Wednesday, October 25, 2017

Unmasking the VooDoo: Chart Patterns

If you are not a charting aficionado you might think that trying to predict future price movements from past price movements is pure folly. And guess what? You would be right.

What? Did I just say that? Well, yes I did but there is more to it than the simple answer. We cannot predict price movements but we can assign porbabilities to what might happen based on what already happened.

"What's the difference?" you ask.

It's a big difference. If you say you can predict what will happen you are gazing into an orbuculum. That's a crystal ball, for you lay folk.

What technical analysts do is gauge future possibilities based on the footprints the elephants leave in the butter. Here is the phrase that pays:

People tend to act in a similar manner when faced with similar situations.

Chart patterns are those similar situations. The "forecast" is the similar manner of action. Nowhere does it say "will" or "must." 

Let me pull a large quote from my favorite author with that phrase embedded.

"Chart patterns are formed by the buying and selling actions of people and people tend to act in a similar manner when faced with similar situations. Some blame this on a self-fulfilling prophecy: if enough investors believe in the significance of a chart pattern ending then they will act, and thus their actions will assure that the assumed result will occur. "

- A Beginner's Guide to Charting Financial Markets: A practical introduction to technical analysis for investors by Michael Kahn.

Wow, that guy makes sense!  Let's continue:

"But investors may not act the same way this time and the charts will tell us when that is the case quickly before losses begin to mount." 

That implies that technical analysis does not expect to be right all the time. If it did, then probabilities would be 100%.  Therefore, trying rate the efficacy of technical calls just by the success rate is a failed methodology.  I challenge you to show me the price/earnings ratio at which a buy signal will always result in profits. Or a percentage of orders or inventory that is 100% right when making a buy or sell call on a stock.

Any single technical analysis-based call can fail. However, string enough of them together with stop losses and re-evaluation at achieved price targets and the Cajun chef will ga-ron-tee a nice positive rate of return.  

What is a pattern, anyway?

Charts are just graphical representations of the tape. They show us how price changes over time and the forms they take actually do tell us a bit about what was going on in the market when they formed. Does the shape of an EKG tell the doctor about what is going on in your heart right now? And can the doctor then "predict" whether you will have heart issues in the future? 

Charts form based on the buying and selling actions of people. And, they really only have three flavors:
  1. They can be pauses for rest in an ongoing trend
  2. They can be pauses as the market decides it wants to go the other way
  3. They can be pauses when the market just has no clue which way it wants to go
Number 1 is called a continuation pattern. Number 2 is a reversal pattern. And number 3 can be a trading range or just a directionless market filled with either uncertainty or apathy.

Patterns have tops and bottoms and they are indeed called resistance and support. Typically, when prices move through one or the other the trend continues in that direction. Something happened to trigger a mood change from the rest or uncertainty of a pattern to the conviction and movement of a trend.

Fortunately, markets have inertia, just as in physics. If they are in motion (trending) they tend to stay in motion. And if they are at rest they tend to stay at rest - both until acted upon by a force.

That force can be truly from the outside, such as news, earnings and legal matters. Or it can be inside the market when sentiment changes or a related stock or index breaks out from a pattern. In either case, pattern becomes trend and it will stay that way until something makes trend turn into pattern.

How do patterns form?

I am not going to get into the nuances of all the things that happen in patterns with volume, momentum and the shapes of various patterns here. Those things give us additional clues as to which way the market can break out but they are not needed to understand just how a pattern forms.

When a pattern first starts to form, the trend hits an insurmountable speed bump. Keeping it simple with a rising trend, it pulls back, finds a new support and rallies again. But it is stopped at the top of the pattern, which now for the first time starts to actually appear. The cycle happens a few more times.

What is happening now is the bulls see a cheap price at the pattern bottom. The bears see an expensive price at the top. Both act accordingly. Nobody has enough nerve to do anything else.

In the case of a triangle or any other pattern with sloping tops of bottoms, the bulls tend to sell out earlier within the pattern. The bears tend to close shorts earlier. Uncertainty builds as neither side is willing to let the market move to their desired prices. They take action earlier and earlier.

Finally, one side finds a reason to push their agenda. For an upside breakout, the bulls no longer sell out at the pattern top. In fact, they buy more. And away we go!

Different pattern shapes are just theme and variation. It is the same dynamic going on in all of them. However, pattern shapes do give us more clues as the eventual resolution.


Chart patterns help us decide what to do based on subjective probabilities that the market will do something somewhat specific once the patterns are broken. The market does not do X just because it broke out. The pattern breaks because the market decided it was time to move. And it lets us know about how far it will go be the way it formed the pattern in the first place.

