A Random Walk Down Wallstreet
October 28, 2008 6:10 AM Subscribe
Another stock market post. Technical traders (or charters) don't look at the fundamentals of an investment, like the earnings per share or even macro economic indicators to understand how a stock will move. Instead, they look at the movements in the market. While many of us might think prices are now low, technicians have the reassurance of fibonacci sequences, relative strength indexes, support levels and other "blackbox" ratios to determine their investments. There are some who blame them for a lot of woe, but they also provide a ray of light when everything else looks glum. Despite some evidence it doesn't work, or at least doesn't work over the long term, the number of true believers in the market mean even true fundamentalists can't ignore their impact.
Investment firms employ analysts to build "black boxes" or trading models, for big money, but in market conditions like now, many models fall down.
Investment firms employ analysts to build "black boxes" or trading models, for big money, but in market conditions like now, many models fall down.
Lol.
Imagine a class of people who invested completly randomly. With absolutly no basis for any of their decisions. You would expect them, on average over time to do about as well as the market. Some doing better, and some doing worse.
The application of this to 'technical traders' is left as an exercise to the reader.
posted by delmoi at 6:34 AM on October 28, 2008
Imagine a class of people who invested completly randomly. With absolutly no basis for any of their decisions. You would expect them, on average over time to do about as well as the market. Some doing better, and some doing worse.
The application of this to 'technical traders' is left as an exercise to the reader.
posted by delmoi at 6:34 AM on October 28, 2008
Actually these people just have better connections to the Hylean Theoric world.
posted by delmoi at 6:38 AM on October 28, 2008 [3 favorites]
posted by delmoi at 6:38 AM on October 28, 2008 [3 favorites]
The basis of technical trading could be applied to selecting winning lottery numbers.
posted by Xoebe at 6:43 AM on October 28, 2008 [1 favorite]
posted by Xoebe at 6:43 AM on October 28, 2008 [1 favorite]
Restate my assumptions:
One: Mathematics is the language of nature.
Two: Everything around us can be represented and understood through numbers.
Three: If you graph the numbers of any system, patterns emerge.
Therefore, there are patterns everywhere in nature. Evidence: The cycling of disease epidemics; the wax and wane of caribou populations; sun spot cycles; the rise and fall of the Nile.
So... what about the stock market?
posted by rokusan at 6:49 AM on October 28, 2008 [3 favorites]
One: Mathematics is the language of nature.
Two: Everything around us can be represented and understood through numbers.
Three: If you graph the numbers of any system, patterns emerge.
Therefore, there are patterns everywhere in nature. Evidence: The cycling of disease epidemics; the wax and wane of caribou populations; sun spot cycles; the rise and fall of the Nile.
So... what about the stock market?
posted by rokusan at 6:49 AM on October 28, 2008 [3 favorites]
Hylean Theoric world
I knew I'd start seeing Anathem refs on MeFi pretty soon. Unfortunately I listened to the audiobook, so I don't know how to spell a lot of the most important stuff.
posted by DU at 6:52 AM on October 28, 2008
I knew I'd start seeing Anathem refs on MeFi pretty soon. Unfortunately I listened to the audiobook, so I don't know how to spell a lot of the most important stuff.
posted by DU at 6:52 AM on October 28, 2008
I've made a lot more money trading on technicals than fundamentals. Just sayin'. Anyone who thinks Wall St is a true random walk is nuts. That said, Fibo stuff and Eliot Wave stuff is complete f'n voodoo. Check out slopeofhope.com if you want to peanut gallery some of this stuff.
posted by unSane at 6:56 AM on October 28, 2008
posted by unSane at 6:56 AM on October 28, 2008
Technical analysis is astrology for traders, but lots of smart people believe in it. I think it taps very effectively into flaws in the way our brains are wired to make decisions. For example, the fact that someone made more money using technicals than fundamentals isn't evidence that technicals work, but I can see how it would be hard for them not to arrive at that conclusion.
There is something to the fact that if enough people believe something (even for the wrong reasons), it can have an effect on the market.
posted by diogenes at 7:05 AM on October 28, 2008
There is something to the fact that if enough people believe something (even for the wrong reasons), it can have an effect on the market.
posted by diogenes at 7:05 AM on October 28, 2008
Great post. I'm favoriting it because I'm learning a great deal about finance from MeFi. Keep up the good work.
posted by up in the old hotel at 8:56 AM on October 28, 2008
posted by up in the old hotel at 8:56 AM on October 28, 2008
Some market cycles are 4 generations trough-to-trough
posted by Rafaelloello at 9:13 AM on October 28, 2008
posted by Rafaelloello at 9:13 AM on October 28, 2008
Fair comment.
And if you'd bought QID this morning at RSI 30 (oversold) and sold at 70 (overbought), you'd have picked up 10%. And you could have done this every day for the last two weeks -- one trade a day, 5-10% move.
posted by A-Train at 9:15 AM on October 28, 2008
And if you'd bought QID this morning at RSI 30 (oversold) and sold at 70 (overbought), you'd have picked up 10%. And you could have done this every day for the last two weeks -- one trade a day, 5-10% move.
posted by A-Train at 9:15 AM on October 28, 2008
With commodities and foreign currencies hitting 10 year lows in two weeks, I'm seriously disillusioned with the ability of the stock exchange to determine prices. Isn't there a way to restrict the number of speculators in the market? Why can't we let only the people actually interested in purchasing and selling goods haggle over price, and leave the weasels out of it? Hell, even central planning doesn't get out of whack this much.
posted by Popular Ethics at 9:17 AM on October 28, 2008
posted by Popular Ethics at 9:17 AM on October 28, 2008
Funny you mention QID. That is my major holding, though I'm beginning to worry about counter-party risk so I'm using QLD puts to add to my position.
posted by Rafaelloello at 9:17 AM on October 28, 2008
posted by Rafaelloello at 9:17 AM on October 28, 2008
This approach to trading seems a bit like trying to predict the next great play Shakespeare might have written based only on analysis of the distribution patterns of iambs, trochees and spondees in Macbeth. I mean, this kind of analysis might be a useful way to identify an entry point before delving into a more thorough, substantive analysis of the factors contributing to price movement over time, but it seems like such a shallow approach to understanding what's actually driving market trends--it confuses the signifier with the signified (or the map for the territory, in other words).
posted by saulgoodman at 9:38 AM on October 28, 2008
posted by saulgoodman at 9:38 AM on October 28, 2008
"This approach to trading seems a bit like trying to predict the next great play Shakespeare might have written based only on analysis of the distribution patterns of iambs, trochees and spondees in Macbeth."
Yeah, it would seem that way, but the market does not move on Shakespeare's opinion or any other single person's. It does have an odd herd mentality to it, especially of late. When the market turns, everything goes at once. Recently, it has tended to stampede in one direction or the other.
posted by A-Train at 9:45 AM on October 28, 2008
Yeah, it would seem that way, but the market does not move on Shakespeare's opinion or any other single person's. It does have an odd herd mentality to it, especially of late. When the market turns, everything goes at once. Recently, it has tended to stampede in one direction or the other.
posted by A-Train at 9:45 AM on October 28, 2008
You can have technical trading or you can have market efficiency. Figure out which religion you believe in more.
posted by GuyZero at 9:48 AM on October 28, 2008
posted by GuyZero at 9:48 AM on October 28, 2008
it confuses the signifier with the signified (or the map for the territory, in other words).
Unless you believe that the market discounts everything, in which case the map contains all the information about the territory.
