Stock Market Technical Analysis

[jimbob]

I hear that alot from people who go to casinos. They have a foolproof system to win at blackjack, and it would have worked last night if only that one guy had not held on 14, or that other guy hadn't left the table early, or....

It is completely different...

Otherwise a dart throwing chimp could beat the market about 50% of the time...

...Oh

 

[oggiesnr]

So did all the people who have all these great systems foresee the current mess and get out before they lost or are they like most of us nursing our losses?

Steve

 

[The Central Scrutinizer]

So did all the people who have all these great systems foresee the current mess and get out before they lost or are they like most of us nursing our losses?

Steve

They all got out right before the market went down.

 

[corplinx]

And all of this is far outside the realm you (and the OP) are talking about - a stock picker using TA either exclusively or as a tool - these studies are mining a hundred years of data across many different markets, testing hundreds to thousands of combinations of rules, and only finding some stocks that yield at best very slight profits, if you ignore transaction costs. It's just not been shown to be a feasible technique for the average investor, and people claiming otherwise really need to present some evidence. Just buying a corrected dataset will set you back a hundred K or so. If Bodhi is running a supercomputer and accessing data subscribed to by his university, who knows, he may be on to something besides luck. This being JREF, I await the evidence, and don't put much stock in the claims, especially when you look at the literature.

That's awesome. Firstly, using TA as a tool for picking stocks, I don't claim to be a proponent of and even urge against. I assume most studies on the subject of TA are regarding this behavior.

As for TA as a part of normal DD (FA, TA, googling the darn stock, reading the industry news, reading the stocks yahoo message board....), it is intrinsically part of DD and I don't see why I need to offer evidence that the following statement is not a bad idea:

"I plan on buying stock in Foo corp. I looked up its earnings. I looked up its book value. I think it is a great value. They issue monthly dividends. The current price is 15 dollars a share, but it drops down to 14 often in this sideways market. Based on the performance of the sector it is in, I think it is undervalued as has room to grow. Also, I think their new services I've found in the news look promising. I plan to buy Foo corp."

Now, if I took the "15 dollars a share, but it drops down to 14 often in this sideways market" part of the DD, and put it on a pretty chart. You would claim its part of that whole TA thing that doesn't work.

Your base position is silly. Much of TA is just a pretty chart on volatility or pricing and volume. Are you also against excel spreadsheets when they contain normal stock due diligence? Pie charts? How about html encoded documents?

Frankly, your position is half this all-or-nothing fallacy and the other half seems to be that of a luddite. I'm not really interested in having you sit back and say "show me the evidence" when I think you have the situation reversed.

 

[egslim]

The current price is 15 dollars a share, but it drops down to 14 often in this sideways market.

Correction, it has often dropped to 14 in the sideways market until now.

There's no reason to assume the market will remain sideways, but even if it does, there's no reason to believe Foo corp's shareprice will continue its behaviour, merely because it did so recently.

By itself it's useless information to base your decision on, much using like the colour of a car to judge its safety-features.

 

[corplinx]

Correction, it has often dropped to 14 in the sideways market until now.

There's no reason to assume the market will remain sideways, but even if it does, there's no reason to believe Foo corp's shareprice will continue its behaviour, merely because it did so recently.

By itself it's useless information to base your decision on, much using like the colour of a car to judge its safety-features.

Correction. This is not roulette and we aren't betting on red because its been black 9 times in a row. I used to work at the HQ for the worlds' largest casino company. I know how that fallacy works.

Stock prices aren't magical nor the result of die throws. They are based on what people offer to sell them for and what other people bid to buy them for. If people in say the past month haven't been willing to pay more than 15 bucks for the stock, the TA person calls this "resistance". ZOMG, what a woo concept right? Now, does that mean the price will never go above 15? Of course not.

However, in lieu of corporate earnings reports, PR, big news, etc, and overall market up/down trends, a reasonable person would surmise that if people aren't willing to buy for more than 15 dollars, that to get a better bargain you should buy in at less than 15. A reasonable person can make this judgement and still realize that (the dow could jumper another 1000, the president of Foo corp could be arrested for pig jumping, Foo corp could announce a new iPhone killer). You really have to quit it with the all-or-nothing fallacy already.

