Let's Get Technical
Ninety percent of science fiction is crap, but then ninety percent of everything is crap.”
-Theodore Sturgeon
>
Yesterday morning, Portfolio's Felix Salmon had a post "Adventures in Technical Analysis, Jim Cramer Edition."
I cannot put aside the fact that Cramer is not, and has never been, any sort of technician. We do not dismiss medicine because accountants cannot do open heart surgery, so it seems kinda odd to use a bad technical call of Cramer's -- an admitted non-technician -- as proof that technicals are worthless ("astrology!").
But what really caught my attention were the following paragraphs, which amount to the standard criticism of Technical Analysis:
"They all do it: even much smarter and much more analytical traders like Barry Ritholtz do it too. Do what? Resort to "technical analysis", which is the art of drawing lines on charts and extrapolating from them what the market is going to do next.
Whenever you hear words like "overbought" or "oversold" or "momentum" or "support" or "resistance", it means that whatever you're hearing is garbage. But it also means that the person you're listening to has no idea what's about to happen, and is therefore resorting to the financial equivalent of astrology. In such cases, it's worth ignoring the message completely, but it's also worth having some serious thoughts about the messenger, too."
There are so many different ways to take this down, its hard figuring out where to start. Let's begin with a definition of what technical analysis is not:
Technicals are not magic. They are not a way to forecast the future, nor are they a guarantee of future profits. They are not based on someone's estimate of what future earnings might be, nor do they require you to guesstimate management's skill set or presume the desirability of a new product. Pure Technicians don't even listen to conference calls or even talk to management.
Technicals can be, however, far less squishy than fundamentals. Technicians use the data that is generated by the markets itself: Price and volume to start, then many other data points and derivatives thereto.
From this basic data, there are many variations of Technical Analysis:
• Trend followers believe markets exhibit persistence, mostly due to big institutional purchasers. This leads to buying uptrends and selling or shorting downtrends. John Henry and Richard Dennis are classic examples of trend follwoers
• Quants use a variety of mathematical data points. A goods example is the fund Renaissance Technologies.
• Contrarians use Sentiment data to determine when markets have moved to far in one direction or another. The goal is to anticipate a major reversal. See Jason Goepfert as a prime example.
• Pattern recognition traders look for various pictures -- pennants, flags, cup & handles, head & shoulders, etc. I find that this last form of TA -- Pattern Recognition -- to be especially prone to Sturgeon's law above.
There are lots of other types of technicals, but this isn't meant to be an exhaustive survey (feel free to discuss Elliot Wave, Fibonacci, and other forms of technicals in comments).
The question as to whether technicals "work" or not is actually framing the wrong issue. There is as much Art as Science in the application of TA. That some people are lousy technicians proves only that it requires skill.
A better question to ask is "What information do charts and related data provide, and how can this be used by investors and traders?"
I posit that, when used appropriately, charts and data can provide tremendous insight:
-Provides a statistical approach to investing, one that describes the probabilities of various outcomes (versus making predictions)
-Charts show you if we are in a bull or bear market, allowing you to manage risk appropriately;
-Trends can keep you away from the wrong sectors (Housing, Autos, and Finance are obvious examples) or keep you in the right sectors (eg., Energy and Ag)
-Developing good risk/reward analyses;
-Tracking what the institutions are doing;
-Identifying specific stocks that might be appealing;
The bottom line is that TA is merely a tool, albeit one used more skillfully by some than others.
Finally, consider this question: If you could look at one and only one source before buying your next stock or fund, which would you choose: a fundamental analyst's report (with no charts in it), or any chart of your choosing? While I like having access to both, I cannot ever imagine buying something without first looking at the chart . . .
Note: We previously addressed these issues a few years ago in Tracking the Elephants (Part I and Part II).
~~~
Tuesday, June 24, 2008 | 10:00 AM | Permalink
| Comments (60)
| TrackBack (0)
add to de.li.cious | 
digg this! | 
add to technorati | 
email this post
TrackBack
TrackBack URL for this entry:
http://www.typepad.com/services/trackback/6a00d8341c52a953ef00e5536c89868833
Listed below are links to weblogs that reference Let's Get Technical:
Comments
Is it not ironic, technically speaking, that the market is right at the March "lows" and a two day Fed meeting begins?
It's almost as if they have planned it this way. Hard to wrap my head around but interesting nevertheless.