The first time I really saw TA work was when I saw a mining stock dribble through a long-term support level. I shorted it. A couple of days later they posted news that their latest hole had missed and the stock cratered. In this case there was no voodoo involved. Presumably some of the lab workers or drill guys had seen the hole come up dry and word had seeped out into the market. Since then I've seen the same pattern a few times in exploration stocks waiting for news, and each time it's worked out. Basically, somebody always knows something and that is reflected in the ticker.
The best thing to read about all of this is Jessie Livermore's brilliantly readable REMINISCENCES OF A STOCK OPERATOR (link goes to free PDF of the book).
posted by unSane at 9:52 AM on October 28, 2008 [1 favorite]
Unless you believe that the market discounts everything, in which case the map contains all the information about the territory.
The first time I really saw TA work was when I saw a mining stock dribble through a long-term support level. I shorted it. A couple of days later they posted news that their latest hole had missed and the stock cratered. In this case there was no voodoo involved. Presumably some of the lab workers or drill guys had seen the hole come up dry and word had seeped out into the market. Since then I've seen the same pattern a few times in exploration stocks waiting for news, and each time it's worked out. Basically, somebody always knows something and that is reflected in the ticker.
The best thing to read about all of this is Jessie Livermore's brilliantly readable REMINISCENCES OF A STOCK OPERATOR (link goes to free PDF of the book).
posted by unSane at 9:52 AM on October 28, 2008 [1 favorite]
Lots of people makes lots of money using technical analysis of the financial markets. Sure, the analysis of financial performance and earnings etc (fundamental analysis) is arguably more useful, but assumes a rationality in the decision process that is often lacking for periods of time, for example, now. Why would using statistical tools to analyze the complex behavior of the market be any less valid than the use of similar tools to try to understand physical phenomena?
posted by sfts2 at 9:57 AM on October 28, 2008
posted by sfts2 at 9:57 AM on October 28, 2008
Basically, somebody always knows something and that is reflected in the ticker.
Well, sure, as long as "somebody... knows something," looking at the patterns on the surface can make a sort of sense. But consider this reductio ad absurdum: imagine if there wasn't a single investor looking at anything but the charts. A company could be on the verge of bankruptcy and all the investors would be caught completely off-guard when the share prices abruptly dropped to zero. So trading on the technicals is ultimately just a less direct way of responding to the fundamentals, news events and other external indicators.
Why would using statistical tools to analyze the complex behavior of the market be any less valid than the use of similar tools to try to understand physical phenomena?
Well, for one thing, the behaviors being analyzed aren't meaningful because they don't take the external drivers for the behaviors into account. A statistical analysis of the population growth of panda bears, for example, would also factor in external conditions--the weather, availability of food, territorial bounds, etc. How do you derive meaningful information from the apparent patterns in an undefined domain?
You can describe patterns of behavior down to the Nth level, but unless you're also modeling how external events influence those behavioral patterns, how do you know what they mean? Obviously, no one yet has figured out the magical formula for reading chart patterns--does a double bottom chart always predict a bullish break? Nope. Sometimes there's another bottom in store. Sometimes the stock just plunges. There's too much information excluded from price charts and other purely technical indicators to make them meaningful. It's like trying to understand an argument by looking only at its schematized, propositional form: The argument might be formally valid, but since you can't evaluate the semantic content of its terms, you can't really conclude that the argument is sound.
posted by saulgoodman at 10:35 AM on October 28, 2008
Well, sure, as long as "somebody... knows something," looking at the patterns on the surface can make a sort of sense. But consider this reductio ad absurdum: imagine if there wasn't a single investor looking at anything but the charts. A company could be on the verge of bankruptcy and all the investors would be caught completely off-guard when the share prices abruptly dropped to zero. So trading on the technicals is ultimately just a less direct way of responding to the fundamentals, news events and other external indicators.
Why would using statistical tools to analyze the complex behavior of the market be any less valid than the use of similar tools to try to understand physical phenomena?
Well, for one thing, the behaviors being analyzed aren't meaningful because they don't take the external drivers for the behaviors into account. A statistical analysis of the population growth of panda bears, for example, would also factor in external conditions--the weather, availability of food, territorial bounds, etc. How do you derive meaningful information from the apparent patterns in an undefined domain?
You can describe patterns of behavior down to the Nth level, but unless you're also modeling how external events influence those behavioral patterns, how do you know what they mean? Obviously, no one yet has figured out the magical formula for reading chart patterns--does a double bottom chart always predict a bullish break? Nope. Sometimes there's another bottom in store. Sometimes the stock just plunges. There's too much information excluded from price charts and other purely technical indicators to make them meaningful. It's like trying to understand an argument by looking only at its schematized, propositional form: The argument might be formally valid, but since you can't evaluate the semantic content of its terms, you can't really conclude that the argument is sound.
posted by saulgoodman at 10:35 AM on October 28, 2008
in the final analysis stock price depends on opinions, and opinions can be contagious. stocks are not that different from barbies, baseball cards and cancelled stamps. and although i technical trading is a recipe for disaster, it is obvious that the price of a stock IS a fundamental, just like a P/E ratio (hey, the astute among you might even notice that price is the P part of that equation). there is a lot of INFORMATION in the price of a stock, and there is information in the relative movement of stock prices. if you believe in the efficient market hypothesis, future changes in the price of that stock should be independent of the information in the price of a stock... but im not sure that anyone _really_ believes the market is efficient on the time scale technical traders are concerned with. furthermore, it really cant be efficient on a super short time scale because information transmission and trading cannot occur instantaneously.
posted by mano at 10:44 AM on October 28, 2008
posted by mano at 10:44 AM on October 28, 2008
I actually did an experiment a few weeks ago. I used various technical indicators to generate entry points for the S&P. (I used random entries, Donchian breakouts, and Moving Average crosses, if you're interested). Then I compared the 'edge ratio' of each indicator, which basically measures how good an entry it was over a particular period. I used 50 years of S&P data.
The result was really interesting to me. There were small differences in edge between the different strategies but the only big takeaway was that you should go long in a bull market (as measured by for example, the 50 day MA being above the 200 day MA). Going short in a bear market was not nearly as good a bet, partly because the overall tendency of the market is to go up not down.
Using timing strategies (particularly the MA crosses) in a bull market definitely enhanced returns and reduced risk, while getting out of the market when it turned bearish vastly reduced drawdowns.
For example, an edge ratio of 1 means that an entry point is neutral. By using MA crosses in a bull market and staying out of the bears I could get that up to 1.5 or 1.6. These figures were for pretty long term holds, too, weeks rather than days.
So everyone's a genius in a bull market. Who knew?
posted by unSane at 10:59 AM on October 28, 2008
The result was really interesting to me. There were small differences in edge between the different strategies but the only big takeaway was that you should go long in a bull market (as measured by for example, the 50 day MA being above the 200 day MA). Going short in a bear market was not nearly as good a bet, partly because the overall tendency of the market is to go up not down.
Using timing strategies (particularly the MA crosses) in a bull market definitely enhanced returns and reduced risk, while getting out of the market when it turned bearish vastly reduced drawdowns.
For example, an edge ratio of 1 means that an entry point is neutral. By using MA crosses in a bull market and staying out of the bears I could get that up to 1.5 or 1.6. These figures were for pretty long term holds, too, weeks rather than days.
So everyone's a genius in a bull market. Who knew?
posted by unSane at 10:59 AM on October 28, 2008
You can have technical trading or you can have market efficiency. Figure out which religion you believe in more.
Thing is, there can be more than one dynamic at play here.
Greed/fear swings do apparently (to me) swing the market to & fro.
