The Flip-A-Coin trading strategy and lessons learned from it

[abogdan]

Quote from Turok:

And with that... post #22, my friend Neutrino may now take his well earned trend following profits. Nice work (and a couple weeks early as well).

Human behavior is rarely random.

JB

Congratulation! You proved one more time that Earth is flat!
Good work!
Here is my prediction: You will post your replies every time that I post mine just so your word would be last. I played this game when I was a kid. I'll make peace with you. Yes you right! Technical analisys is a true science, trend is your friend and moving average is your GOD! I wish you good luck in 2004 and many many profitable days! But if it is all right with you, I'll still discuss with whoever is interested my, slightly different, point of view. Happy New Year!

 

[axeman]

Congratulation! You proved one more time that Earth is flat!Yes you right! Technical analisys is a true science, trend is your friend and moving average is your GOD!

Strawman fallacy.

Turok never made such silly claims.
You must use sarcasm now, because he already clearly
proved you are wrong.

You obviously could have brought your "Trend" to a dead
stop until nuetrino's deadline past, and you would have
proved him wrong.

Instead, you proved him right by continuing your posting trend.



peace

axeman

Quote from abogdan:

Congratulation! You proved one more time that Earth is flat!
Good work!
Here is my prediction: You will post your replies every time that I post mine just so your word would be last. I played this game when I was a kid. I'll make peace with you. Yes you right! Technical analisys is a true science, trend is your friend and moving average is your GOD! I wish you good luck in 2004 and many many profitable days! But if it is all right with you, I'll still discuss with whoever is interested my, slightly different, point of view. Happy New Year!

 

[axeman]

1) "positive associative memories" is an error the vast majority of
people make. At least those who dont have a basic understanding
of statistics (MOST).
2) Most TA indicators are BS. I agree.
3) "I mean that the probability of the trend to continue it self versus the probability of its reversal is still 50/50."
FALSE.
You simply havent found a way to do it yet.
Want proof?

Pick any year. Buy the index. Wait 50 years. Sell.
If there is NO trend, we explect a win% of 50/50.
This is obviously false. In this case, I would wager over
a 90% chance of a gain going with the trend.

4) Futher.... I have several trend systems with over 60% predictive
value of the future. No...I wont share them to prove it, so ill
just use the previous example in #3 as proof.


peace

axeman

It is not a question whether trends could describe movements in observed (historical) data the question is when do you know that the trend has been established and when do you say that the trend is over. When I say trends do not exist I mean that the probability of the trend to continue it self versus the probability of its reversal is still 50/50.


You just said your self "Weekly charts of the dow since march …" meaning: General direction of the price movement in DOW between March and Now was Up. Please note the word "WAS". Now, tell me what is the probability for DOW to continue this trend for the next 2 weeks? I'll tell you precisely - 50/50. Believe me, I have fell in this trap many times. Everybody is a genius when analysing historical data. Sure you can show me on any historical chart those reversal moments but in real time any prediction that you make will exhibit the probability of 50/50. This is, by the way, the main tool that all the people who sell any technical analysis bullshit use to convince you. Let me expand on that.
There is a very known psychological phenomenon called "positive associative memories". Without going to deep I'll use an example to illustrate it. Let's assume that I have predicted that the price of Gold will go up between now and two months from now. And, miracle, price of gold indeed went $10 up. In this case you will definitely remember my prediction and tell your friends that this guy knew something that we didn't. You will listen to my predictions from now on very carefully because of your "positive associative memories". On another hand, if the price of gold went down you will forget about my prediction and erase me from your memory. So, your memory, for the most part, consists of positive associations that will make you to believe those kinds of predictions over and over again!
I'll tell you more. I have tested over 1000 different technical analysis indicators over more than 20 years of tick-by-tick historical data (for which I paid enormous amount of money). Do you know what was the average objective probability of their accuracy? You guessed it right - 50/50. I have used more computing power to conduct this analysis than you could imagine! So, you would say: "was it a waste of time and money?" No. Although the results were negative I have discovered something that gives us all some hope! That is what I wanted to share with you all in a first place.
400 years ago people were convinced that the Earth was flat and they burned alive whoever thought otherwise (buy the way, it did not interfere with their abilities to grow crop and profit from it). 300 years ago people knew that the smallest animal on earth is Dicopomorpha echmepterygis till sir Van Leeuwenhoek invented the microscope!
2 days ago you knew that trends exist and they have more the 50/50 chance to continue. Imagine what you will know tomorrow! [/B]

 

[ig0r]

Quote from axeman:

Congratulation! You proved one more time that Earth is flat!Yes you right! Technical analisys is a true science, trend is your friend and moving average is your GOD!

