Is technical analysis all smoke and mirrors?

[protagonist]

Seems like it to me, but what do I know?

[willthrill81]

It depends on what you're referring to, specifically. Some say that trend following is 'technical analysis mumbo jumbo', but Larry Swedroe has come full circle on the topic and now is in favor of it.

[RRAAYY3]

My buddy works on Wall Street - another is a financial advisor

Both, in casual conversation, refer to “the market” as “BS”

Because it is.

[Solo Prosperity]

Truth is, whatever keeps you disciplined to a long-term strategy, is fine by me. I personally believe that trend-following has a lot of merit to it (I have never bought into making investment decision on chart patterns but to each their own). It's a very un-boglehead stance to take, but having some of my investments follow a trend-following approach allows me to sleep better at night and stick to my portfolio allocation with a lot less behavioral interference. Sometimes I wish I could buy VASGX like others and never look at my accounts again, but I can't. The important thing is that I know that about myself.

Bottom line, if following a little technical analysis on a small portion of a portfolio helps someone stick to their portfolio as a whole, it's probably worth the price of admission...and who knows, maybe it will add alpha

[mega317]

but Larry Swedroe has come full circle on the topic and now is in favor of it.

This is an honest question: then why does he bother writing and advising others instead of trading constantly and becoming fabulously wealthy?

[willthrill81]

but Larry Swedroe has come full circle on the topic and now is in favor of it.

This is an honest question: then why does he bother writing and advising others instead of trading constantly and becoming fabulously wealthy?

This is a very common misconception about trend following. Even the most ambitious 'serious' trend following methods aren't believed to be even capable of outperforming the market by more than 1-2% annually over the long-term. If you add that to the market's returns, you aren't going to get rich quickly. And for most of us who follow trends, myself included, we do so more to manage downside risk than as an attempt to outperform the market's absolute returns.

Another objection people raise is why hedge fund managers don't do this. The answer is that some have and do, but the problem is that trend following can lag behind buy-and-hold for long periods of time. Something like the 200 day moving average strategy beat the pants off buy-and-hold from 2000-2009, but since then buy-and-hold has come out on top. It's hard for investors to stick with a 'niche' strategy, or any strategy for that matter, that can lag much from the overall market for nearly a decade.

[nedsaid]

My take is that technical analysis does a good job looking backwards and can explain in great deal what happened. The problem is projecting that into the future, smart alecks would say this is similar to academic research. I remember Louis Rukeyser and Wall $treet Week, he had a panel of technical analysts called the elves that would make bullish or bearish calls on the market. Problem was the elves were almost a perfect contrary indicator. They were bullish when they should have been bearish and bearish when they should have been bullish. The elves were so consistently wrong that Rukeyser fired them all.

In the run-up before the 2000-2002 crash and bear market, Gail Dudack was the only elf who was correctly bearish. Rukeyser fired her first and after the bear market hit, fired the rest.

[lack_ey]

but Larry Swedroe has come full circle on the topic and now is in favor of it.

This is an honest question: then why does he bother writing and advising others instead of trading constantly and becoming fabulously wealthy?

Well, the kind of technical analysis he likes, if you call it that, is the kind of medium-term signals of academic studies of trend following (hold typically on the order of months based on months-long momentum signals). As a strategy it loses money in plenty of years and is not that powerful in the sense that many trades lose money. i.e. the "edge" or economic significance of any advantage, if any, is not nearly large enough to become fabulously wealthy on the basis of investment results.

It's not like something you can execute on individual US stocks and make an average of 20 trades a day each averaging 1% return +/- 2%.

[Northern Flicker]

Sequences of coin tosses also exhibit “momentum”.

[Ari]

but Larry Swedroe has come full circle on the topic and now is in favor of it.

This is an honest question: then why does he bother writing and advising others instead of trading constantly and becoming fabulously wealthy?

Where does this idea come from that anyone who can make money investing in anything but index funds can get ultra wealthy in mere days? It's a really weird statement.