FWIW, I think fibonacci stuff is bullshit since any sufficient number of lines on the graph will intersect something or come close enough, but the EW approach does have a thesis that bears analysis and understanding, if not belief.
posted by troy at 11:05 AM on October 28, 2008
Thing is, there can be more than one dynamic at play here.
Greed/fear swings do apparently (to me) swing the market to & fro.
FWIW, I think fibonacci stuff is bullshit since any sufficient number of lines on the graph will intersect something or come close enough, but the EW approach does have a thesis that bears analysis and understanding, if not belief.
posted by troy at 11:05 AM on October 28, 2008
Gee. Fundamentals don't matter in the stock market. The major driving force of the American Economy has no regard for quality, fiscal responsibility or good management. No matter how hard they try to sell you otherwise, it has long been all speculation, no investment in anything real.
posted by wendell at 11:11 AM on October 28, 2008 [2 favorites]
posted by wendell at 11:11 AM on October 28, 2008 [2 favorites]
unSane The first time I really saw TA work was when I saw a mining stock dribble through a long-term support level. I shorted it ... and the stock cratered.
Have you ever shorted a stock after it broke a long-term support and then it didn't crater? Would you take that to mean that TA didn't work? One time I bought stock on a whim and it went up. Does that mean buying stock on a whim works?
posted by diogenes at 11:17 AM on October 28, 2008
Have you ever shorted a stock after it broke a long-term support and then it didn't crater? Would you take that to mean that TA didn't work? One time I bought stock on a whim and it went up. Does that mean buying stock on a whim works?
posted by diogenes at 11:17 AM on October 28, 2008
One thing that causes me to cock an eyebrow is when the TAers draw support lines all the way back into the 90s, eg. comparing this week's 825 low with the 2002 800 low or the mid-90s plateau of 650.
M3 money supply was around $4T in 1996, and its' $14T+ today.
posted by troy at 11:31 AM on October 28, 2008
M3 money supply was around $4T in 1996, and its' $14T+ today.
posted by troy at 11:31 AM on October 28, 2008
M3 money supply was around $4T in 1996, and its' $14T+ today.
Which is the trouble with TA IMO: there's always more data. Imagine drawing those price charts over a curved surface that reflects money supply. It's non-euclidean technical analysis which is so awesome it moves into the pure bullshit zone. Per my previous comment, I go with market efficiency as TA looks too much like staring into your own navel to discover the secrets of the universe.
posted by GuyZero at 11:35 AM on October 28, 2008
saulgoodman - I have to disagree with your statement "the behaviors being analyzed aren't meaningful because they don't take the external drivers for the behaviors into account." The external drivers are taken into account through the price and volume, money flow, and relative strength behaviors. Not everyone, nor even perhaps a majority of market players utilize technical analysis, basically you have a portion of the players, watching the other players. I do not believe TA is the best approach to market participation - I don't even use it much myself, but to say that it is voodoo or astrology is a bit much.
posted by sfts2 at 12:10 PM on October 28, 2008
posted by sfts2 at 12:10 PM on October 28, 2008
sfts2: to say that it is voodoo or astrology is a bit much.
Why? They are all pseudosciences.
posted by diogenes at 12:18 PM on October 28, 2008
Why? They are all pseudosciences.
posted by diogenes at 12:18 PM on October 28, 2008
Have you ever shorted a stock after it broke a long-term support and then it didn't crater? Would you take that to mean that TA didn't work? One time I bought stock on a whim and it went up. Does that mean buying stock on a whim works?
I said that was the *first* time, not that it proved anything in and of itself.
Think about it this way. A stock has traded above $10 for a year. It's bounced off $10 several times. Then one day it sinks through $10 to $9.50, even though there is no obvious news driving the move. Is that a useful piece of information or not?
Another stock has never broken $20. Ever. Then one day, in the absence of any obvious driver, it suddenly gaps up above $20 on strong volume. Is that a useful piece of information or not?
According to the fundamentals, the first stock just became better value and the second stock became worse value. But would you really buy the first stock or short the second unless you were hoping for a short term rebound?
That's all TA is. People make it too complicated. It isn't.
As for fundamentals, go look at any Bullboard for a stock that recently hit highs and is now cratering. GCE.TO or WTN.TO are two great examples. They both declined from around $10 to around $1.75 in a matter of weeks. The bullboards were full of fundies saying what great bargains they were, all the way down...a bargain at $8... an incredible bargain at $6... at $4 they were backing up the truck... Meantime the TA guys had all stopped out weeks before as the prices broke support after support after support.
All indicators lag but fundamentals lag horribly.
posted by unSane at 12:30 PM on October 28, 2008 [1 favorite]
I said that was the *first* time, not that it proved anything in and of itself.
Think about it this way. A stock has traded above $10 for a year. It's bounced off $10 several times. Then one day it sinks through $10 to $9.50, even though there is no obvious news driving the move. Is that a useful piece of information or not?
Another stock has never broken $20. Ever. Then one day, in the absence of any obvious driver, it suddenly gaps up above $20 on strong volume. Is that a useful piece of information or not?
According to the fundamentals, the first stock just became better value and the second stock became worse value. But would you really buy the first stock or short the second unless you were hoping for a short term rebound?
That's all TA is. People make it too complicated. It isn't.
As for fundamentals, go look at any Bullboard for a stock that recently hit highs and is now cratering. GCE.TO or WTN.TO are two great examples. They both declined from around $10 to around $1.75 in a matter of weeks. The bullboards were full of fundies saying what great bargains they were, all the way down...a bargain at $8... an incredible bargain at $6... at $4 they were backing up the truck... Meantime the TA guys had all stopped out weeks before as the prices broke support after support after support.
All indicators lag but fundamentals lag horribly.
posted by unSane at 12:30 PM on October 28, 2008 [1 favorite]
Exactly, fundamentals are backward looking and only periodic snapshots at that. Whatever your personal investment philosophy is, its hard to argue that fundamentals are basically meaningless at the current time - from a TRADING standpoint, less so from an investment standpoint.
posted by sfts2 at 1:05 PM on October 28, 2008
posted by sfts2 at 1:05 PM on October 28, 2008
saulgoodman - I have to disagree with your statement "the behaviors being analyzed aren't meaningful because they don't take the external drivers for the behaviors into account." The external drivers are taken into account through the price and volume, money flow, and relative strength behaviors.
Assuming trading action is rational, which it likely isn't, then yes. But I don't disagree completely--my point was more to elaborate on the earlier reductio ad absurdum: technical indicators are just a way of getting information about investor behavior, which is (presumably) a response to something external to price. Otherwise, the price is just chasing its own tail.
Let's say we create a fake stock, assign it ten years worth of arbitrary price data, and then begin trading it alongside a real stock (let's assume both stocks are traded at around the same volume, and nobody questions the authenticity of the fake stock). Say we make all of our trades based purely on technical indicators and chart patterns. Would our predictions for the fake stock be any more or less accurate than for the real stock? My guess is not.
But just to be clear--I'm only talking about the extreme case: the exclusive use of technical indicators. As I said, it makes sense that technical indicators and a little common sense could be put to good use.
posted by saulgoodman at 1:11 PM on October 28, 2008
Assuming trading action is rational, which it likely isn't, then yes. But I don't disagree completely--my point was more to elaborate on the earlier reductio ad absurdum: technical indicators are just a way of getting information about investor behavior, which is (presumably) a response to something external to price. Otherwise, the price is just chasing its own tail.