 

[drkitten]

You really have to quit it with the all-or-nothing fallacy already.

I don't think anyone except you is using this "all-or-nothing" fallacy.

We're using the "expected value is positive or not" line of reasoning, which is not at all fallacious.

What, for example, does "but it drops down to 14 often in this sideways market" tell you that "Mercury is in Pisces" doesn't?

And, in particular, what is the expected value of that additional knowledge? And does it exceed transaction costs?

 

[jimbob]

corplinx,

So why is the approach to forcasting this stock price any different to trying to forecast weather by pattern recognition?

Except that the feedback systems in share prices are even mre nonlinear, as the rules can change as people suddenly feel optimisic or pessimistic, and can factor in what they think other people will be thinking about that stock...

"I plan on buying stock in Foo corp. I looked up its earnings. I looked up its book value. I think it is a great value. They issue monthly dividends. The current price is 15 dollars a share, but it drops down to 14 often in this sideways market. Based on the performance of the sector it is in, I think it is undervalued as has room to grow. Also, I think their new services I've found in the news look promising. I plan to buy Foo corp."

 

[corplinx]

I don't think anyone except you is using this "all-or-nothing" fallacy.

We're using the "expected value is positive or not" line of reasoning, which is not at all fallacious.

What, for example, does "but it drops down to 14 often in this sideways market" tell you that "Mercury is in Pisces" doesn't?

And, in particular, what is the expected value of that additional knowledge? And does it exceed transaction costs?

Its all or nothing because you assign nil value if it doesn't work as a completely accurate picking/prediction system (which no one here is arguing that it is).

 

[drkitten]

Its all or nothing because you assign nil value if it doesn't work as a completely accurate picking/prediction system (which no one here is arguing that it is).

No. We assign nil value to it because it has not been shown to generate any value.

I understand the idea of expected value. If I knew that a particular roulette wheel NEVER came up 13, I could make a packet betting on evens, since I win more often than I lose.

On the other hand, if I knew that a particular roulette wheel NEVER came up 13, that woudn't help me at all if I were at the craps table.

And if I believe in error that it never comes up 13, I'll actually lose money betting on evens.

If you think that technical analysis generates a positive expected value, show us how.

 

[corplinx]

No. We assign nil value to it because it has not been shown to generate any value.

56 out of 95 studies show positive results. I don't have a breakdown of that number further. I don't think it would be going out on a limb to assume that they mean TA for picking versus TA as a part of DD. TA as a part of DD would be much harder to run a study on.

Well, in short. You're wrong.

 

[GreyICE]

56 out of 95 studies show positive results. I don't have a breakdown of that number further. I don't think it would be going out on a limb to assume that they mean TA for picking versus TA as a part of DD. TA as a part of DD would be much harder to run a study on.

Well, in short. You're wrong.

If I flipped a coin 95 times and it came up heads 56, you'd say that coins come up heads when you flip em?

I find your methodology poor.

 

[roger]

Correction. This is not roulette and we aren't betting on red because its been black 9 times in a row. I used to work at the HQ for the worlds' largest casino company. I know how that fallacy works.

Stock prices aren't magical nor the result of die throws. They are based on what people offer to sell them for and what other people bid to buy them for. If people in say the past month haven't been willing to pay more than 15 bucks for the stock, the TA person calls this "resistance". ZOMG, what a woo concept right? Now, does that mean the price will never go above 15? Of course not.

However, in lieu of corporate earnings reports, PR, big news, etc, and overall market up/down trends, a reasonable person would surmise that if people aren't willing to buy for more than 15 dollars, that to get a better bargain you should buy in at less than 15. A reasonable person can make this judgement and still realize that (the dow could jumper another 1000, the president of Foo corp could be arrested for pig jumping, Foo corp could announce a new iPhone killer). You really have to quit it with the all-or-nothing fallacy already.

All evidence disagrees with you.

I like how you completely ignore the evidence.


And what's with the snittiness?