[just frank]

Of course technical analysis is all 'smoke an mirrors', its socially constructed. But then so is all money, so who cares?

The question is whether technical analysis can be used to make money by traders. The answer seems to be clearly that yes it can. Any method that is not valuation based is IMO technical analysis, and traders use them all the time.

Philosophically, we can say that it only 'works' because people believe in it, like money itself. But we don't care about philosophy, we care about success.

Stated another way, the market is a chaotic system with a fractal-like mathematical structure. You can't predict earthquakes (usually) but you can predict that after a big quake there will be little ones for awhile (aftershocks)...which an emergent 'feature' of that system with fractal dynamics. So technical analysis need not merely be all about traders exploiting different market behaviors created by a mass delusion that technical analysis works...it IS possible, that like aftershocks there are some (statistical) patterns in price movements that are robust features of market systems despite trader and investor activity.

I would argue that in a world with perfectly efficient markets, no corrupt trading, and ample information, traders would not exist and we would all be Bogleheads. The fact that traders exist (in turn b/c they make money) implies that some of those assumptions are not true...or that there are durable patterns in the market that can be exploited to make money 'honestly' (i.e. technical analysis is at least partially 'real').

[tmcc]

tech analysis is pure economic supply and demand.

order depth at a price per share

if you get an API data feed from a CME partner and start analyzing S&P futures, you can find some interesting things in order flow, trades and daily range. however, this forum doesn't think these things exist

for the curious, it costs like $50-100/mo and you'll need to program your own interface and trade a couple times a month.

[nisiprius]

I don't think calling time-series momentum "technical analysis" does anything but muddy the waters... any more than calling smart beta "stock-picking" or rebalancing "market timing." Or saying that antibiotics are like rubbing green mold on a wound. Or that vaccination is a kind of homeopathy. Or that astronomy is a kind of astrology. Or that chemistry is a kind of alchemy...

[Hillview]

A "Random Walk Down Wall Street" has a an interesting section on technical analysis. Worth a read if it interests you.

[just frank]

I don't think calling time-series momentum "technical analysis" does anything but muddy the waters... any more than calling smart beta "stock-picking" or rebalancing "market timing." Or saying that antibiotics are like rubbing green mold on a wound. Or that vaccination is a kind of homeopathy. Or that astronomy is a kind of astrology. Or that chemistry is a kind of alchemy...

OK then....what is technical analysis?

I would say that acting like Uncle Warren, analyzing a company, its markets and potential for future sales and profits is valuation based analysis (and revered around here).

I would say that there is financial engineering, like the packaging of investments into index funds to reduce volatility and risk of ruin and to help manage investor psychology is investing (and is also revered around here).

I guess profits made on selling valuation analyses and index (and other engineered) funds is a business model that is different from trading.

But if we are talking about traders who make money just by looking at the shape of squiggly curves to decide when to buy and sell (irrespective of the details)....I think they are all technical analysts. And around here they are openly disdained or it is denied they exist (profitable daytraders).

And like Santa, I think they exist.

[Taylor Larimore]

The question is whether technical analysis can be used to make money by traders. The answer seems to be clearly that yes it can. Any method that is not valuation based is IMO technical analysis, and traders use them all the time.

Bogleheads:

These are valuable quotes from those who know:

David Dreman, author of Contrarian Investment Strategies: "The performance of 185 tactical asset allocation mutual funds was compared with buy-and-hold strategies and equity mutual funds over the years 1985-97. Over this period the S&P 500 Index increased 734%, average equity funds increased 598%, and tactical asset allocation funds increased 384%."

Barber Odean Study: "Of 66,465 households with accounts at a large discount broker during 1991 to 1996, those that trade most earn an annual return of 11.4 percent, while the market returns 17.9 percent. Our central message is that trading is hazardous to your health."

Don Phillips, Managing Director of Morningstar: "I can't point to any mutual fund anywhere in the world that's produced a superior long-term record using market timing as its main investment criteria."

Jim Schmidt, Editor: "For the 10 years that ended 12-31-2000, only one newsletter out of the 112 that Timers Digest follows managed to beat the S&P 500 Benchmark."