Let's say we create a fake stock, assign it ten years worth of arbitrary price data, and then begin trading it alongside a real stock (let's assume both stocks are traded at around the same volume, and nobody questions the authenticity of the fake stock). Say we make all of our trades based purely on technical indicators and chart patterns. Would our predictions for the fake stock be any more or less accurate than for the real stock? My guess is not.
But just to be clear--I'm only talking about the extreme case: the exclusive use of technical indicators. As I said, it makes sense that technical indicators and a little common sense could be put to good use.
posted by saulgoodman at 1:11 PM on October 28, 2008
I just like the arbitrary distinction between "technical" trading and "fundamentals" trading. In one case you look at a bunch of numbers and guess if the stock is going to go up. In the other...
Stock is worth what the vast herd of buyers and sellers are willing to trade it for. Unless it pays a regular dividend it has no/none/zero/zilch value outside of that. Period.
The only reason the various bunches of numbers might be interesting is because other people might be basing decisions off of them, and therefore you might be able to predict what the other people are going to do by watching the numbers.
(Note that this technique works best if every in the kingdom is blindly reacting to a particular set of numbers while you, the one-eyed man, are predicting what they'll be doing. Good luck with that)
posted by tkolar at 1:31 PM on October 28, 2008
Stock is worth what the vast herd of buyers and sellers are willing to trade it for. Unless it pays a regular dividend it has no/none/zero/zilch value outside of that. Period.
The only reason the various bunches of numbers might be interesting is because other people might be basing decisions off of them, and therefore you might be able to predict what the other people are going to do by watching the numbers.
(Note that this technique works best if every in the kingdom is blindly reacting to a particular set of numbers while you, the one-eyed man, are predicting what they'll be doing. Good luck with that)
posted by tkolar at 1:31 PM on October 28, 2008
I've used to be a hater on TA for a really, really long time -- calling it reading tea leaves and little more than a mentally challenged 4-year old with a box of crayola crayons. Then I started trading. I don't know a single active trader that doesn't understand that TA on some level isn't bullshit.
As a simple example, support lines (sort of a price firewall, when it's crossed, it'll pass it hard) exist because historically that's where huge volumes of shares were traded, and if it crosses below, it means those who bought shares at that price are now losing money, and it's emotionally hard to be losing money. So those who bought at support are now going to sell en mass, which means those interested in TA are going to see on their squiggly line chart that a support at $20, when it goes to $19, means it may drop significantly below that.
It's true that prices will normalize to fundamentals in the long run, but when prices and fundamentals are going to equalize. We can all agree that it will equalize, but it might be the next hour, or the next year. As the overused quote by Keynes, "In the long run, we're all dead."
posted by amuseDetachment at 1:35 PM on October 28, 2008
As a simple example, support lines (sort of a price firewall, when it's crossed, it'll pass it hard) exist because historically that's where huge volumes of shares were traded, and if it crosses below, it means those who bought shares at that price are now losing money, and it's emotionally hard to be losing money. So those who bought at support are now going to sell en mass, which means those interested in TA are going to see on their squiggly line chart that a support at $20, when it goes to $19, means it may drop significantly below that.
It's true that prices will normalize to fundamentals in the long run, but when prices and fundamentals are going to equalize. We can all agree that it will equalize, but it might be the next hour, or the next year. As the overused quote by Keynes, "In the long run, we're all dead."
posted by amuseDetachment at 1:35 PM on October 28, 2008
Would our predictions for the fake stock be any more or less accurate than for the real stock? My guess is not.
You are assuming that you can fake price data convincingly. A plain random walk is not going to look *anything* like real price data. Real price data has long and short term trends, support levels and so on, and moves in concert with other securities and markets.
This is the big straw man of the anti-TA zealots. If you assume that stock prices are a random walk then of course TA is voodoo. But that assumption relies on all participants in the market pricing in all news perfectly and instantly. If you don't believe that hardline version, then clearly trends can exist as news percolates outwards. If trends exist then some version of TA can be used to identify them and give an edge to market participants who use them.
posted by unSane at 1:38 PM on October 28, 2008
You are assuming that you can fake price data convincingly. A plain random walk is not going to look *anything* like real price data. Real price data has long and short term trends, support levels and so on, and moves in concert with other securities and markets.
This is the big straw man of the anti-TA zealots. If you assume that stock prices are a random walk then of course TA is voodoo. But that assumption relies on all participants in the market pricing in all news perfectly and instantly. If you don't believe that hardline version, then clearly trends can exist as news percolates outwards. If trends exist then some version of TA can be used to identify them and give an edge to market participants who use them.
posted by unSane at 1:38 PM on October 28, 2008
unSane: Yeah, to pile on, I notice that those who dislike TA, tend to have a heavy math (and especially statistics) background. They think the market is just Brownian Motion and you can't predict that stuff. Those that do probably just need to read up on Nassim Taleb's Black Swan or anything by Mandelbrot.
Summary for those with that mindset: You can't apply those models to the market (they are physical equations attempted to fit into social systems). Applying those models creates problems like the subprime and derivatives disaster we face today (in fact, the current crisis is directly caused by that line of thinking). Thinking that market movements are gaussian and can fit in your standard deviation models is going to lose you a LOT of money.
posted by amuseDetachment at 1:45 PM on October 28, 2008
Summary for those with that mindset: You can't apply those models to the market (they are physical equations attempted to fit into social systems). Applying those models creates problems like the subprime and derivatives disaster we face today (in fact, the current crisis is directly caused by that line of thinking). Thinking that market movements are gaussian and can fit in your standard deviation models is going to lose you a LOT of money.
posted by amuseDetachment at 1:45 PM on October 28, 2008
Stock is worth what the vast herd of buyers and sellers are willing to trade it for. Unless it pays a regular dividend it has no/none/zero/zilch value outside of that. Period.
No, no--a stock's real value is a function of the underlying company's future earning potential. A share of stock entitles the owner to participate in the profits the company makes in the form of dividends. That's why the fundamentals are important: You invest in a company that's likely to make big profits, and then when that happens, the company returns a portion of those profits to its shareholders, who are after all the owners of the company, in the form of dividends...
Bwuh-ha-ha-ha! I couldn't even finish saying it with a straight face.
Well. That was the original idea anyway. But now it's all about reading charts and making bets and playing upsides and downsides, and yet, it's still the same as it ever was.
posted by saulgoodman at 1:51 PM on October 28, 2008
No, no--a stock's real value is a function of the underlying company's future earning potential. A share of stock entitles the owner to participate in the profits the company makes in the form of dividends. That's why the fundamentals are important: You invest in a company that's likely to make big profits, and then when that happens, the company returns a portion of those profits to its shareholders, who are after all the owners of the company, in the form of dividends...
Bwuh-ha-ha-ha! I couldn't even finish saying it with a straight face.
Well. That was the original idea anyway. But now it's all about reading charts and making bets and playing upsides and downsides, and yet, it's still the same as it ever was.
posted by saulgoodman at 1:51 PM on October 28, 2008
Thinking that market movements are gaussian and can fit in your standard deviation models is going to lose you a LOT of money.
Right... this is the whole incredible Black-Scholes debacle in a nutshell.
posted by unSane at 2:12 PM on October 28, 2008
Right... this is the whole incredible Black-Scholes debacle in a nutshell.
posted by unSane at 2:12 PM on October 28, 2008
I saw this recently. Apologies for the cut'n'paste but the original has disappeared. Anyhoo:
Once upon a time a man appeared in a village and announced to the villagers that he would buy monkeys for $10 each.posted by unSane at 2:23 PM on October 28, 2008 [10 favorites]
The villagers, seeing that there were many monkeys around, went out to the forest and started catching them.
