Looking at technical indicators doesn't imply that they work, no more than drawing unicorns implies unicorns exist. And by "work", doesn't just mean you can make money from them, but that you can outperform the broader market on risk adjusted basis.

Sent from my SAMSUNG-SM-G925A using Tapatalk

Yeah, the notion that the market isn't efficient if one person does something dumb is silly. EMH doesn't even require half rational actors let alone all rational actors.

Oh, I thought his argument was "you wouldn't be able to make money using TA if weak EMH is true, but people make money using TA". But saying 1 person uses TA isn't even an argument.

I agree that philosophically it seems a little too convenient. But again, with the following definition: claims that past price movements and volume data do not affect stock prices. As weak form efficiency is theoretical in nature...

All it takes is one. Stock prices at any given point in time are the intersection of the bid and the ask. Let's do a thought experiment, at t = 0 all the bids and all the asks are entered based upon fundamental information. The last transaction was at the bid. Then at t=1, one trader places a market buy order based upon a technical signal. This changes the price, and because it was a technical signal, demonstrates that the current price (even if to a small degree) reflects past price action.

Further, this buy (new price) could change someone's risk and cause them to place another market buy order to rebalance. This is called a feedback loop, another characteristic of complex systems. And something that George Soros has written about extensively (he calls it market reflexivity). Dalio, Bass, and anyone worth their salt openly admit and discuss feedback loops in the markets.

Buying incentivizes more buying. Buy stops are hit, short covering ensues, which incentives more buying...

This conversation is ****ing ******ed...To think that past price action or volume data does not affect current or future action is at best ignorance.

Are either one of you now going to argue against feedback loops in financial markets? A child with a modicum of common sense would immediately know that you are wrong...

You're not refuting weak EMH with your argument. Even hardware entropy sources act on deterministic single events, but together, are not predictable and act randomly in aggregate. The same thing happens with your supposed TA trader. The market is still random if a gaggle of millions of fundamental, TA, news, algo etc traders all act according to their own desire and models. If you can't predict or determine the aggregate of the inputs to the system, it's random and weak EMH is true.

Re-read the thread title! With that in mind it appears to me that there seems to be no concern whatsoever from the politicians in aggregate - and apparently from many of the professionals - as things stand at this time so I ask the question 'Are they out of their minds?' here. I'll admit that I don't understand most of the discussion as it's never been an interest of mine so I didn't study it. I will say that Tooth's contributions appear to be deep and thoughtful so I'm getting something out of that.

The principle of one technical trader disproving weak emh has a kernel of truth. Either the guy has a working theory and is capturing an inefficiency himself, or he's creating an inefficiency by trying and failing.

And in a world where there were no transaction fees and stocks were bought/sold down to the nth decimal place in price the existence of one technical trader (even if he was clueless) would prove there to be some very minor inefficiency but it's not really material to the argument when institutional money is as significant as it is. They're generally squeezing the price in opposite directions all the way down to the penny.

Yes, this is true and very fair. But it is a theory that we are discussing.

And further, there are many technical traders out there, not just one. Institutions as well perform technical analysis. Things like moving averages, RSI, etc are certainly used by institutions.

I promise that there are plenty of algos that take the current / last price as a parameter to the function that determines their next action. This is like, the definition of past price action affecting current price action.

I've yet to see convincing data that proves individuals are doing it with any real success.

I suspect the vast majority of the value is eaten up by a very small number of actors using software that automates the processes and tests an endless list of hypotheses using all available data.

Yeah, IDK if you are ever going to see good data on the signals (technical or otherwise) that institutions use to place trades.

But I promise you, pretty much everyone looks at things like the 50 day moving average.

I don't remember very well because I have a lot going on right now, am reading a lot of books etc, but one of the ones I am reading is Ray Dalio's Principles. I highly recommend this book by the way.

Anyway, like I said, I am not certain, but I am pretty sure he talks about technical signals in it. Like Bridgewater establishes most of its positions for fundamental reasons but manages risk based on technical signals or something to that effect.

What level of abstraction do you live in? Do you not consider/model a roulette wheel as random or a shuffled deck of cards as random? Clearly, these are also not random due to the underlying physics. In fact, nothing really is random except for quantum physics, and even that is debatable.

I love it when people say "this discussion is stupid/******ed" when they think they made an insightful point or think they know it all. You're second guy (next to BoredSocial) in this thread to do this.

Regardless, while random markets imply weak EMH, weak EMH does not imply random markets, so the whole talk on randomness is totally relevant. At the end of the day, weak EMH simply states no edge can be derived by examining historical data (as I've stated probably 3 times now). And by edge, I mean risk-adjusted excess return.

See definition of weak EMH above.

Where?

Its a fair question. I consider those things random - practically random. To be clear, I am not claiming that the markets are not random walks because I believe that nothing is random.

I think your typo drastically changed the meaning of what you meant to say. It seems to me that there are two conversations going on:
1. random walk
2. EMH (weak form)

I think both are wrong and not terribly useful. But models that were derived from these theories had (past tense) their place. They are, in my opinion, becoming less and less relevant (I meant to omit the ir...).

Where have i NOT seen convincing data? This thread. But i'm not claiming it doesn't exist.

The definition of weak EMH is such that you can't get an edge by looking at data from the past. If people are trading based on historic price data they're either doing it successfully or doing it unsuccessfully and, at least for as long as they're using their failed model, creating a discernible ripple effect on a very small scale.

One of us has misread / misinterpreted Addbabadbaa's (butchered almost on purpose) post.

My money is on you.

"Discernible ripple effect", even if we assume this exists, does not disprove weak EMH. Again, to disprove weak EMH, you have to prove there is an edge from Abbaddabba's Ripple Effects, which I would imagine would entail being able to identify it and knowing how it influences future price movement and/or volatility etc.

You mean like Japan? With their complete lack of growth over the last 27 years combined with their absolutely massive debt load their bond rates must be extremely high!

*checks Japan bond rates*

Nevermind. It doesn't seem to work that way.

You'd get an incremental upward pressure on price when they buy and downward when they sell. Snatch up that millionth of a cent in value accordingly.

And how do you know that that incoming trade is dubious? And even if it's dubious, how do you know price will behave as you expect?

Moreover, this is sort low-level discussion is like saying rolling a dice can't be modeled as random because of physics. It's not a very practical model of the world for someone trying to formulate a backgammon strategy.

I tend to think it's not important because the country technically could start paying down the debt without significantly slowing growth by enacting sensible changes to federal taxation and spending. Or we could at least get to a point where the deficit is smaller than inflation and the debt shrinks in real terms.

Congress and the world's economists would love to hear from you.

I built a fairly straightforward trading model using TA that made decent $ on SPY consistently for about 18 months. It ran into a dry spell and I gave up, but pocketed a sizeable gain. Not sure if it would work now.

Buying SVXY (short vol) two years ago paid 600% without doing a damn thing or looking at technicals.

In my case I made 100s of trades, a very high % of which were profitable.

Also hard to think of a more liquid market than SPY.

How do you think a pure TA "buy any dip and hold until 5 points of profit" trade would have gone on this random walk? Pretty darn well I'd say. If you made hundreds of trades using that strategy, what percentage do you think would have profitable? (hint: a very high %).