Jason Zweig, author and Wall Street Journal columnist: "If you buy, and then hold a total-stock-market index fund, it is mathematically certain that you will outperform the vast majority of all other investors in the long run."

Best wishes
Taylor

[Watty]

Is technical analysis all smoke and mirrors?

It really does not matter because if it does work, then it works for everyone so the next question is if you are better and faster at doing technical analysis than the other million people that are using it to trade stocks. If not then any advantage will be gone before you can use it.

The term "technical analysis" is also a little bit dated now and "quantitative analysis" is more current and can get into things like nano-second trades.

https://www.investopedia.com/articles/i ... alysis.asp

[JBTX]

From a practical perspective is it BS because most investors can’t really translate the information into something useful. The only way to use it is to actively trade and that clearly over the long term has not been a successful approach for most.

[bigred77]

I don't know if I'd call it "smoke and mirrors". I prefer "the observed human tendency to associate patterns with random data".

I think technical analysis is not worth your time or attention.

I'm open to cases being made for fundamental analysis, but I've yet to be convinced, and I doubt I ever will.

[grayfox]

I am with nisiprius. It seems people here don't know what the term "technical analysis" means. One guy linked to an article about quants. That's not technical analysis. And Larry is not a technical analyst.

If you don't know what "technical analysis" and "fundamental analysis" is, you should read A Random Walk Down Wall Street. Then learn about the weak, semi-strong and strong forms of EMH.

BTW, I occasionally watch YouTube videos of technical analysis. It's hilarious! They draw a bunch of lines on a candlestick chart showing support and resistance. A few days later, when the stock market has crossed their random lines, they draw a new bunch of random lines. This process repeats endlessly.

[nisiprius]

...And like Santa, I think they exist.

??? If that's a joke, it's one level too deep for me to get.

I would say this. The question with technical analysis and many investment strategies is not "could there be something to it?"

The question is "Well, for the sake of argument, let's suppose there is. What then?"

Do I think I can do it myself? Do I think it's easy, if I just read a book or take a $395 course or buy some software?

Do I think that almost every competent technical analyst can beat the market?

If not, do I think I can identify the ones that can?

If I find a technical analyst who can beat the market, is he going to work for me for free? How do I do that? If not for free, do I think he's going to give me most of the alpha he creates out of the goodness of his heart? Or is he going to basically going to keep the lion's share for himself?

In contrast, yes, I really do believe that anyone can tie the market to within a few basis points, using a strategy that can be explained in a sentence (and justified in a few paragraphs): use a total market index fund.

[dm200]

Years ago, there were many sponsored radio "personal finance" shows on the radio. I don't know what ever happened to this guy, but one regular guy (selling his "services") was a PhD Physicist who claimed that the patterns charting stock prices/performance resembled those of things he studied and researched in Physics. He claimed to be able to predict stock performance. I think he sold subscriptions to his newsletter.

I never believed him - but listened to the show as entertainment.

[CokeSlurpee711]

I think there may be something there, but strictly reserved in the 5-10% "funny money" portion. Like the idea of incorporating it with other analysis like fundamentals, but do think there's value in technical analysis as far as support and resistance to determine good entry/exit points.

[cfs]

Thousands and thousands of TA personalities out there making a living and we don't want this bunch to be unemployed damaging the economy accelerating "stocks in freefall" . . . Good luck with our without the TA, y gracias por leer ~cfs~

[2015]

It depends on what you're referring to, specifically. Some say that trend following is 'technical analysis mumbo jumbo', but Larry Swedroe has come full circle on the topic and now is in favor of it.

Tells me all I need to know about this voodoo and the people who write about it. Technical analysis can best be referred to as narrative fallacy, among other things.

Why is it every new generation must fall for this stuff, only to learn the hard way? Many of the responses in this thread coupled with Taylor's excellent quotes above clearly demonstrate why the overconfident will always be with us, if only in never ending new waves of virgin investors (suckers).