Strawman fallacy.

Turok never made such silly claims.
You must use sarcasm now, because he already clearly
proved you are wrong.

You obviously could have brought your "Trend" to a dead
stop until nuetrino's deadline past, and you would have
proved him wrong.

Instead, you proved him right by continuing your posting trend.



peace

axeman

The fact that neutrino had a 50/50 chance of predicting the trend still remains. Abogdan could have just as easily stopped posting and ignored the thread, back when the prediction was made you couldn't have known. I want to see neutrino make this kind of prediction on a newbie that posts a 'how do I trade' thread, I want him to predict whether the newbie will start posting regularly or whether after a few posts will leave the site. Your chances of success (without cheating) are 50/50. Lets say you get it right the first time, well the idea of positive associative memories will lead all of us to believe you are some kind of predictive genius and that you can predict it every time, which is quite obviously false.

 

[ig0r]

Quote from axeman:

1) "positive associative memories" is an error the vast majority of
people make. At least those who dont have a basic understanding
of statistics (MOST).
2) Most TA indicators are BS. I agree.
3) "I mean that the probability of the trend to continue it self versus the probability of its reversal is still 50/50."
FALSE.
You simply havent found a way to do it yet.
Want proof?

Pick any year. Buy the index. Wait 50 years. Sell.
If there is NO trend, we explect a win% of 50/50.
This is obviously false. In this case, I would wager over
a 90% chance of a gain going with the trend.

4) Futher.... I have several trend systems with over 60% predictive
value of the future. No...I wont share them to prove it, so ill
just use the previous example in #3 as proof.


peace

axeman

This is hindsight axeman, you KNOW that the market has been strong for the last 60 years or so, you are not predicting trends... On the other hand, buying now and waiting 50 years, you have a 50/50 chance of being profitable. The fact that the market has had an upward tendency for the last 60 years doesnt prove the existence of trends either, only in hindsight can you see the start and end, for all you know we could have just entered a 60 year bear market after the 2000 top, but we don't do we? It's just as likely that we'll end up still strong after 60 years. The same thought process works in all timeframes.

 

[abogdan]

Pick any year. Buy the index. Wait 50 years. Sell.
If there is NO trend, we expect a win% of 50/50.
This is obviously false. In this case, I would wager over
a 90% chance of a gain going with the trend.

I agree, extremely long trading cycles typically have a DC component in them. It has nothing to do with TA it is just a devaluation of the currencies. It is like saying "My utility bills are twice as more expensive as they were 20 years ago". And if you simply take the ratio between the value of your dollar and the current prices you will see very different picture. I humbly ask you to go and buy the index in Feb 2000 because it was clearly a bullish trend. Wait a year and see what you come up with. I don't have time to wait 50 years. The whole purpose of this discussion is to analyze much shorter trends (1 - 10 days). Now, I belive you when you said that you have a trend system that works! Trading is like chess it is possible to have infinite amount of winning combinations. It is, again, not my point. But I don't believe that you use for your systems only trend indicators whatever they might be. You must have "counter trend" staff in them too. But anyway, Happy New Year! I sincerely wish you all a very profitable year! Truth is born in discussions!

 

[axeman]

Im looking at a chart back to 1905.
There is a CLEAR trend and upward bias to the markets.
There exist published papers by the EMT guys who even admit this.

Sure...we COULD go into a 60 year bear market, but we could
also be destroyed by a meteor tomorrow. Who cares.
It's a worthless point without some statistical context.

The point is... the larger the dataset, the higher the confidence
that a trend can exist and that the markets are not random.

100 years of data is a large enough data set to
convince me that trends DO exist.

The probability that you could flip a coin that many times and
come out with a result THIS skewed are well outside of mere chance.


peace

axeman

Quote from ig0r:

This is hindsight axeman, you KNOW that the market has been strong for the last 60 years or so, you are not predicting trends... On the other hand, buying know and waiting 50 years, you have a 50/50 chance of being profitable. The fact that the market has had an upward tendency for the last 60 years doesnt prove the existence of trends either, only in hindsight can you see the start and end, for all you know we could have just entered a 60 year bear market after the 2000 top, but we don't do we? It's just as likely that we'll end up still strong after 60 years. The same thought process works in all timeframes.

 

[Turok]

>Congratulation! You proved one more time
>that Earth is flat! Good work!

I just love your flair for the dramatic. You and I both know that it takes a lot more of a sample than that to "prove" anything statistically. No, nothing was proven...but "profits" were taken in a quite interesting example of entering a relatively low risk trade based on the loose expectation of a particular human behavior.

