A message to some day traders.

[Amahrix]

My opinion is that you are mathematically illiterate OP. This is probably why you can't maintain objectivity with regard to Nassim Taleb's book. Taleb has become some kind of mystic for non quantitative people. Taleb worshippers love saying 'fat tails' but have no clue of the complexity involved in quantitative modeling.

You are using the word 'random' in a naive sense. There are much more sophisticated ideas than those of this author.

Compounded distributions (this causes Kurtosis AKA 'fat tails')
Non-stationary Parameters
Semi-stationary Models

Smart people who actually understand mathematical statistics study how the accuracy of estimates are affected by different characteristics of random processes. They understand the limitations of statistics and other mathematical theories (and what effect violations of assumptions have on the predictive power of models).

Statistics is used to estimate (under a set of assumptions) theorized relationships, distributions of theorized parameters, and theorized distributions of errors that arise when modeling processes where we do not and cannot have complete information about the variability of phenomena.

Every attempt to predict markets will be subject to error, and even estimating the size of the errors is problematic. However, these models have been weaponized and are being used to compete for profits in financial markets.

Hi Real Money,

Please do not go away from original argument...there is no signal in the charts that day traders use to predict future price movement. (e.g. looking at last 15 minutes to predict next 15 minutes, or looking for a TA alert that went off and trading on this pseudo nonsense.

Today's Apple chart..... constructed purely randomly.

 

[MKTrader]

Every attempt to predict markets will be subject to error, and even estimating the size of the errors is problematic. However, these models have been weaponized and are being used to compete for profits in financial markets.

Maybe, but the models pushed here (chart patterns, popular lagging indictors, etc.) are beyond useless. If they ever provided a small edge, it's been gone since the late 80s/early 90s as even retail traders got access to basic charting/backtesting tools. Taleb is more right than wrong.

 

[Amahrix]

Look at Roku's chart over the same time frame.

This proves my point. The Roku chart doesn't have any information embedded in today's data that you can use to predict future price movement, the same way that yesterday's chart, or last weeks chart, or last months charged did not have embedded information that would have told you about this random -18% drop. Now if you look at a 2 year Roku chart, it has more information than today's daily chart.

 

[Real Money]

Today's Apple chart..... constructed purely randomly

Doesn't look random to me. But then again I'm a quantitative genius.

 

[jayboy]

"there are protracted periods of time during a trading session where the market wanders in the noise of more or less equitable buy and sell orders. We call that “chop”."

Even the sideways movement within a narrow range is setting up a future price move. There are many predictive patterns, but the best ones are not part of traditional TA. The OP has not paid his (I assume the OP is a male with ego issues; few females would waste their time on such a thread.) tuition, i.e., years of observation and experimentation.

 

[gaussian]

My opinion is that you are mathematically illiterate OP. This is probably why you can't maintain objectivity with regard to Nassim Taleb's book. Taleb has become some kind of mystic for non quantitative people. Taleb worshippers love saying 'fat tails' but have no clue of the complexity involved in quantitative modeling.

You are using the word 'random' in a naive sense. There are much more sophisticated ideas than those of this author.

Compounded distributions (this causes Kurtosis AKA 'fat tails')
Non-stationary Parameters
Semi-stationary Models

Smart people who actually understand mathematical statistics study how the accuracy of estimates are affected by different characteristics of random processes. They understand the limitations of statistics and other mathematical theories (and what effect violations of assumptions have on the predictive power of models).

Statistics is used to estimate (under a set of assumptions) theorized relationships, distributions of theorized parameters, and theorized distributions of errors that arise when modeling processes where we do not and cannot have complete information about the variability of phenomena.

Every attempt to predict markets will be subject to error, and even estimating the size of the errors is problematic. However, these models have been weaponized and are being used to compete for profits in financial markets.

I don't know about OP's math ability but you are right in that Taleb, in particular Fooled By Randomness has become another mystical text for non-quants. A Random Walk Down Wall Street is the same.

Taleb's premise - that models based on the normal (gaussian...heh) distribution have a weakness in that they do not properly model tail risk. Namely, tail risk occurs more frequently due to the fat tails (in intro probability courses you use a T distribution to get around this...).

Naturally non-quants take this as everything is bad and unpredictable instead of taking what Taleb IS saying - the random effect of extreme events is not properly accounted for in present day models. Taleb has, to my knowledge, never came out and said "everyone is stupid just shove all your money in an index fund and go home". That's generally the territory for guys like Dave Ramsey and his ilk.

Smart people who actually understand mathematical statistics study how the accuracy of estimates are affected by different characteristics of random processes. They understand the limitations of statistics and other mathematical theories (and what effect violations of assumptions have on the predictive power of models).

Yes, a model is a model because it models an event. If it was 100% accurate it would be the event itself.

 

[Amahrix]

"there are protracted periods of time during a trading session where the market wanders in the noise of more or less equitable buy and sell orders. We call that “chop”."

Even the sideways movement within a narrow range is setting up a future price move. There are many predictive patterns, but the best ones are not part of traditional TA. The OP has not paid his (I assume the OP is a male with ego issues; few females would waste their time on such a thread.) tuition, i.e., years of observation and experimentation.

Answer me Jayboy, do you believe today's 1 day chart of Apple is constructed with a precise intended message or purely random? Which do you realllly believe? Or the 45 minute chart of Apple a few hours ago... did that get printed and constructed to the tape randomly or with a precise intended message about the future.

....

 

[gaussian]

So wrong!

He never said that tail risks occur more frequently, in fact, he said in today's world, they appear even less frequently, but when they appear, it's impact is highly disproportional and therefore evidently is not gaussian.

I see you haven't even read Fooled By Randomness. Perhaps you ARE illiterate.

 

[Amahrix]

I see you haven't even read Fooled By Randomness. Perhaps you ARE illiterate.

Is todays chart on Apple constructed purely randomly or with a precise intended message about the future?

 

[jayboy]

I have no idea why you would ascribe intentionality to price action. The appropriate question is whether there are repeatable patterns that represent high R-value opportunities.

I don't, and never will, give a shit about an Apple chart because I trade index futures, on which I had five of my indicators flashing sell signals at the Open, three of which specified targets that were reached around 1300 Eastern.

To a mouse, a Gabon Viper has a random pattern that cannot be perceived as anything but leaf litter. You're a mouse and the best thing you can do is to keep trading your swing method and keep believing that day trading is impossible.