• Quick note - the problem with Youtube videos not embedding on the forum appears to have been fixed, thanks to ZiprHead. If you do still see problems let me know.

Stock Market Technical Analysis

roger said:
You are right. If people are just doing random ****, then yes, they'll more or less cancel each other out.

But we are not positing random ****, we are positing a system that actually, demonstrably works.

Furthermore, if the # of systems is nearly infinite, what are the odds of finding a system that actually, demonstrably works?

I leave the latter as an exercise for the reader.
Click to expand...

I've done the calculations. I come up with a number that is indistinguishable from zero.
 
roger said:
But we are not positing random ****, we are positing a system that actually, demonstrably works.
Click to expand...

We aren't. You are. Last warning before you go on ignore.

You keep insisting that others have a position that they do not and then writiing essays attacking that position.
 
I think you have to use fundamental and technical analysis. I want to get the CFA/CMT designations. I will do CFA next year. I'm studying for it now.
 
Technical analysis works fine as long as everyone uses your technical analysis. It's herd behavior at it's finest (and by finest I mean lemmings off a cliff).
 
roger said:
From Security Analysis, 1951 edition, page 652:
Quote:
"1.Chart Reading Not a Science That chart reading cannot be a science is clearly demonstrable. If it were a science, its conclusions would be as a rule dependable. In that case everybody could anticipate tomorrow's or next week's price changes, and hence everyone could make money continuously by buying and selling at the right time. This is patently impossible. A moment's thought will show that there can be no such thing as a scientific prediction of economic events under human control The very "dependability" of such a prediction will cause human actions that will invalidate it.... It follows that there is no generally known method of chart reading that has been continuously successful for a long period of time. If it were known, it would be speedily adopted by numberless traders. This very following would bring its usefulness to an end."


I don't think anything has changed in the ensuing 60 years to change Graham's or jimbob's conclusions.
Click to expand...

Of course you are free to believe whatever you want. Let me just state this clear. There are a number of assumption in this quote that are, at best, untested. First of all, I agree in that chart reading is not deterministic science, it is obtuse to even believe this is true. I agree in that it is comparable to astrology, to put an example.

But the second frase is very weak, (that its conclusions would be rule dependable for it to be science). Sometimes science is based on probabilities and not straight, causal and direct facts. A planet might follow an easy set of rules but an electron certainly does not.

His third frase is not any better, it naively assumes that if someone developed a working probabilistic based system EVERYONE would have access to it and then use it. This is absurd.

I could go further, but it doesn't worth my time.
 
Last edited:
Bodhi Dharma Zen said:
But the second frase is very weak, (that its conclusions would be rule dependable for it to be science). Sometimes science is based on probabilities and not straight, causal and direct facts. A planet might follow an easy set of rules but an electron certainly does not.
Click to expand...
Science based on probabilities gives both possible predictions and their likelyhoods. Therefore the conclusion becomes rule dependable if we use a sufficiently large sample size.

His third frase is not any better, it naively assumes that if someone developed a working probabilistic based system EVERYONE would have access to it and then use it. This is absurd.
Click to expand...
It's absurd to believe a succesful method can be kept secret, because a) so many people try to find it, and b) entire books are written about methods of the world's most succesful investors.
 
It seems that there is this all-or-nothing fallacy where because TA is not 100 percent effective at prediction that it is a useless tool (drkitten, et al).

I think that is plain silliness. Fundamental Analysis is not 100 percent effective either! Every stock has risk. I am a fundamentals guy myself. I observe other people using TA as a tool, but for my kind of investing it just isn't very helpful.

One of the earliest anecdotes I heard when learning finance in college, was from a professor who had become part of a cabal. The cabal had used TA and data mining to come up with a foolproof formula for predicting stock. They pooled their money and got ready to buy on the predictor date. On the morning they were going to place their order, the CEO of the company was arrested for financial fraud. She said they had not yet bought in. Could have been a made up anecdote to scare students. However, the point was made. There is no panacea.

The TA guys who aren't flaming zealots seem to understand that there are factors beyond TA and that there are downsides to TA. They also seem to understand that if a stock gets near 15 dollars, sells off, near 15, sells off, near 15, sells off, that it is probably not a good idea to buy in at 15. Crazy guys, right? Of course, they aren't investors, they are traders. Trading isn't my bag.
 
Last edited:
roger said:
We haven't discussed it much here, but there is an excellent book on the topic:

Evidence-Based Technical Analysis: Applying the Scientific Method and Statistical Inference to Trading Signals

Summary: 6400 TA methods were tested and found not to produce statistically significant returns.

Which is not to say that there is no possible TA rules that will produce significant returns. But it all reads like voodoo to me.
Click to expand...