[jonbois]

It adds discipline to the trading equation

[Pajamas]

The basis of technical analysis is data, which describes what has actually happened. I doubt many would deny that there is some predictive value in data, for instance in identifying trends, but data can easily be misinterpreted or misused to reach conclusions that are not actually supported, and more importantly, conditions change constantly. The more extreme or whacky technical analysis is amusing in the same way that astrology is amusing. So, it is not all smoke and mirrors, but there is frequently an element of smoke and mirrors to it.

Technical analysis can also be self-fulfilling to some degree, like that time in the 1970s that Johnny Carson created a short-lived toilet paper shortage.

Another useful comparison might be standardized admission tests used in education. They definitely measure something, and the scores have some predictive value, but they can't be relied on absolutely to prognosticate future scholastic success.

[fortyofforty]

It's very useful to explain what happened. It's of zero use to predict what will happen.

[Socal77]

Here's a rhetorical question. Does technical analysis foretell the exogenous shocks that could happen at any time in the future?

[randomizer]

Yes.

[Walkure]

David Dreman, author of Contrarian Investment Strategies: "The performance of 185 tactical asset allocation mutual funds was compared with buy-and-hold strategies and equity mutual funds over the years 1985-97. Over this period the S&P 500 Index increased 734%, average equity funds increased 598%, and tactical asset allocation funds increased 384%." ....

Jason Zweig, author and Wall Street Journal columnist: "If you buy, and then hold a total-stock-market index fund, it is mathematically certain that you will outperform the vast majority of all other investors in the long run."

I think there's a really interesting point here that gets missed in the rush to exult over the inability of active management to "outperform" the market. In fact, I think there's a quasi-immutable feature of economics at work here, which I propose to call Walkure's Rule (apologies if someone has already invented it) First, it appears to be an almost linear inverse relationship between the relative turnover of the passive/active/trader and their respective returns. But that's not the rule. Rather, it is my proposed explanation for the relationship: that the difference in total returns is a function of the ratio of realized to unrealized gains in their respective performance.

Thus, the index is expected to have the highest gain over any given finite time frame because the underlying securities have only been sold (and the gains realized) to the extent that total inflows lag total outflows, or in the case of a particular security being dropped from the index. The active manager, who has presumably been moving stuff around to justify his commission, will have accrued some amount of realized gains. And the trader, who must perforce closeout each position in the near term, has almost exclusively realized gains. The point being that the act of realizing the gain necessarily incurs not only transaction costs but also depresses (by a miniscule amount per trade) the actual amount of the gain, and therefore his "underperformance" is really due to being compared against an intrinsic feature of the index's float vs. the paper gains of it's NAV.

[dm200]

Years ago, there were many sponsored radio "personal finance" shows on the radio. I don't know what ever happened to this guy, but one regular guy (selling his "services") was a PhD Physicist who claimed that the patterns charting stock prices/performance resembled those of things he studied and researched in Physics. He claimed to be able to predict stock performance. I think he sold subscriptions to his newsletter.

I never believed him - but listened to the show as entertainment.

Wish, now, that I had saved this guy's (entertaining) sales pitch.

[David Jay]

I don't know if I'd call it "smoke and mirrors". I prefer "the observed human tendency to associate patterns with random data"

This, and it has been going on for centuries. Shamans used to examine chicken entrails, looking for patterns. Now "technical" analysts examine squiggly lines on a chart.

[hornet96]

According to the CFA Institute:

Technical analysis is a form of security analysis that uses price and volume data, which is often graphically displayed, in decision making. Technical analysis can be used for securities in any freely traded market around the globe.[....] Technical analysis of any financial instrument does not require detailed knowledge of that instrument. As long as the chart represents the action in a freely traded market, a technician does not even need to know the name or type of the security to conduct the analysis. [.....]