Also I noticed that a Fed meeting has been scheduled for approx. 1 full week before the Nov. election.
One more thing....that SEC filing for LEH that was done on Sat. is such a joke:
"terms TBD".....how they get away with that is what is really the "crisis"....
Ciao
MS
Posted by: michael schumacher | Jun 24, 2008 10:07:59 AM
TA is most useful IMO for risk management. Stops, targets, and entry points. The way I see it...if people are buying and selling based on any certain strategy - in a significant amount - then you better pay mind to it. Even if its astrology lol
Posted by: Zo | Jun 24, 2008 10:16:31 AM
Let's be honest. The guy is a writer, correct? People who can, do. There are very few, yes this is a compliment Barry, that do both well.
Posted by: Steve Colglazier | Jun 24, 2008 10:18:21 AM
Charts are useful primarily as time savers. Without looking at any financials and possessing only common knowledge about their businesses, if someone were to show me a chart of Apple and General Motors, I could safely reach all sorts of conclusions about what is going with truck sales and Mac sales. And I'd be right. Why? Because markets are largely efficient. None of us is so smart that we can afford to ignore what everyone else is seeing.
Posted by: Eric | Jun 24, 2008 10:18:58 AM
It is not that everything is "crap" but everything with TA and fundamentals look nowdays to be as deriding as a golfer searching his holes on a rugby field and asking rugby men for a direction and his tee.
Posted by: Philippe | Jun 24, 2008 10:26:03 AM
Anyone who dismisses TA is either blind or stupid. As a pure technician, I can tell you it works. I primarily use elliot wave theory and Fibbonacci anlysis. I also like to look at japanese candlesticks, RSI divergence as well as some long term Moving averages.
You think it's a coincidence that the S&P 500 precisely held the 38.2% retracement on on the MLK holiday scare?
You think it's a coincidence that the 10 yr bond precisely held the 61.8% retrace in June 2007?
You think it's a coincidence that the S&P500 failed at 1440 when the 20 mo. moving average came in at 1438? (The 20 mo m.a. is a powerful longer term average)
I could list hundreds of examples of fibbonacci retracements holding and elliot wave analysis predicting turning points in various markets. The 10 yr bond this decade, in particular, has been a tutorial in EW theory and fibbo's.
- AT
Posted by: Andy Tabbo | Jun 24, 2008 10:31:25 AM
I personally like the slow stochastic combined with the MACD. If an oversold sector or grouping is underwater on the MACD, and it's only in an unpopular phase, then I see potential gold. At that point, you have to rely on education, experience, and luck. And sometimes patience before it takes off again. But it usually takes off big time.
Right now, a lot of the market is in a blood on the street phase. I'm betting on oil retreating in a reasonably short time ... maybe 90 days or less. I also don't think the market is going much lower unless oil spikes again significantly. This is an enhanced summer malaise that will pass when oil reverses. Until then, it will be pretty depressing, though. These are generally the best times to start looking around for good buys.
Posted by: cinefoz | Jun 24, 2008 10:57:18 AM
>>You think it's a coincidence that the S&P 500 precisely held the 38.2% retracement on on the MLK holiday scare?>>
I think it was a coincidence and it's just like a party that everyone is invited to. Why go if everyone else does to. The system knows what technicians look for and COUNT on them buying those levels. If enough of them do them mission accomplished however the actual meaning or reason has nothing to do with it.
YOu think that the last 15 minutes was purely technical??? erasing a 100 point loss (almost)at the March "lows" Not a chance...
Ciao
MS
Posted by: michael schumacher | Jun 24, 2008 11:02:43 AM
Technical analysis is an excellent tool, used to model what has occurred, and/or what may occur (or be occurring). As BR paraphrased in an earlier post, ALL MODELS ARE WRONG, strictly speaking. Detractors of TA never seem to notice that any successful trading strategy utilizing TA is a combination of the technicals with a heavy dose of risk control. So when your initial analysis goes bad, the technical will tell you to get the hell out!
Posted by: HCF | Jun 24, 2008 11:03:16 AM
Barry: Another thing that charts do is level the playing field. In other words, we all know that the price chart should contain all the information that is known about a stock; all information should be reflected in the stock price. So as a small trader/ investor, with the price chart I have the same information as the "big boys". When it comes to fundamental information, this isn't the case; I cannot talk to management or get on a plane and go and look at operations, etc.
