzic
as an academic, I can’t help but wonder by comparing the human brain’s func- [2009]. Heretics Of Finance,
what the commonalities are in profit- tions to a computer’s because there are Bloomberg Press.
able trading strategies. For example, some obvious differences between what Lo, Andrew [2008]. Hedge Funds: An
two of the most venerated investors of humans and computers do. Analytic Perspective, Princeton Uni-
our time are Warren Buffett and George versity Press.
Soros, who are wildly different in their JG: Which are? Lo, Andrew, and A. Craig MacKinlay
approach to investing, or so it seems. In One of the biggest differences is pat- [2001]. A Non-Random Walk
fact, I suspect that they share some com- tern matching. Humans are better at Down Wall Street, Princeton Uni-
mon approaches to managing money that pattern recognition than computers. The versity Press.
are worth studying. Similarly, the body of
Copyright (c) Technical Analysis Inc.
 I work hard at my trading and
want a platform that works as
hard as I do. That’s why
I chose NinjaTrader.
It not only lets me execute like
a pro - their Ecosystem,
support and training have
me thinking like a pro too.

I may not be a pro,

but I trade like one.
Professional traders move through the markets with confidence and precision.
NinjaTrader’s extensive video library, partner Ecosystem and live training
webinars provide you with the foundation to become a more confident trader.
Our award winning software lets you execute trades with professional
precision, and the best part is that getting started with NinjaTrader is FREE.
So what are you waiting for? Start trading like a pro today!
ninjatrader.com

decisionbar.com topdogtrading.com ampclearing.com

Featured Ecosystem Partners | to view more go to ninjatrader.com/partners

 How do you make sense
of stock ratings?
Like this.
Bullish
8.1

Equity Summary Score by StarMine

It translates the different analyst opinions and rating systems out there into one
simplified stock score.
• Combines up to 10 independent analyst ratings
• Weighted by historical analyst accuracy
• Considers how commonly or rarely analysts issue a given rating
• Easily compares stocks within or across industries using a standardized
“bullish, neutral, or bearish” sentiment

Equity Summary Score. One more innovative reason serious investors are choosing Fidelity.

Get 200 free trades* today and

explore your next investing idea
with Fidelity.

800.343.3548 | Fidelity.com/trading

Investing involves risk, including the risk of loss.

The Equity Summary Score is provided for informational purposes only, does not constitute advice or guidance, nor is it an endorsement or recommendation,
and is not a guarantee of future performance for any particular security or trading strategy. The Equity Summary Score is provided by StarMine, an independent
company not affiliated with Fidelity Investments. For more information and details, go to Fidelity.com.

*Trades are limited to online domestic equities and options must be used within 90 days. Options trades are limited to 20 contracts
per trade. Offer valid for new or existing Fidelity customers opening or adding to an eligible Fidelity IRA or Fidelity brokerage
retail account. Accounts receiving more than $100,000 of net new assets will receive 200 free trades. An account balance of $50K
must be maintained for at least 9 months; otherwise normal commission schedule rates may be retroactively applied to any free
trade executions. See Fidelity.com/200free for further details.

© 2014 FMR LLC. All rights reserved. Fidelity Brokerage Services LLC, Member NYSE, SIPC 595749.5.0
Take Control of Your Trading with the
Professional Traders’ Starter Kit  ™

*Based on the 5-year subscription rate of $299.99. Shipping extra for foreign addresses.

The Foremost Collection For Traders 4. Traders.com Advantage™, premium website content
This massive collection packages the best tools for trading and delivering real-world technical analysis to you! You’ll get five
investing in any market! years of charts, indicators, and “how-to” advice for specific
markets, currencies, stocks, and commodities; near-term
1. Technical Analysis of Stocks & Commodities, the Traders’
opportunities; price movement; new techniques. Posted in
Magazine™. The premier magazine for technical analysis.
real-time with an archive of thousands of articles.
You’ll get five years — 65 issues — including our annual
Bonus Issues with our Readers’ Choice Awards. 5. Working Money™, the Investors’ Magazine online. You’ll get
five years of market observations; explanations of charts,
2. S&C Digital Edition. Recent complete issues available in their
markets, and market sectors; money management; and
entirety as PDFs for you to either download or read directly
interviews with money people that will help you trade and
in your browser. No more waiting for the mail to deliver your
invest wisely. Articles added several times a month.
magazine! You will still receive the printed magazine unless
you opt for a digital-only subscription. 6. Article Code. Download or copy & paste code presented in
past issues of Stocks & Commodities — no need to type it
3. Complete Digital Archive. The complete archives as PDFs
out manually.
— more than 17,000 pages — from Technical Analysis of
Stocks & Commodities from 1982 through the present. The 7 . Optimized Trading. The optimized indicator values can be
articles can be read in your browser or download to your used as starting points when trying to decide what values
computer (or any device with Internet to input into your charting software. Search for a certain
access and the ability to read a PDF)! symbol or company or build your own portfolio.

ALSO RECEIVE A FREE✝ BOOK

WITH YOUR PTSK ORDER!
Charting The Stock Market: The Wyckoff
Method, edited by Jack K. Hutson, is yours Online: www.TRADERS.com
free✝ with your order for the Professional Toll free: 1-800-832-4642
Traders’ Starter Kit. 206-938-0570 • Fax: 206-938-1307
✝
Shipping and handling charges apply for foreign orders.
Email: Circ@Traders.com
Join us on Facebook at www.facebook.com/STOCKSandCOMMODITIES Follow us on Twitter @STOCKSandCOMM