The man bought thousands at $10 and, as supply started to diminish, the villagers stopped their effort. He next announced that he would now buy monkeys at $20 each. This renewed the efforts of the villagers and they started catching monkeys again.
Soon the supply diminished even further and people started going back to their farms. The offer increased to $25 each and the supply of monkeys became so scarce it was an effort to even find a monkey, let alone catch it!
The man now announced that he would buy monkeys at $50 each! However, since he had to go to the city on some business, his assis tant would buy on his behalf.
In the absence of the man, the assistant told the villagers: "Look at all these monkeys in the big cage that the man has already collected. I will sell them to you at $35 and when the man returns from the city, you can sell them to him for $50 each."
The villagers rounded up all their savings and bought all the monkeys.
They never saw the man or his assistant again, only lots and lots of monkeys.
Now you have a better understanding of how the stock market works.
"The basis of technical trading could be applied to selecting winning lottery numbers."
Only if lottery ticket numbers became more or less likely based on what people thought was going to happen.
When everyone is optimistic/greedy and buying (possibly borrowing money to buy), stocks go up. When everyone is pessimistic/fearful and selling! selling! selling!, stocks go down. And the technical analysts say you can interpret greed and fear in the numbers.
Is there some analogous way in which lottery tickets work?
posted by A-Train at 3:09 PM on October 28, 2008
Only if lottery ticket numbers became more or less likely based on what people thought was going to happen.
When everyone is optimistic/greedy and buying (possibly borrowing money to buy), stocks go up. When everyone is pessimistic/fearful and selling! selling! selling!, stocks go down. And the technical analysts say you can interpret greed and fear in the numbers.
Is there some analogous way in which lottery tickets work?
posted by A-Train at 3:09 PM on October 28, 2008
unSane: "39I saw this recently. Apologies for the cut'n'paste but the original has disappeared. Anyhoo:
Jeez, my mind is going. I found this and emailed it to someone 6 or 10 or more weeks back, but I forgot where I found it. Can someone remind me? Thanks in advance. I'll find out tomorrow when I get to work when I sent it, but I forget where I got it.
posted by Rafaelloello at 4:48 PM on October 28, 2008
Once upon a time a man appeared in a village and announced to the villagers that he would buy monkeys for $10 each."
The villagers, seeing that there were many monkeys around, went out to the forest and started catching them.
The man bought thousands at $10 and, as supply started to diminish, the villagers stopped their effort. He next announced that he would now buy monkeys at $20 each. This renewed the efforts of the villagers and they started catching monkeys again.
Soon the supply diminished even further and people started going back to their farms. The offer increased to $25 each and the supply of monkeys became so scarce it was an effort to even find a monkey, let alone catch it!
The man now announced that he would buy monkeys at $50 each! However, since he had to go to the city on some business, his assis tant would buy on his behalf.
In the absence of the man, the assistant told the villagers: "Look at all these monkeys in the big cage that the man has already collected. I will sell them to you at $35 and when the man returns from the city, you can sell them to him for $50 each."
The villagers rounded up all their savings and bought all the monkeys.
They never saw the man or his assistant again, only lots and lots of monkeys.
Now you have a better understanding of how the stock market works.
Jeez, my mind is going. I found this and emailed it to someone 6 or 10 or more weeks back, but I forgot where I found it. Can someone remind me? Thanks in advance. I'll find out tomorrow when I get to work when I sent it, but I forget where I got it.
posted by Rafaelloello at 4:48 PM on October 28, 2008
I saw it on stockchase.com but a Google search shows it was everywhere a few weeks ago.
posted by unSane at 5:12 PM on October 28, 2008
posted by unSane at 5:12 PM on October 28, 2008
I was thinking it was here or one of my other haunts. There is someone at the company I'm consulting for that has only been in this country less than a year. He has recently transitioned from contractor to employee and I help him daily with a lot of what goes on in America. He was distressed this week that he has lost $118 in his 401(k).
posted by Rafaelloello at 5:30 PM on October 28, 2008
posted by Rafaelloello at 5:30 PM on October 28, 2008
Forgot to mention, that I sent this to him at about the time that he became eligible fro the 401(k)
posted by Rafaelloello at 5:36 PM on October 28, 2008
posted by Rafaelloello at 5:36 PM on October 28, 2008
You guys just don't get it. Let me give you a hint: Time is Money
Which means that the stock market is obviously a moneycube. By knowing the 33-dimensional position of some vertices, you can predict the other vertices exact position, and thus know when to invest and when to pull out the money
posted by qvantamon at 6:09 PM on October 28, 2008 [1 favorite]
posted by qvantamon at 6:09 PM on October 28, 2008 [1 favorite]
The real money is to be made selling black box systems for trading platforms with trading figures based on overfitted backtesting results. I would do it, too, if it didn't make me feel dirty.
posted by unSane at 6:12 PM on October 28, 2008
posted by unSane at 6:12 PM on October 28, 2008
So... what about the stock market?
Survival of the fittest Max, and we got the fucking gun.
posted by A dead Quaker at 6:25 PM on October 28, 2008
Survival of the fittest Max, and we got the fucking gun.
posted by A dead Quaker at 6:25 PM on October 28, 2008
The best thing to read about all of this is Jessie Livermore's brilliantly readable REMINISCENCES OF A STOCK OPERATOR (link goes to free PDF of the book).
This book is cited frequently as an endorsement of technical analysis. Unfortunately, the ironic thing is that Jesse Livermore lost his entire fortune during the Great Depression, and declared bankruptcy.
posted by storybored at 6:55 PM on October 28, 2008
This book is cited frequently as an endorsement of technical analysis. Unfortunately, the ironic thing is that Jesse Livermore lost his entire fortune during the Great Depression, and declared bankruptcy.
posted by storybored at 6:55 PM on October 28, 2008
I didn't post it as an endorsement of TA so much as a narrative of TA.
Livermore made and lost fortunes over and over again. He made a mass of money (more than $100m by some accounts) in 1929 which, as you say, he lost during the depression. However at the time of his death in 1940 he was worth more than $5m.
REMINISCENCES naturally doesn't tell the story of his death. He blew his brains out in 1940 in a New York hotel cloakroom, having suffered from depression for many years, and never having regained his trading confidence after the years of the GD.
Livermore's HOW TO TRADE IN STOCKS is another classic.
And yes, I've been thinking about writing a movie about him for a long time!
posted by unSane at 7:09 PM on October 28, 2008
Livermore made and lost fortunes over and over again. He made a mass of money (more than $100m by some accounts) in 1929 which, as you say, he lost during the depression. However at the time of his death in 1940 he was worth more than $5m.
REMINISCENCES naturally doesn't tell the story of his death. He blew his brains out in 1940 in a New York hotel cloakroom, having suffered from depression for many years, and never having regained his trading confidence after the years of the GD.
Livermore's HOW TO TRADE IN STOCKS is another classic.
And yes, I've been thinking about writing a movie about him for a long time!
posted by unSane at 7:09 PM on October 28, 2008
I've used to be a hater on TA for a really, really long time -- calling it reading tea leaves and little more than a mentally challenged 4-year old with a box of crayola crayons. Then I started trading. I don't know a single active trader that doesn't understand that TA on some level isn't bullshit.
Sign my name to this too. I used to call it 'haruspices'. Then I started trading.
I would no sooner trade without a look at the technicals than I would trade without a look at the fundamentals. "Reminiscences of a Stock Operator" is a great book if you want to learn - it's surprisingly timeless.
posted by ikkyu2 at 10:10 PM on October 28, 2008
Sign my name to this too. I used to call it 'haruspices'. Then I started trading.