Now where I think technicians and others who poo - poo TA get in trouble is that they don't do their homework. They don't define their strategies or they don't follow their own analysis. They don't take the trouble to findout what works and what doesn't. Most stuff is just dogma.
Posted by: GM Lerner | Jun 24, 2008 11:04:06 AM
A comparable quote from "Adventures in Fundamental Analysis" would sound like something like this:
"Whenever you hear words like “missed earnings target” or “strong buy” or “buy-out rumour” or “the CEO just bought stocks” or “the yield curve inverted” or "multiple expansion", it means that whatever you're hearing is garbage. But it also means that the person you're listening to has no idea what's about to happen, and is therefore resorting to the financial equivalent of astrology. In such cases, it's worth ignoring the message completely, but it's also worth having some serious thoughts about the messenger, too."
Posted by: Jay | Jun 24, 2008 11:08:35 AM
Technical analysis can be great at risk management, the exercise of choosing a stop-loss price level at which your trade idea will be "proven wrong" is one of TA's greatest virtues.
Fundamental analysis is just about useless without analyzing the expectations about the fundamentals. Fundamental analysis too often emphasizes buying a low valuation and selling a high valuation when the investor would potentially be better served by making an assessment of whether they are buying low expectations or selling high expectations....which ultimately would bring them back to asking themselves questions about any potential for mean reversion in whatever fundamental data series on which they may be basing their decisions.....
price action is pure, price action is king
Posted by: I just want to be right | Jun 24, 2008 11:29:41 AM
Do "fibbo's" work by some arcane law of nature or do they work because there are enough manipulators with enough money to make them work whenever they want? I wonder how well those "fibbo's" would have worked this morning if somebody had dropped a bomb on Goldman Sucks right at DJIA 11,725.
Posted by: Tom F. | Jun 24, 2008 11:31:36 AM
IMO, read properly, technical indicators can be superior to sentiment surveys because they can tell you what investors are actually doing rather than what they say they are doing -- e.g., an average true range says more about stock sponsorship than a survey does -- but none of it is really worth arguing about because anything that gives you an edge when you're making money rather than losing it is okay and telling everyone exactly how you deploy that edge is usually not (okay) so folks like Salmon who don't actively engage in the game have little chance of ever seeing the whole picture because most of what anyone is going to tell them is ...well, refer back to Sturgeon.
I'd guess the main point is, can a connection between a technique or principal and the real-world actions of humans be made? I personally don't see how that can be done with a fairly big hunk of economic theory (the assumption of a rational actor for example) as well as most techniques including many that are considered quantitative or even fundamental so I don't use them or refer to them unless they are very popular and I want to understand what other investors are basing their decision-making on.
OTOH I don't think it's a stretch to see supply/demand for an equity or sector at play in a simple long-term moving average, or growing investor excitement in an accelerating RSI, MACD or Stochastic, or, for that matter, support and resistance levels as the psychological commitment to the price range where a significant number of investors bought an asset or sold it.
It cost me a lot to learn how to make more good calls than bad ones within the larger context of risk and money management and I don't waste much time defending it. If I was trying to sell something I suppose it would be another matter -- it's never a good idea to forget that when it comes to money everyone talks their own book -- but otherwise it's just different strokes for different folks and Salmon is perfectly welcome to his opinion: Seems to work for him and has no impact on me ...unless he should find himself on the other side of my desk in which case I would be delighted to eat his lunch while telling him it was really a personalized diet plan ...hmmmm, think I could charge a fee for that too (hee, hee).
Posted by: RW | Jun 24, 2008 11:44:29 AM
A. Felix Salmon arguments regarding TA show he really does not know what he's talking about. B. He has no clues as to what astrology really is either.
Posted by: OldBrokenRecord | Jun 24, 2008 11:46:10 AM
Tom F.-
I think you are closer to what is actually happening at this point. We both know that "they" would never do that in this environment so the actual reasons mean very little to me other than how I've opined above about counting on buyers at those levels.
It's a rigged game at this point. Less so in the past.
WHat "happy news" are we going to get out of the Fed at the March "lows" at the start of it's meeting.......that's really what I want to know. Don't get me wrong.....Fib. #'s have it's own merit (used them many times) but decidedly less so as a reliable and true indicator in this heavily "helped" market.