This book has an excellent overview and lay explanation of how testing enough technical trading rules will yield some that appear to have statistical significance. Taleb touches upon this when he looks at groups of funds that have been selected as having produced good returns for, say, 5 years and why the actual return the next year will be highly unlikely to be similarly high.
 
corplinx said:
One of the earliest anecdotes I heard when learning finance in college, was from a professor who had become part of a cabal. The cabal had used TA and data mining to come up with a foolproof formula for predicting stock. They pooled their money and got ready to buy on the predictor date. On the morning they were going to place their order, the CEO of the company was arrested for financial fraud. She said they had not yet bought in. Could have been a made up anecdote to scare students. However, the point was made. There is no panacea.
Click to expand...

This anecdote suggests that it would have worked otherwise but that there were "events my dear boy, events" which upset the system.

I would argue that the most mechanistic interpretation of a generalised market that has a reasonable realism is of a chaotic system. I would also argue that it is a general feature of chaotic systems that you can't look at past patterns and use statistics to reliably predict what will happen. People did try that with weather forcasts, and beyond the level of, "red sky at night - shepherd's delight," the success was very low.
 
egslim said:
Science based on probabilities gives both possible predictions and their likelyhoods. Therefore the conclusion becomes rule dependable if we use a sufficiently large sample size.
Click to expand...

No. We don't have a clue regarding some fields, all we have are "possible outcomes", "range of probabilities for this and that to happen". Every sufficiently complex system will make our simplistic cause and effect models useless. Hint: stocks behavior is one of such systems.

egslim said:
It's absurd to believe a succesful method can be kept secret, because a) so many people try to find it, and b) entire books are written about methods of the world's most succesful investors.
Click to expand...

Wrong again. As I have explained before, there are countless of possible systems based on approximately the same rules. It is not sufficient to "know" that a particular trading system uses "RSI and candle sticks patterns" you can use the same on different stocks and never accomplish the same results.

Trading, for the few of us who take it seriously and approach it with an skeptical mind, is an hypercomplex subject, there are countless variables involved and you should know where to look, and what to look. As I have said before, I do not base my system on any of the current methods available in books or around the net, and yes, I use some TA as a base to take some decisions.. and yes, it is useful for me.
 
Last edited:
jimbob said:
This anecdote suggests that it would have worked otherwise but that there were "events my dear boy, events" which upset the system.
Click to expand...

The anecdote is what it was. The market is larger than charts and trends.

The fact that their system did not work was evidence that it failed. Not that it would have worked except for that dastardly CEO.

The point of the anecdote is that there are un-chartables. There is risk that is not empirical.
 
corplinx said:
It seems that there is this all-or-nothing fallacy where because TA is not 100 percent effective at prediction that it is a useless tool (drkitten, et al).

I think that is plain silliness. Fundamental Analysis is not 100 percent effective either! Every stock has risk. I am a fundamentals guy myself. I observe other people using TA as a tool, but for my kind of investing it just isn't very helpful.
Click to expand...
No, that is not the claim. The claim is that no reasonable evidence exists that TA provides excess returns compared to the market. That means over multiple transactions. There is no requirement, or expectation that any investment strategy is successful 100% of the time. However, any strategy that returns less than the market is worthless, since it would be less risky, and return more, to just buy the appropriate index (S&P, or whatever). It is in that sense that we say an investment strategy doesn't "work".

As for using TA as a tool, again, studies show that TA does not allow you to buy in at a lower price, sell at a higher price, identify trends, etc. As in the last paragraph "does not allow" means over a large enough set of transactions. Naturally, I am open to a counter example. However, I will point out this has been studied extensively, and no consistently profitable schemes have been identified. Some studies have shown slightly profitable returns in things such as the carry trade. However, these studies are strongly contested. I note that previous studies that showed positive returns in the stock market subsequently failed once new data came in. It is easy to data mine and find correlations. There are also recent interesting papers suggesting the possibility of extracting useful information from price data. But let us be clear, no conclusive evidence has ever been provided that it actually does work. Period. You cannot claim it does work. I happily concede that some sophisticated forms of TA will be shown to work in the future. However, the forms touted in the books, on forums, in newsletters, etc., have been studied, and they do not generate excessive returns.

The TA guys who aren't flaming zealots seem to understand that there are factors beyond TA and that there are downsides to TA. They also seem to understand that if a stock gets near 15 dollars, sells off, near 15, sells off, near 15, sells off, that it is probably not a good idea to buy in at 15. Crazy guys, right?
Click to expand...
Well, yes they are. Again, studies show that there is no profitable way to identify and capitalize on trends like these. For example, see Griffioen Technical Analysis in Financial Markets.
 
roger said:
Again, studies show that there is no profitable way to identify and capitalize on trends like these. For example, see Griffioen Technical Analysis in Financial Markets.
Click to expand...