Technical analysis can be thought of as the study of collective investor psychology, or sentiment. Prices in any freely traded market are set by human beings or their automated proxies (such as computerized trading programs), and price is set at the equilibrium between supply and demand at any instant in time. Various fundamental theorists have proposed that markets are efficient and rational, but technicians believe that humans are often irrational and emotional and that they tend to behave similarly in similar circumstances. [....] Technicians believe that market trends and patterns reflect this irrational human behavior.

[Cannot paste link as you have to be a member to access the reading]

Essentially, technical analysis is an attempt to capture value from any inefficiencies caused by human irrationality in the market, as witnessed in the past market data itself.

FWIW, I am a CFA Charterholder, and this topic received very little (if any) attention or weighting on the exams (if that gives you any sense of the relative importance of this topic in the investment world).

[telemark]

The market is not simply random data: it actively resists attempts to study it.

There is a theory which states that if ever anyone discovers exactly what the Universe is for and why it is here, it will instantly disappear and be replaced by something even more bizarre and inexplicable.

There is another theory which states that this has already happened.

[nisiprius]

A characteristic feature of most of the "technical analysis" that I have looked at is it uses hedged, allusive, ambiguous Delphic language that makes it impossible to tell if any prediction is actually being made, and, if so, what the prediction is and when it is supposed to come true. As a result, it mostly isn't possible to test it or keep score. Here's an example:

Why Bitcoin's death cross may become a bear trap

Bitcoin risks entering a technical "death cross" soon, but the bearish signal will likely not be as severe as has been made out in reports.... Some strategists are saying that the death cross could yield a big sell-off in BTC, possibly to as low as $2,800, a level last seen in September 2017. However, such fears are likely overstated, as the crossover tends to work as a contrarian indicator - that is, they tend to occur at the end of a big bear move, with prices rallying soon after....

To cut a long story short, BTC had to drop by $14,000 (from $20,000 to $6,000) to push the 50-day MA so far towards the 200-day MA. Hence, it's likely that the bears will run out of steam by the time the actual death cross occurs.

In fact, it could end up being a bear trap, at least in the short term. And the historical data seems to support the argument, as explained below....the death cross failed to yield a big sell-off in two out of the last three events, and the odds are high that it would end up being a bear trap for the third time.

A move to $7,240 (recent low) will likely confirm the death cross and may yield further drop towards $6,600. That said, the support will likely hold, with the daily RSI likely to show oversold conditions by then. In the subsequent days, bitcoin may trap the bears on the wrong side of the trade, as seen in April 2014 and September 2015.

However, if BTC finds acceptance below $6,600, a further sell-off to sub-$6,000 levels cannot be ruled out.

That's only part of it, and notice just how squishy it is. If there's a big sell-off, it confirms the validity of the "death cross," and it doesn't falsify this analyst's predictions because he only says it "could" be a bear trap. If there is only a small sell-off, then he is right that "such fears are likely overstated." But then again "a sell-off to sub-$6,000 levels cannot be ruled out." In short, anything can happen--and there isn't any time frame on when any of is supposed to happen so no matter how bad the prediction seems to be, someone can always say "he wasn't wrong, just early."

[cfs]

Smoke and Mirrors is winning by a landslide . . . keep the posts coming y gracias por leer ~cfs~

[BolderBoy]

I personally believe that trend-following has a lot of merit to it (I have never bought into making investment decision on chart patterns but to each their own).

For 3 years in the mid-2000s I tried "trend following" with a guy who had a large following and claimed he had made a lot of money doing it. I lost money doing it, following his precise instructions.

Looking back, it is just another form of market timing.

[MinhN]

Technical analysis is a self fulfilling prophecy. If enough people believe in it, it could work.

[just frank]

...And like Santa, I think they exist.

??? If that's a joke, it's one level too deep for me to get.

I would say this. The question with technical analysis and many investment strategies is not "could there be something to it?"

The question is "Well, for the sake of argument, let's suppose there is. What then?"

Do I think I can do it myself? Do I think it's easy, if I just read a book or take a $395 course or buy some software?

Do I think that almost every competent technical analyst can beat the market?

If not, do I think I can identify the ones that can?