It is our job to read these signals and do one of the only three things we can do - buy, sell or hold. We do not know exactly where the market will go and how long it will take to get there. However, based on the signs, we will get a very good idea if we should ride that tiger with a lot of money or with a little or forget about it completely.

Nobody cares if you predicted a market target. Only your wallet cares that you made it fat.

Wednesday, October 11, 2017

Future of Capitalism

In another round of file cleaning on the old computer I found my notes taken when reading The Future of Capitalism by Lester Thurow. The book was written in 1996 and I read it in 1998.

Wikipedia - Lester Carl Thurow (May 7, 1938 – March 25, 2016) was an American political economist, former dean of the MIT Sloan School of Management.

Actually, this is kind of spooky. I made (in 2017) comments in parens places.

Surging inequality (a big political talking point)
China focus will change to India focus (not so much)
Communism had good education (don't know what this meant)
No competitor to capitalism now so it will not implode (true dat)

Economic Surface of the Earth

End of Communism (Plate 1)   (well, Cuba pushed that way a little and North Korea will if they push war)
Offers lots of cheap labor (another political talking point

Migration (plate 4) pushes unskilled workers into 1st world (ditto)

New technologies (plate 2) generates skill shift (in spades)
Need to offset this (by massive skill investment) is resisted by the elderly (plate 3)

Global economy (plate 4) forces wages down (indeed)

Without dominant power (plate 5) there is no economic locomotive. (better watch out)

Progress occurs when technology meets ideology

Inflation is extinct (true dat, too)

Japan will crack (it is number 3)
Europe is biased towards social programs so jobs will flee (Nostradamus)
Global bubbles burst together (looks like that could happen very soon)
Uncertainty breeds religious fundamentalism (its familiar in uncertain times) (O...M...G...)

Capital is human and no longer ownable

Need to spur long-term thinking and R&D - Government's only role  (we wish)

Bill Meehan in the archives

I was looking through old documents buried on my computer and found this raw text excerpts that I believe eventually made it's way into the Market Technician's Association newsletter in 2000 or 2001. It was comments about a recent MTA seminar written by Bill Meehan, whom we lost in the Towers on 9/11.
-------------------- comments on the Seminar

excerpted with permission from a column written by Bill Meehan.

Voodoo, as I use it here, refers to only one of the colorful descriptors that many on Wall Street use to denigrate those who spend their waking hours devoted to the art of technical analysis. Kooks, I believe, is how academia prefers to label chart readers. However, nothing was quite so satisfying as listening to Dr. Andrew Lo of the MIT Sloan School of Management present mathematical proof at the Market Technicians Association's annual meeting that technical analysis cannot be dismissed, at least when it comes to pattern recognition, as mere voodoo practiced by kooks.

It's good to have one's belief verified by such an acclaimed professor. The work continues, especially as to why the results were so significant when looking at Nasdaq stocks and so inconclusive when looking at those traded on the New York Stock Exchange and the Amex. Perhaps readers might have some ideas after reviewing the data.

Oops! I noticed that I made a big mistake above. Having the chance to meet and spend quite a bit of time with Dick Arms, who so graciously fills in for me, was actually the highlight of the meeting for me, with all due respect for Dr. Lo and his paper. And meeting many other leaders in the field was a real thrill that alone is worth the modest cost of joining the MTA.

Passing on Wisdom

I've been a bit out of touch with the market virtually all of last week, without many of the technical tools that I rely on. Therefore, I'll relay a bit of what another presenter at the MTA meeting had to say about the current technical state of the market. In addition, I'll throw in my two cents about the Federal Reserve and some of the psychological indicators that keep me from becoming more bullish in the short term. As time allows this week, I'll also do my best to respond to readers who sent me email while I was away.

While most of the meeting's presenters were very interesting (although a few were speaking a foreign language to my mathematically challenged brain), Phil Erlanger's observations about short interest struck me as particularly important. The MTA's current president focused on adjusted short interest, a sentiment tool that he developed, which has a very strong track record. The Erlanger short interest ratios for the Dow, S&P 500 and Nasdaq Composite are all at multi-decade lows and at levels far below those generally seen at times of high market risk. The charts were startling and alone should make one pause, although relying on one technical indicator isn't wise, in my opinion.

Bill Meehan is the chief market analyst for Cantor Fitzgerald, a Manhattan-based institutional trading and research firm, and writes daily for the Cantor Morning News. He can be reached at