I would no sooner trade without a look at the technicals than I would trade without a look at the fundamentals. "Reminiscences of a Stock Operator" is a great book if you want to learn - it's surprisingly timeless.
posted by ikkyu2 at 10:10 PM on October 28, 2008
I just thought I'd offer this: My dad has been trading using nothing but TA for the last eight years or so, helped put me through school, and two years ago my sister was diagnosed with a very, very expensive medical condition. He has zero debt. He takes the television out of his office, reads no news sites. He even tried to teach me the tricks of (the) trade once, but I'm still young and full of dreams.
Anyway, I'm not bringing myself or or my sister's medical issues into this to write about how technical analysis saved my dog's life or something, only to point out that at some point, you can't at the very least, discount it. I thought it was a buncha crap too until he told me he'd been trading this way for ages.
posted by bam at 11:17 PM on October 28, 2008
Anyway, I'm not bringing myself or or my sister's medical issues into this to write about how technical analysis saved my dog's life or something, only to point out that at some point, you can't at the very least, discount it. I thought it was a buncha crap too until he told me he'd been trading this way for ages.
posted by bam at 11:17 PM on October 28, 2008
It also helps to understand how the herd mentality uses TA. For example, there are lots and lots of junior/assistant fund managers on Wall Street that *will* make moves based on what a stock does in the vicinity of the 50 day MA. Also, day traders. I spent many years real-time scanning at 2:30PM what I knew day traders would find later that evening after hours. I would look for their chart setups I knew they would be finding that night, buy at 3:30PM, hold overnight and sell to them at a nice profit the next morning. Like taking candy from a baby.
posted by Rafaelloello at 2:03 AM on October 29, 2008
posted by Rafaelloello at 2:03 AM on October 29, 2008
Is it just me, or is there something unsettling about the fact that everybody is trying to read charts faster than the other guy in order to make some money? It seems so far removed from investing. It's just gambling (with a little more data). It feels unsustainable. It's just a bunch of people watching numbers and guessing what other people are going to do based on the movement of those numbers. I just want to be able to conservatively invest my money for the long term and have it grow faster than inflation. It seems like that's getting harder and harder. I don't want to play this game.
posted by diogenes at 4:02 AM on October 29, 2008 [1 favorite]
posted by diogenes at 4:02 AM on October 29, 2008 [1 favorite]
unSane: I said that was the *first* time, not that it proved anything in and of itself.
I understand. But the second and third time didn't prove anything either. Does there ever come a point where this collection of anecdotes proves anything? I don't think so. And that's the definition of a pseudoscience.
posted by diogenes at 4:16 AM on October 29, 2008
I understand. But the second and third time didn't prove anything either. Does there ever come a point where this collection of anecdotes proves anything? I don't think so. And that's the definition of a pseudoscience.
posted by diogenes at 4:16 AM on October 29, 2008
I understand. But the second and third time didn't prove anything either. Does there ever come a point where this collection of anecdotes proves anything?
I'm not trying to prove anything to you. But you do understand the concept of statistical significance, right?
posted by unSane at 5:22 AM on October 29, 2008
I'm not trying to prove anything to you. But you do understand the concept of statistical significance, right?
posted by unSane at 5:22 AM on October 29, 2008
diogenes, your search for an honest market is quaint. If you want a conservative investment, but high quality bonds or treasuries. Have a long term view. This market today is function of a broken regulatory environment causing too many sellers, generating fear, feeding on itself, and is not an indicator of the usefulness of fundamental or technical analysis. Both are out the window now, with I think TA being slightly more useful at the present time. Do you blame TA for current market conditions? I'm not sure why. It seems like you are looking for something that isn't there. Markets are the ultimate reality. Projecting some ideal or wish onto them is silly. If you don't want to play, then don't - that is a perfectly understandable point of view. Everything that has happened is completely irrational, you're right. TA provides a toolset that allows someone try to make some sense of the behavior of large numbers of irrational people.
posted by sfts2 at 5:25 AM on October 29, 2008
posted by sfts2 at 5:25 AM on October 29, 2008
Is it just me, or is there something unsettling about the fact that everybody is trying to read charts faster than the other guy in order to make some money? It seems so far removed from investing. It's just gambling (with a little more data). It feels unsustainable. It's just a bunch of people watching numbers and guessing what other people are going to do based on the movement of those numbers. I just want to be able to conservatively invest my money for the long term and have it grow faster than inflation. It seems like that's getting harder and harder. I don't want to play this game.
I think it's just you. No doubt short term trading has real similarities to gambling, but that doesn't hinder long term investment. No one is making you play any game. Your difficulty in outpacing inflation has much more to do with deceitful government statistics and reckless expansion of credit. If you buy an index stock, your money grows at the same rate as the economy and any short term fluctuation in a particular stock has a negligible effect. But check out this page to see how difficult it is, and always has been, to grow wealth through an index fund.
posted by BigSky at 5:37 AM on October 29, 2008
I think it's just you. No doubt short term trading has real similarities to gambling, but that doesn't hinder long term investment. No one is making you play any game. Your difficulty in outpacing inflation has much more to do with deceitful government statistics and reckless expansion of credit. If you buy an index stock, your money grows at the same rate as the economy and any short term fluctuation in a particular stock has a negligible effect. But check out this page to see how difficult it is, and always has been, to grow wealth through an index fund.
posted by BigSky at 5:37 AM on October 29, 2008
But you do understand the concept of statistical significance, right?
I don't think statistical significance can be used to show that TA isn't pseudoscience. People have mined large amounts of data and found statistically significant examples of ESP. I don't think that proves that ESP works.
From Wikipedia: Given a sufficiently large sample, extremely small and non-notable differences can be found to be statistically significant, and statistical significance says nothing about the practical significance of a difference.
posted by diogenes at 6:18 AM on October 29, 2008
I don't think statistical significance can be used to show that TA isn't pseudoscience. People have mined large amounts of data and found statistically significant examples of ESP. I don't think that proves that ESP works.
From Wikipedia: Given a sufficiently large sample, extremely small and non-notable differences can be found to be statistically significant, and statistical significance says nothing about the practical significance of a difference.
posted by diogenes at 6:18 AM on October 29, 2008
Do you blame TA for current market conditions?
I don't blame TA. It's a symptom, not a cause. I don't blame people for trying to use it to understand a fundamentally broken and corrupt system. I'm just frustrated that the system is such a mess that TA is seen as the best method of investing.
I do have to play the game if I want to have any kind of financial security. I have a long term plan that includes conservative investments and index funds as part of an asset allocation strategy, but I'm afraid it isn't going to work. I'm not going to beat inflation that way.
posted by diogenes at 6:28 AM on October 29, 2008 [1 favorite]
I don't blame TA. It's a symptom, not a cause. I don't blame people for trying to use it to understand a fundamentally broken and corrupt system. I'm just frustrated that the system is such a mess that TA is seen as the best method of investing.
I do have to play the game if I want to have any kind of financial security. I have a long term plan that includes conservative investments and index funds as part of an asset allocation strategy, but I'm afraid it isn't going to work. I'm not going to beat inflation that way.
posted by diogenes at 6:28 AM on October 29, 2008 [1 favorite]
I don't think statistical significance can be used to show that TA isn't pseudoscience.
That's pretty much how science works, Diogenes.
Hypothesis: TA works
Null hypothesis: TA doesn't do any better than buying and selling randomly.
Test: Make a large number of trades using TA and an equally large number of trades using the null strategy (eg random buys/sells).