If I find a technical analyst who can beat the market, is he going to work for me for free? How do I do that? If not for free, do I think he's going to give me most of the alpha he creates out of the goodness of his heart? Or is he going to basically going to keep the lion's share for himself?

In contrast, yes, I really do believe that anyone can tie the market to within a few basis points, using a strategy that can be explained in a sentence (and justified in a few paragraphs): use a total market index fund.

Not advocating for technical analysis, nor do I think this is an actionable topic.

But as a purely philosophical and mathematical point, I think there can be structure to market price movements (and it would be impossible to prove otherwise) AND that some of them can probably be exploited for gains.

The fact that most people who try to 'time' or whatever fail relative to an index likely has more to do with human psychology than the market. And it does not mean that an algorithm couldn't do its thing and generate a little alpha.

[protagonist]

This is not my field, but it seems to me that it should be easy to say whether it is smoke or mirrors or not, or at least if it is justified as a reliable tool for predicting future market movement, by looking at the math behind it.

If the math justifies the conclusions, then it is science. If not, then it is guesswork. Which isn't to say that guesswork is necessarily wrong...that is often the first step towards scientific discovery.....just that it is not science.

Looking back over a limited time period and finding some parameter a posteriori that led to increased returns is not good enough- if the math is absent to show why it should work in the future and if there is no good testable and falsifiable hypothesis going forward, then there is no real scientific basis for belief, beyond faith. Otherwise, "philosophy" is just a fancy word for faith.

[Alexa9]

From Bogle:

https://www.youtube.com/watch?v=A0gQiz0pCyI

All you need to know about investing in three words: nobody knows nothing!

[Solo Prosperity]

I personally believe that trend-following has a lot of merit to it (I have never bought into making investment decision on chart patterns but to each their own).

For 3 years in the mid-2000s I tried "trend following" with a guy who had a large following and claimed he had made a lot of money doing it. I lost money doing it, following his precise instructions.

Looking back, it is just another form of market timing.

1. It is very clearly market-timing. No denying that.
2. Not sure who this "guy" was or what his method was, but the 2000-2010 decade was one of the best trend-following decades in history for equities (See below).
3. Like any investment style, 3 years is not long enough to make a call. I hope you didn't stop following the trend rules right before 2008...Regardless, and this is insanely important, whatever strategy you choose, you must be prepared to invest that way for the long-haul. I like to think about it as my entire lifetime. If I want to invest in value, momentum, trend etc., I better be prepared to follow the tenets of said strategy for my entire investment career...if not, just buy VASGX and call it a day.

[triceratop]

I thought this was an engaging post: Everyone is a Closet Technician

Everyone is a closet technician. Everyone. And in a panic or a market correction, this truism is even more, um, truistic.

First, what is a technician? Here’s my own handy definition, I think you’ll like it: A technician is someone who cuts right to the chase and studies actual prices and behavior instead of puzzling over the causes of prices and behavior like everyone else.

Discussing causes is a much more interesting conversation and it gets you on all the talk shows. Discussing price – the sum total of all investor fear and greed, both historical and real-time – tells you the truth about what’s actually going on, it does not offer an opinion.

Besides, price dictates what the news is, not the other way around. Consider:

If the price of Yahoo common stock was higher today than it was when Marissa Mayer first joined, she’d be hailed as the second coming of Lou Gerstner or Steve Jobs, not a punching bag whose business initiatives and party-throwing receipts are dissected in the tech press each week.

...

[GibsonL6s]

Years ago, there were many sponsored radio "personal finance" shows on the radio. I don't know what ever happened to this guy, but one regular guy (selling his "services") was a PhD Physicist who claimed that the patterns charting stock prices/performance resembled those of things he studied and researched in Physics. He claimed to be able to predict stock performance. I think he sold subscriptions to his newsletter.

I never believed him - but listened to the show as entertainment.

Wish, now, that I had saved this guy's (entertaining) sales pitch.

Was he an Indian guy I think I remember him people would give him a stock symbol and he would chart it Nd tell them what he thought based on the technical factors he saw.