>Here is my prediction: You will post your replies
>every time that I post mine just so your word
>would be last.

Well, now we are getting somewhere. See, you have the reason wrong but the reason doesn't matter - trade what you see, not what you think. Go with the trend and take your profits. Congrats.

>Yes you right! Technical analisys is a true
>science, trend is your friend and moving
>average is your GOD!

Well, one out of three ain't that bad I guess.

My extensive testing has shown that 99.9% (at least) of the typical TA indicators and the like is rubbish and ends up on the 50/50 basket (see, the earth is round after all).

Moving averages are OK some of the time but are secondary to most of the systems I work with.

But, yes, the trend is my friend!

>But if it is all right with you, I'll still discuss
>with whoever is interested my, slightly
>different, point of view.

It's perfectly fine with me and I'm enjoying the exchange but of course you don't need permission from me to do this. Continue...I'm all ears.

JB

 

[NET]

Quote from abogdan:



There is a very known psychological phenomenon called "positive associative memories"....

First, Abogdan, let me say I'm really impressed with your comments--very compelling and intellectually stimulating. I'm not so sure I agree, however...

Are not price charts simply a visual picture of what the masses are doing? And if you agree with this, then wouldn't you also have to agree that the one can recognize a visual representation of where the masses have changed behavior? Does human behavior repeat itself?

Best Buy had a pre-Christmas sale on computers in limited quantity. I think something like $350.00 for a complete system. This price was two things: It "sounded" great, and it was perhaps a hundred dollars cheaper than one could normally buy the system. Hundreds of people were lined up for hours and hours at each location for the handful that were available.

So what's the point...? Well, you could have probably bidded up the price, and there would definitely come a point at which the crowds would leave.

Isn't this what happens when stock prices are advancing? Human behavior, driven by the greed emotion (this is not a negative by the way) will have many jumping on board up until the value does not seem so good. Resistance is established, and profit taking sets in.

Well, there's a crowd who's greed emotion causes a feeling of disappointment (because they did not buy before prices went up) and as such missed the profit opportunity. People expect patterns to repeat, therefore, the next time that price comes back to the starting point, they become a buyer (support).

Well, others are smart enough to figure that out, so they want to make sure they're able to "get in first" and therefore buy at a slightly higher price (rising lows). Once the crowd recognizes that prices will get no lower, the greed emotion causes a flood of purchasing, and prices spike up again.

Is it not this repeating of human behavior that causes a successive sequence of higher lows (the above cycle repeating itself) that others identify as a trend?

I'll tell you more. I have tested over 1000 different technical analysis indicators over more than 20 years of tick-by-tick historical data (for which I paid enormous amount of money). Do you know what was the average objective probability of their accuracy? You guessed it right - 50/50.

This is indeed a compelling argument. I've actually stopped using mathematical indicators. But I do use fibonacci retracements and projections, and trendlines. These indicators are not random from my experience and observation. Did you also test these methods of analysis?

 

[axeman]

I humbly ask you to go and buy the index in Feb 2000 because it was clearly a bullish trend. Wait a year and see what you come up with. I don't have time to wait 50 years. The whole purpose of this discussion is to analyze much shorter trends (1 - 10 days).


So you concede that longer term trends do exist then?
...and only wish to discuss shorter times frames less than 10 days?


Also.... I wager that if you bought the index in 2000, you have
a better than 50% chance of making money in 50 years.
Even the great crash of the 30's only took 15 years to recover from.

Guess we will just have to wait and see
I'll still probably be alive by then ....hehe...


peace

axeman

Quote from abogdan:

Pick any year. Buy the index. Wait 50 years. Sell.
If there is NO trend, we expect a win% of 50/50.
This is obviously false. In this case, I would wager over
a 90% chance of a gain going with the trend.

I agree, extremely long trading cycles typically have a DC component in them. It has nothing to do with TA it is just a devaluation of the currencies. It is like saying "My utility bills are twice as more expensive as they were 20 years ago". And if you simply take the ratio between the value of your dollar and the current prices you will see very different picture. I humbly ask you to go and buy the index in Feb 2000 because it was clearly a bullish trend. Wait a year and see what you come up with. I don't have time to wait 50 years. The whole purpose of this discussion is to analyze much shorter trends (1 - 10 days). Now, I belive you when you said that you have a trend system that works! Trading is like chess it is possible to have infinite amount of winning combinations. It is, again, not my point. But I don't believe that you use for your systems only trend indicators whatever they might be. You must have "counter trend" staff in them too. But anyway, Happy New Year! I sincerely wish you all a very profitable year! Truth is born in discussions!