Result: Did the TA strategy make more money than the null strategy? If so, what is the probability that the results are simply due to chance (ie a series of lucky trades)?
How else would you go about 'proving' it?
posted by unSane at 6:55 AM on October 29, 2008
That's pretty much how science works, Diogenes.
Hypothesis: TA works
Null hypothesis: TA doesn't do any better than buying and selling randomly.
Test: Make a large number of trades using TA and an equally large number of trades using the null strategy (eg random buys/sells).
Result: Did the TA strategy make more money than the null strategy? If so, what is the probability that the results are simply due to chance (ie a series of lucky trades)?
How else would you go about 'proving' it?
posted by unSane at 6:55 AM on October 29, 2008
TA tells me when to buy stocks with excellent fundamentals. That way, when I sell them at my pre-defined point, I can live to regret having sold too early--In spite of the fact that my returns beat the market.
If I didn't trade short-term like this, I'd loose my shirt, because my attention would wander far far away.
posted by Goofyy at 7:06 AM on October 29, 2008
If I didn't trade short-term like this, I'd loose my shirt, because my attention would wander far far away.
posted by Goofyy at 7:06 AM on October 29, 2008
This is starting to be above my pay grade (my statistics course was 12 years ago), but I'll try.
I think the problem is that when people look at TA, they look into the past. They take a huge set of data, and look for patterns that show a particular strategy would have been effective. When you look at a data set as large as the financial market for 20 or 30 years, you are bound to get some statistically significant "hits." It's more like data mining than verifying the accuracy of a prediction.
To prove that TA works, you would need to pick one TA strategy ahead of time, apply it consistently starting now, and then measure whether or not it worked better than random.
posted by diogenes at 7:15 AM on October 29, 2008
I think the problem is that when people look at TA, they look into the past. They take a huge set of data, and look for patterns that show a particular strategy would have been effective. When you look at a data set as large as the financial market for 20 or 30 years, you are bound to get some statistically significant "hits." It's more like data mining than verifying the accuracy of a prediction.
To prove that TA works, you would need to pick one TA strategy ahead of time, apply it consistently starting now, and then measure whether or not it worked better than random.
posted by diogenes at 7:15 AM on October 29, 2008
The standard way to backtest strategies is that you divide the data into two parts, optimize on one part and test on the other. Or you do a so-called 'walk forward' test where you optimize on N data points, test on the next N, optimize on the next N and so on.
Really there are two kinds of TA. There's pure quant stuff which is easy to code programatically (eg buy on a 50-day breakout, sell on a 20-day breakdown) and there's the charting approach, which is much more hand-wavy but, in my experience, works surprisingly well if you keep it simple.
For example, a classic pattern is a 'descending triangle', where a stock bounces between a particular low price and a descending series of highs. You can, and some people have, analyse collections of descending triangles and determine how many historically break to the upside, how many to the downside, and how far the price typically travels after that.
For the record, I'm another person who started out thinking of TA as complete voodoo (and as I say upthread, I still think some of it is) but it wasn't until I started using TA that I made any money in the markets at all.
posted by unSane at 7:32 AM on October 29, 2008
Really there are two kinds of TA. There's pure quant stuff which is easy to code programatically (eg buy on a 50-day breakout, sell on a 20-day breakdown) and there's the charting approach, which is much more hand-wavy but, in my experience, works surprisingly well if you keep it simple.
For example, a classic pattern is a 'descending triangle', where a stock bounces between a particular low price and a descending series of highs. You can, and some people have, analyse collections of descending triangles and determine how many historically break to the upside, how many to the downside, and how far the price typically travels after that.
For the record, I'm another person who started out thinking of TA as complete voodoo (and as I say upthread, I still think some of it is) but it wasn't until I started using TA that I made any money in the markets at all.
posted by unSane at 7:32 AM on October 29, 2008
Voodoo works when enough people believe it. I think charters should embrace the voodoo label :)
I'm still not going to start using TA though. I don't have enough time anyway. That's part of the problem. Any trading/investing system that requires so much interaction isn't realistic for most people.
posted by diogenes at 7:47 AM on October 29, 2008
I'm still not going to start using TA though. I don't have enough time anyway. That's part of the problem. Any trading/investing system that requires so much interaction isn't realistic for most people.
posted by diogenes at 7:47 AM on October 29, 2008
Markets are the ultimate reality.
Wow, and here all this time I thought the ultimate reality was made out of quantum foam and the operation of certain physical laws and all sorts of, you know, kind of real and ultimate-y things. Ultimate reality as markets is much tidier (and avoids all the messy business of explaining why gravity is a weaker than expected force).
posted by saulgoodman at 7:54 AM on October 29, 2008
Wow, and here all this time I thought the ultimate reality was made out of quantum foam and the operation of certain physical laws and all sorts of, you know, kind of real and ultimate-y things. Ultimate reality as markets is much tidier (and avoids all the messy business of explaining why gravity is a weaker than expected force).
posted by saulgoodman at 7:54 AM on October 29, 2008
Any trading/investing system that requires so much interaction isn't realistic for most people.
I think the best way to think of it is that TA suggests *when* to buy rather than *what* to buy.
Therefore if you are going to invest in anything, TA is helpful for timing your entry/exit.
I mean, if it works, you make money, and if it doesn't, it's no worse than random timing, right?
It is really astonishing how many times you look at stock picks on fundie sites, then go to the chart, and the chart is absolutely hair-raisingly bad. I don't mean bad in the sense of 'stock is beaten down and is now a bargain', I mean bad in the sense of 'stock is a falling knife'.
At the very least, TA allows you to pick a sensible entry point and stop-loss level. That right there is about 90% of investing. It doesn't matter if you check the charts once a minute or once a month really, so long as you have stops set.
posted by unSane at 9:24 AM on October 29, 2008
I think the best way to think of it is that TA suggests *when* to buy rather than *what* to buy.
Therefore if you are going to invest in anything, TA is helpful for timing your entry/exit.
I mean, if it works, you make money, and if it doesn't, it's no worse than random timing, right?
It is really astonishing how many times you look at stock picks on fundie sites, then go to the chart, and the chart is absolutely hair-raisingly bad. I don't mean bad in the sense of 'stock is beaten down and is now a bargain', I mean bad in the sense of 'stock is a falling knife'.
At the very least, TA allows you to pick a sensible entry point and stop-loss level. That right there is about 90% of investing. It doesn't matter if you check the charts once a minute or once a month really, so long as you have stops set.
posted by unSane at 9:24 AM on October 29, 2008
Seriously, try to logically refute my example above. If a lot of shares were bought at $20 because it seems to drop down to $20 and rise up to $22 with most shares traded at, say $20.50, when the stock drops down to $19, the majority of people that bought at $20.50 will FREAK THE FUCK OUT. One penny in green compared to one penny in red isn't a difference of a couple cents, it's huge. While TA calls $20.50 a support line (with some give down to $20), what's really going on is people freaking out at losing money. This causes a wave of selling down to $15. So people versed in TA will start pushing short positions at $19.50 (not exactly $20.50 or $20 obviously).
Now you can say that this is useless when it comes to fundamentals, but there is some legitimate arguments that it's going to $15 anyways, might as well get there faster with the TA people in the game (and on the other side, other TA people will be quicker to support it and buy at $15).
This accelerates market pricing and does serve a good in the markets, otherwise this process of selling will take a lot longer and create more mispricing.