According to wikipedia, 56 out 95 studies disagree with your assertion.

As for me, I'm a Buffetologist. I have no skin in the game. I just dispute factual errors like yours.
 
corplinx said:
According to wikipedia, 56 out 95 studies disagree with your assertion.

As for me, I'm a Buffetologist. I have no skin in the game. I just dispute factual errors like yours.
Click to expand...
If you read the paper:
[FONT=Myriad Roman, Arial, Helvetica, Sans-serif;]Among a total of 95 modern studies, 56 studies find positive results regarding technical trading strategies, 20 studies obtain negative results, and 19 studies indicate mixed results. Despite the positive evidence on the profitability of technical trading strategies, most empirical studies are subject to various problems in their testing procedures, e.g. data snooping, ex post selection of trading rules or search technologies, and difficulties in estimation of risk and transaction costs. Future research must address these deficiencies in testing in order to provide conclusive evidence on the profitability of technical trading strategies.[/FONT]
Click to expand...
[FONT=Myriad Roman, Arial, Helvetica, Sans-serif;]

As I said, "[/FONT]There are also recent interesting papers suggesting the possibility of extracting useful information from price data. But let us be clear, no conclusive evidence has ever been provided that it actually does work. Period."

With that said, of course my statement was wrong. Studies have not shown there is no profitable way to make money, they have just failed to conclusively find that method. Entirely different things.

I suggest reading at least part of of the Griffioen paper.For example,
Hence, in short, after correcting for sufficient transaction costs, risk, data snooping and out-of-sample forecasting, we conclude that objective trend-following technical trading techniques applied to the AEX-index and to stocks listed in the AEX-index in the period 1983-2002 are not genuine superior, as suggested by their performances, to the buy-andhold benchmark. Only for transaction costs below 0.10% technical trading is statistically profitable, if the best strategy is selected by the Sharpe ratio criterion.
Click to expand...
I don't know anyone who pays 0.1% taxes. If we ignore taxes, 0.1% is hard to achieve even for a floor trader, who has control over the bid/ask spread and pays the lowest possible transaction costs. Okay, I guess if you are a floor trader, wily, and trading your personal IRA while paying your firm's rates, you can get a bit of coin in the Amsterdam market, for now, for certain stocks. Opps, there's the currency exchange risk, that one gets blown out of the water. Yes, you can use the forex market to neutralize the risk, but then you've just added to your transaction costs, and don't ask me the IRA rules on that little strategy.

But rather more interesting is when you look at the numbers behind this. The upshot is that he finds strategies that work, but only for specific stocks. i.e. telecom stock XYZ has a successful strategy, but fertilizer maker ABC doesn't. The study itself gives us no information on why this may be. I find it very interesting. Search the whole world for securities, commodities, currencies, cut transaction costs to the bone, completely ignore taxes, and you can find narrow profits on some things. Some times.

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.

edit: I find those 0.1% transaction fee requirement particularly galling. It should tell you how tiny the effects they are finding. Let's suppose for a moment that it's real, not just data mining. We know that markets shift, and that techniques that previous papers found that work no longer work when we look at new data. My hypothesis is that it is bull caca, but let's assume that the big money has implemented automated trading that arbitraged the real advantage away. So, in either case we have the risk of our strategy not working. How much added risk does it take to kill that 0.1%? The papers do not say. We do know that we have no way to detect that the strategy is not working until you've taken quite a beating. So, I contest the claims for profit after adjusting for this risk.
 
Last edited:
Bodhi Dharma Zen said:
No. We don't have a clue regarding some fields, all we have are "possible outcomes", "range of probabilities for this and that to happen". Every sufficiently complex system will make our simplistic cause and effect models useless. Hint: stocks behavior is one of such systems.
Click to expand...
Bolding mine.

And that's exactly my point: The TA-model is useless.

Because, as you say, we don't have a clue, ergo no scientific theory, so no conclusions that are rule dependable.

Note that:
A scientific theory need not be generally applicable, but the range of situations to which it applies does need to be specified.
Probabilistic conclusions are fine, because they can always be narrowed down with a sufficiently large sample size, but they do need to be better than chance.

As I have said before, I do not base my system on any of the current methods available in books or around the net, and yes, I use some TA as a base to take some decisions.. and yes, it is useful for me.
Click to expand...
I never considered myself modest, but if I believed I had discovered a system that was both superior to others, and unknown to most in my field, I would seriously consider myself delusional.

By definition very few people are that exceptional, and the probability that one of us in this thread is, is very small.
 
You must log in or register to reply here.

Back
Top Bottom