Ciao
MS
Posted by: michael schumacher | Jun 24, 2008 11:47:50 AM
I hate to agree with Cramer, I've done very well shorting lenders and homebuilders whenever he idiotically recommended them over the last couple years, but I totally agree with the astrology analogy.
But I've come to the conclusion that since so many fools trade on their technical analysis (read horoscope), it's necessary to pay attention to it when trading short-term.
Posted by: Bob_in_MA | Jun 24, 2008 11:50:50 AM
The power of technical analysis is undeniable in today's market, but I believe it is much more attributable to the amount of money being managed with a technical slant.
15 yrs ago, most TA was little more than chart reading and very little money was managed by TA. I was a fundy analyst at a bulge bracket and frankly most of the salesmen and major funds thought the technical analysts were quirky and funny, but not to be taken seriously.
Today there are huge funds being run with a technical bias. All TA may be B/S, but if everyone's buying and selling according to the same B/S, then it's hard to deny it's importance to your bottom line.
Posted by: Brian | Jun 24, 2008 11:52:39 AM
Charts and TA are just indications of the past and are therefore completely useless to determine the future.
Posted by: John Borchers | Jun 24, 2008 12:01:24 PM
Hi Barry,
With regards to Cramer, it isn't at all that he made a bad call that is the problem, rather it is the fact that he made the bad call on one show and a week later he claimed that he had never recommended the stocks that he had recommended only one week earlier. He even went as far as to say that people who had bought the stocks that he had recommended one week earlier were not paying attention to his show because he had never recommended such foolishness. There is a video of the two shows - juxtaposing him saying "buy! buy! buy" (for mortgage companies, banks, etc) followed by one week later where he denies ever making those recommendations. So the problem is that he is a liar, not that he is doing technical analysis. Cramer should have said "I screwed up last week, and I am sorry for recommending banks, mortgage companies and the other screwed up investments and I will try to do better". Instead he told his audience that they weren't listening to him if they had been stupid enough to buy the very stocks he recommended. Lying is the problem, not technical analysis.
Posted by: Tim | Jun 24, 2008 12:19:17 PM
There is so much academic research supporting the validity of technical analysis it is embarrassing at this point to be unaware of it. Most people in the field do not call it "technical analysis" but rather momentum, or pattern recognition, or tactical asset allocation, or behavioral finance.
After all, the most famous factor model in finance (Fama) uses TA.
Andy Lo has a new book coming out on the subject:
"The Heretics of Finance: Conversations with the Leading Practitioners of Technical Analysis"
Posted by: WorldBeta | Jun 24, 2008 12:24:58 PM
Brian above makes a good point.
Put all other arguments aside and the "self-fulfilling prophecy" angle still plays. Early in my investing/trading journey I also thought TA was voodoo b.s. These days I am almost purely a technical trader. In many trades I don't know anything about the company other than the ticker.
Posted by: Alaskan Pete | Jun 24, 2008 12:25:42 PM
Listen to this recent speech by Andrew Lo at MIT, "Technical Analysis: An Academic Perspective":
http://event.on24.com/clients/default/presentation/default.html?titlecolor=000000&eventid=110220&sessionid=7&username=&partnerref=&format=rmmulti&key=28A828822AD18023FD5FDB8F5F090792&text_language_id=en&playerwidth=880&playerheight=640&eventuserid=16993198&contenttype=A&mediametricsessionid=15075148&mediametricid=293267&usercd=16993198&mode=launch
Posted by: WorldBeta | Jun 24, 2008 12:30:36 PM
I think it's amazing how seldom the technical vs fundamental argument resolves into overall strategy and timing, in the sense that a financial planner rather than a trader understands the terms. It would be foolish, for example, to take a position you only intend to hold for a few days based solely on fundamentals. And it is (probably) foolish to make a ten-year buy-and-hold investment based entirely on technicals.
The main problem I have with the constant media coverage of technicals is that for most people, stocks are mostly in a retirement account and buy-and-hold is probably a better strategy, and the technical stuff is just kinda distracting.
Over the long run, business are driven by the fundamentals. But technicals can tell us a lot about the timing. If you're gonna be actively trading, you need to be aware of them.
Posted by: JBL | Jun 24, 2008 12:35:58 PM
Query: Have we spent too many of our limited minutes debunking Salmon's piece?
Posted by: Eric | Jun 24, 2008 12:36:23 PM