[willthrill81]

I personally believe that trend-following has a lot of merit to it (I have never bought into making investment decision on chart patterns but to each their own).

For 3 years in the mid-2000s I tried "trend following" with a guy who had a large following and claimed he had made a lot of money doing it. I lost money doing it, following his precise instructions.

Looking back, it is just another form of market timing.

1. It is very clearly market-timing. No denying that.
2. Not sure who this "guy" was or what his method was, but the 2000-2010 decade was one of the best trend-following decades in history for equities (See below).
3. Like any investment style, 3 years is not long enough to make a call. I hope you didn't stop following the trend rules right before 2008...Regardless, and this is insanely important, whatever strategy you choose, you must be prepared to invest that way for the long-haul. I like to think about it as my entire lifetime. If I want to invest in value, momentum, trend etc., I better be prepared to follow the tenets of said strategy for my entire investment career...if not, just buy VASGX and call it a day.

Over the long-term, I think that most buy-and-holders and trend followers will have similar returns (though I think trend followers will experience less downside volatility). But the key is not switching your strategy at every whim, particularly when the market is gyrating intensely.

[wootwoot]

Watch this documentary about Paul Tudor Jones and let me know if you think technical analysis is real or not.

https://m.youtube.com/watch?v=-38x671CUQw

[protagonist]

Watch this documentary about Paul Tudor Jones and let me know if you think technical analysis is real or not.

https://m.youtube.com/watch?v=-38x671CUQw

It looks interesting from the beginning, but it is 55 minutes long and I am hungry. Can you please indulge me and summarize?

[Karamatsu]

There is a definition problem but I think of "pure" technical analysis as being analysis of patterns in price action totally irrespective of what the underlying issue is (though of course, the analysis parameters differ depending on what type of asset you're talking about) or any other outside data. My take is that it's mostly just data fitting. You can find all kinds of patterns in past data, and people are amazingly creative when it comes to such things. None are absolute, of course, or somebody using them would already own the whole market. Probably the best anyone can hope for is to win more than they lose over the long run, but of course... the financial graveyards are full of people for whom the long run was just too long, and they went bankrupt before their system could get make them rich (or if they used margin, even get them back to zero).

Computer power has, of course, turned it into an endless game with all kinds of machine learning algorithms battling each other in real time, but that's something on a different order since in many cases no human knows what they're doing. For all we know the programs are already sentient, and just messing with us to see how we react.

Anyway, long and short of it, there ARE patterns in the past, even double dip candlestick parabolic Fibonacci (partial) Ehrenberg crosses if you look hard enough. But they're unlikely to be of much help in the future. Better to invest in a few index funds, avoid trading too much, keep costs down, and use the time saved to live your life. That's what investment is for, after all.

[Pelerus]

Sequences of coin tosses also exhibit “momentum”.

Ludic fallacy?

However I do concede that it’s a compelling argument. Many players insist that the game of blackjack is unusually “streaky,” but this is not supported by evidence: each hand is essentially an an independent random trial in terms of win/loss outcome. The correct explanation for the apparent streakiness is instead selective memory - players remember long strings of wins or losses more than equivalent strings of a more uniform distribution, because the former have more emotional impact.

Studies have also shown that in general, people overestimate the uniformity of random trials, like coin tossing. That is, people tend to underestimate the frequency with which random coin tosses result in clumpy sequences of heads or tails.

A counter argument against this cognitive explanation for apparent momentum/streakiness in financial markets is that such explanations succumb to Taleb’s ludic fallacy: the misuse of games of chance to model real world situations. The fundamental flaw in doing so is that games of chance have no “unknown unknowns,” insofar as all the variables in these games are known in advance.

In the real world of financial markets, by contrast, there are a multitude of unknown unknowns, many relating to human behavior, and there is wide disagreement on the correct approach to decision making under uncertainty, even at a theoretical level, let alone in practice. This stands in stark contrast to the world of advantage play at casino games, in which there is a mathematical level of certainty and unanimity over the statistical edge one can have over the house by employing particular strategies.