 

[roger]

56 out of 95 studies show positive results. I don't have a breakdown of that number further. I don't think it would be going out on a limb to assume that they mean TA for picking versus TA as a part of DD. TA as a part of DD would be much harder to run a study on.

Well, in short. You're wrong.

Yes you do have a breakdown. I gave it to you. It's in the paper. You are ignoring the evidence, and getting snitty about people who actually read the papers.

It's 56 positives with bad methodology, with such a slim margin that > 0.1% transaction rates kills the 'profit', with no analysis of the risk you are taking.

But hey, ignore this (repeated) information and insult me. You seem to get some joy out of it for some odd reason.

 

[roger]

Its all or nothing because you assign nil value if it doesn't work as a completely accurate picking/prediction system (which no one here is arguing that it is).

Why do you say this? I explicitly explained how TA is judged to work or not, and it is not all or nothing. I told you this expressly because you made this claim before. I point you to post #76.

 

[roger]

Correction, it has often dropped to 14 in the sideways market until now.

There's no reason to assume the market will remain sideways, but even if it does, there's no reason to believe Foo corp's shareprice will continue its behaviour, merely because it did so recently.

By itself it's useless information to base your decision on, much using like the colour of a car to judge its safety-features.

Quite. If the poster you corrected actually read the study I referenced, he would see that in fact such a simple technique does not generate information with a positive expectation value. They have to use quite robust sets of signals, randomly generated by a computer, selected with the Sharpe ratio, tested with the RC and SPA algorithms, and then forward tested with data outside the data set. Even then, you only get some stocks in some markets to show positive expectation, with a requirement that transaction costs are less than 0.5%.

In short, no, looking at past prices does not allow you to say anything useful about the future prices.

 

[Tippit]

I've gotten involved on another message board in a discussion on "technical analysis" of stock prices. (Here if you're interested).

If you're not familiar with technical analysis or chart analysis, it's a school of thought among stock market traders who think that they can predict the movement of the stock market or of individual stocks by looking for certain patterns in a chart of past prices, without knowing anything at all about the company.

I love technical analysis. It amplifies volatility, and therefore opportunities for those of us who use FA. Zodiac-based stock picking is fantastic too. Bring it on

 

[Francesca R]

As for TA as a part of normal DD (FA, TA, googling the darn stock, reading the industry news, reading the stocks yahoo message board....), it is intrinsically part of DD and I don't see why I need to offer evidence that the following statement is not a bad idea: [ . . . ]

I wouldn't go as far as to tell someone that technical analysis is a bad idea. I wouldn't really deign to tell anyone how to invest.

The last few months before I left my job in March, managing a fund at a bank, my (recently installed) boss (who I had little time for) was very big on "Let's use anything that works . . . I'm a sceptic but show me backtests and I'm in . . . What's wrong with marketing lit that says we use all available information and have a diversified approach? . . . It can't have negative value-added can it; why rule things out?" He wanted me to reinstate fixed income and FX momentum models that I've worked with in the past. This came after a period (July 2007 through November 2008, AKA the meltdown) where my fund's performance had been extremely strong using no technical analysis, and in stark negative correlation to almost everything else that was happening at the bank, and in the overwhelming majority of the hedge fund business. To my mind he was staring in the face something that was fixed (and quite victorious) and wanting to break it. It was an "icing on the cake" reason why I (and my co-manager) resigned. Now they can do what they like. Good luck to them.

I am now part of a start-up fund and to date our AUM only consists of my, and other partners' money (although we are talking to former clients), but we are not going to be using technical analysis and our material is quite concrete about that.

 

[Francesca R]

for those of us who use FA.

I don't recommend using ****-all.

 

[drkitten]

If I flipped a coin 95 times and it came up heads 56, you'd say that coins come up heads when you flip em.

In particular, 56/95 is not statistically significant or distinguishable from "no result at all." (The actual p value is 0.0501 for a one-sided test (which is not justifiable) or 0.1002 for a two-sided test (which is justifiable).

That's elementary probability theory.

If I tried to publish a paper with a p value greater than 0.10, I'd get laughed at. I guess corplinx has less critical standards.