A message to some day traders.

For those who don't know, the OP's complaint is a good one, but is also reasonably met by the most basic sort of Times Series Decomposition methods. To wit:
Forecast = f[trend, annual, seasonal, [even daily], random] content. Your job as a data user is to weddle out the most of the systematic content, leaving the least: the *random* "unsystematic" content. Just not a big deal. Not a hard assignment. Not something not done all the time, in any multivariate context -- whether of weather, user data, cars-in-parking-lots, corn yields, or market prices.

Done every day, from Stats 201, to Ren. Cap......

{FWIW, there IS a real "data crime" going on out there, having do to with data stability and the recognition [or not!] of changes in data/serial correlation/etc over time. Most of those who wish to paint their work with "A.I." or "Machine Learning" glibly ignore these tenets, and end up with crappy results. It's a real thing, but not without its humor, either...} https://xkcd.com/1838/
 
Last edited:
tomorton said:
I'll go so far as to admit that the deeper traders drill into progressively smaller time-frames, the less reliable their conclusions based on past TA. No doubt at an infinitely short time-frame they are infinitely unreliable.
More...

Yes and I’m saying that the 1 minute time frame noise to signal ratio and the daily are virtually the same, weeklies not far out. There’s no pattern. It’s coincidences, self fulfilling prophecies, noise, Randomness. Get a million day traders in the room, a handful will be successful by sheer chance but they’ll think their gods and will write ebooks and sell you in on their “secret”... soon enough the ebook income will be more than there trading income (since the luck goes away) and then they are retired stock advisors all of a sudden. Look at these YouTube Day trader dream sellers. Y’all fooled by Randomness. WAKE UP
 
tommcginnis said:
For those who don't know, the OP's complaint is a good one, but is also reasonably met by the most basic sort of Times Series Decomposition methods. To wit:
Forecast == f[trend, annual, seasonal, [even daily], random] content. Your job as a data user is to weddle out the most of the systematic content, leaving the least: the *random* "unsystematic" content. Just not a big deal. Not a hard assignment. Not something not done all the time, in any multivariate context -- whether of weather, user data, cars-in-parking-lots, corn yields, or market prices.

Done every day, from Stats 201, to Ren. Cap......

{FWIW, there IS a real "data crime" going on out there, having do to with data stability and the recognition [or not!] of changes in data/serial correlation/etc over time. Most of those who wish to paint their work with "A.I." or "Machine Learning" glibly ignore these tenets, and end up with crappy results. It's a real thing, but not without its humor, either...} https://xkcd.com/1838/
More...

T,

A reliable forecast will not be had when the data that’s sampled in the function is noise. There are no patterns, no trends, no nothing. It’s rubbish.

Edit: day traders do not have signal in the data they get that’s required to day trade in the traditional sense.

P.S. I’m also saying that day trading in the traditional sense doesn’t work and never has. It’s a hoax.

Renaissance doesn’t day trade in the traditional sense. Stop using them in examples of day trading success. They hold many types of positions that range from daily to non daily as well as derivatives and insurance products. They know what they’re doing. Day traders don’t.
 
TheBigShort said:
Its random right?
More...

It’s more random (attributable to chance/luck) than you think. Hedge risk of ruin, take as many chances as you can take (increase frequency of bets) and you’ll hopefully increase luck in your favor, but maybe not, hence hedging risk of ruin. That’s another topic.

Bro btw thx for unfollowing the thread and still talking with us. We would love to continue having you here. This is the most excitement ET has had this month.
 
Amahrix said:
Yes and I’m saying that the 1 minute time frame noise to signal ratio and the daily are virtually the same, weeklies not far out. There’s no pattern. It’s coincidences, self fulfilling prophecies, noise, Randomness. Get a million day traders in the room, a handful will be successful by sheer chance but they’ll think their gods and will write ebooks and sell you in on their “secret”... soon enough the ebook income will be more than there trading income (since the luck goes away) and then they are retired stock advisors all of a sudden. Look at these YouTube Day trader dream sellers. Y’all fooled by Randomness. WAKE UP
More...


I can't accept that long-term trends are random, either in their inception, termination or continuation, though the trend probably can be shown to start or maybe both start and end due to FA. But once a trend is established the next most likely outcome is that price continues the trend.
 
IMPORTANT

@tomorton @TheBigShort @themickey @tommcginnis @soulfire and others

“Construct a population of 10,000 fictional investment managers […]. Assume that they each have a perfectly fair game; each one has a 50% probability of making $10,000 at the end of the year, and a 50% probability of losing $10,000. Let us introduce an additional restriction; once a manager has a single bad year, he is thrown out of the sample. […]

Toss a coin; heads and the manager will make $10,000 over the year, tails and he will lose $10,000. We run it for the first year. At the end of the year, we expect 5,000 managers to be up $10,000 each, and 5,000 to be down $10,000. Now we run the game a second year. Again, we can expect 2,500 managers to be up two years in a row; another year, 1,250; a fourth one, 625; a fifth, 313.

We have now, simply in a fair game, 313 managers who made money for five years in a row. Out of pure luck.

Meanwhile if we throw one of these successful traders into the real world we would get very interesting and helpful comments on his remarkable style, his incisive mind, and the influences that helped him achieve such success. Some analysts may attribute his achievement to precise elements among his childhood experiences. His biographer will dwell on the wonderful role models provided by his parents; we would be supplied with black-and-white pictures in the middle of the book of a great mind in the making.” — Nassim Nicholas Taleb
 
Last edited:
tomorton said:
I can't accept that long-term trends are random, either in their inception, termination or continuation, though the trend probably can be shown to start or maybe both start and end due to FA. But once a trend is established the next most likely outcome is that price continues the trend.
More...

Never said long-term trends are all random. My post is specifically about day traders and day trading.

Also, btw, a trend is hard to define because if you zoom out far enough there may be another underlying trend the opposite direction etc etc.

But a 1 year is highly more random than a 5 year which is more random than a 10 year which is more random than a 20 year which is more random than a 100 year which is more random than 1,000 years.

The further you zoom out, the better you can decipher the message behind the data.

A 1 year Apple, Inc chart tells me more than a 3 month Apple chart, but a 10 year really has the story if you want to talk data.

Same with news... if I read every headline will I be more informed of the world or more lost versus getting the news at the end of every month, because at the end of every month, at least the bullsh*t headlines will wash away and i'll be left with the stories that matter, even more so if that turns into checking the news every 2 or 3 months.

Lesson: Ironically, paradoxically, you make better, more quality decisions when you have LESS DATA. I'll be more informed of Apple if I check up on Tim Cooks cell phone line every 3 months than if I do every single day, can you imagine if I checked up on Tim Cook every single....MINUTE. I'll be even MORE INFORMED about the trajectory of Apple Inc's future if I call him once every 2 years and take in the data at a less frequent rate. #heuristics

day traders got their head stuck in the mud.
 
Last edited:
You must log in or register to reply here.
Back
Top