Also, you can see TA participants eventually approaching the maximum point of risk-reward ratio (I wouldn't exactly call it zero-sum but it's something like that), in which the ratio between TA participants price equities quick enough for market participants with enough relative risk for both parties. In other words, the potential for moneymaking for people that do TA has an upper-limit, and reaches equilibrium.
posted by amuseDetachment at 10:25 AM on October 29, 2008
Now you can say that this is useless when it comes to fundamentals, but there is some legitimate arguments that it's going to $15 anyways, might as well get there faster with the TA people in the game (and on the other side, other TA people will be quicker to support it and buy at $15).
This accelerates market pricing and does serve a good in the markets, otherwise this process of selling will take a lot longer and create more mispricing.
Also, you can see TA participants eventually approaching the maximum point of risk-reward ratio (I wouldn't exactly call it zero-sum but it's something like that), in which the ratio between TA participants price equities quick enough for market participants with enough relative risk for both parties. In other words, the potential for moneymaking for people that do TA has an upper-limit, and reaches equilibrium.
posted by amuseDetachment at 10:25 AM on October 29, 2008
Also, many many traders use stops just below support levels. Thus in the above example, if I bought at $20 I might put in a stop at $19.50. So there's a pyramiding effect when support is broken, as the sell stops kick in. Large volume traders sometimes use this to their advantage, running prices down far enough to trigger stops before a big buy (known as an 'assblast' in my circles, because it often tempts technical traders to enter a short position just before the stock takes off like a rocket).
posted by unSane at 10:49 AM on October 29, 2008
posted by unSane at 10:49 AM on October 29, 2008
Therefore if you are going to invest in anything, TA is helpful for timing your entry/exit.
I mean, if it works, you make money, and if it doesn't, it's no worse than random timing, right?
That seems reasonable. I can live with that. *Shakes hands with unSane. They part amiably*
posted by diogenes at 11:36 AM on October 29, 2008
I mean, if it works, you make money, and if it doesn't, it's no worse than random timing, right?
That seems reasonable. I can live with that. *Shakes hands with unSane. They part amiably*
posted by diogenes at 11:36 AM on October 29, 2008
FWIW, my criticism all along was meant to be against trading on the technicals, it was meant to be against the idea of trading only on the technicals--because I think it's a bad, even dangerous, idea to encourage trading only on the technicals in principle. I'm applying something like Kant's Categorical Imperative to reach this conclusion, and here's my reasoning:
If everyone traded only on the technicals, you'd end up with the market effectively going into a feedback loop, and ultimately becoming extremely unpredictable and sensitive to minor short-term price fluctuations. Prices would never settle into consistent, stable trends over the long-term. Why? Because the price of the stock would effectively become exclusively a nonlinear function of the price of the stock. Prices would be determined recursively, which would give rise to chaotic behavior. (That's what I meant about prices chasing their own tails earlier.)
Still, obviously investors don't just look at the technicals, and there will (hopefully) always be plenty of investors taking long positions as investments in companies or otherwise basing their investment strategies on the fundamentals or other external realities. It makes perfect sense that you'd want to base the timing of trades on the technical indicators. And even as a trading strategy, TA probably works as good as any, assuming every investor in the market doesn't suddenly decide to use TA-driven strategies exclusively. But it's important, I think, to keep the big picture in sight.
posted by saulgoodman at 1:08 PM on October 29, 2008
If everyone traded only on the technicals, you'd end up with the market effectively going into a feedback loop, and ultimately becoming extremely unpredictable and sensitive to minor short-term price fluctuations. Prices would never settle into consistent, stable trends over the long-term. Why? Because the price of the stock would effectively become exclusively a nonlinear function of the price of the stock. Prices would be determined recursively, which would give rise to chaotic behavior. (That's what I meant about prices chasing their own tails earlier.)
Still, obviously investors don't just look at the technicals, and there will (hopefully) always be plenty of investors taking long positions as investments in companies or otherwise basing their investment strategies on the fundamentals or other external realities. It makes perfect sense that you'd want to base the timing of trades on the technical indicators. And even as a trading strategy, TA probably works as good as any, assuming every investor in the market doesn't suddenly decide to use TA-driven strategies exclusively. But it's important, I think, to keep the big picture in sight.
posted by saulgoodman at 1:08 PM on October 29, 2008
...my criticism all along was NOT meant to be against trading on the technicals...
heh.
posted by saulgoodman at 1:29 PM on October 29, 2008
heh.
posted by saulgoodman at 1:29 PM on October 29, 2008
The point of TA for me is that I assume (correctly) that many of the biggest market participants have much better information than I do, and are probably smarter too, at least collectively.
The only sane assumption in this case is that securities are priced, if not correctly, at least more correctly than I can price them.
So for example, if a stock is trading at a P/E of 4 and yielding 20%, there's probably a reason why it isn't a screaming bargain. Because if it was, smarter people than me would have driven the price up. It's like walking onto a used car lot and seeing a Ferrari for $1000.
(There are exceptions to this rule for thinly traded or penny stocks but bear with me).
However, what TA tracks and fundamentals don't are the dynamics of market sentiment, which is a distillation of all those brains and all that fundamental analysis and all that news and all that emotion, plus drivers like a hedge fund slowly unloading a big position, and so on and so on.
TA also allows you to see very quickly at what point in the business cycle a security, a sector or even a whole market is in. For example, a glance at a monthly chart of the Dow sectors shows you instantly that banking and real estate are much closer to a bottom than, say, biotech, which began its decline much later.
By tracking these kinds of changes it's possible to make much better bets when buying stocks. Buying a biotech stock when the whole biotech sector is turning down is fighting traffic, even if it's a great stock. Similarly, if a stock is in a long term downtrend on a monthly chart, it's going to make you think twice about buying into a rally on the daily chart.
Pretty much all you need to know about TA at that level is in Stan Weinstein's book with the ridiculous cover, and most of it is in a single chapter.
posted by unSane at 2:30 PM on October 29, 2008
The only sane assumption in this case is that securities are priced, if not correctly, at least more correctly than I can price them.
So for example, if a stock is trading at a P/E of 4 and yielding 20%, there's probably a reason why it isn't a screaming bargain. Because if it was, smarter people than me would have driven the price up. It's like walking onto a used car lot and seeing a Ferrari for $1000.
(There are exceptions to this rule for thinly traded or penny stocks but bear with me).
However, what TA tracks and fundamentals don't are the dynamics of market sentiment, which is a distillation of all those brains and all that fundamental analysis and all that news and all that emotion, plus drivers like a hedge fund slowly unloading a big position, and so on and so on.
TA also allows you to see very quickly at what point in the business cycle a security, a sector or even a whole market is in. For example, a glance at a monthly chart of the Dow sectors shows you instantly that banking and real estate are much closer to a bottom than, say, biotech, which began its decline much later.
By tracking these kinds of changes it's possible to make much better bets when buying stocks. Buying a biotech stock when the whole biotech sector is turning down is fighting traffic, even if it's a great stock. Similarly, if a stock is in a long term downtrend on a monthly chart, it's going to make you think twice about buying into a rally on the daily chart.
Pretty much all you need to know about TA at that level is in Stan Weinstein's book with the ridiculous cover, and most of it is in a single chapter.
posted by unSane at 2:30 PM on October 29, 2008
« Older When your dream of perfect beauty comes true | BBBRRRPPP-tschk BBBRRRPPP-tschk... Newer »
This thread has been archived and is closed to new comments
Mathematical psychologist Not Amadeupname demonstrated that human expectations occur a natural order. Hence variation in stock prices largely reflect human opinions, valuations and expectations.
Meaningful! I should try exploiting these "numbers" when trade.
posted by DU at 6:30 AM on October 28, 2008