Don't Be Fooled By Randomness

[BeginnerJoe]

I have been conducting a personal experiment in randomness for some weeks now.
I can't tell on here where, in case they chuck me off, but I have a demo account of a forex nature. The entries are all done with a coin toss. The exits are made when they are 50+ pips up or stopped out on their stops ( about 300) pips .

Erm yes it is currently showing a loss.

Why the reluctance in putting in the money, not like you need to bet more than 0.0001 pence ? Your can't be that hard up, surely ? But if you use this random BS with real money, you'd get cleaned up down to your underpants. It would make a much more interesting story.

 

[BeginnerJoe]

BTW, no need for a physical coin.

True. Pick your method, any method. The market wouldn't know when you are coming or going. In other words, you are already random. So coins are not necessary, C# is not necessary, just be yourself and you will get all the random goodness you have always dreamed of. But will it help you ? Not the slightest.

To point to one form of randomness and say that's a better randomness than another randomness is just pure arbitrary and random, and a sign of grasping at straws.

BTW, computer generated random is less random than coin flips. Computer random is selecting a random entry from a known mathematical set and sequence.

With all the talk about random, people are forgetting why the guy on the other side of your trade is in business, and why he should give you a profit when he can quite easily keep it for himself ? In fact he does.

 

[Pat494]

I'm surprised noone has come up with quantum physics for an explanation.

So if someone goes short say then someone somewhere else will take the opposite view and go long. A bit like sitting on one end of the seesaw. Down your end and up elsewhere and then maybe down a bit more than your down is pushed down elsewhere so you lose.

 

[BeginnerJoe]

So if someone goes short say then someone somewhere else will take the opposite view and go long.

The deal will not between the two someone's. The deal will be between:

Someone number 1 <--> institutional market maker/counterparty
Someone number 2 <--> institutional market maker/counterparty

The resulting price move or non-move will be determined by the institutional market maker/counterparty for his benefit. He profits by making both someone's loosers.

 

[the hare]

BTW, computer generated random is less random than coin flips. Computer random is selecting a random entry from a known mathematical set and sequence.

You might find this amusing

bookofjoe: A coin flip isn't random

The reliability of the random generator isn't really that much of an issue, a coin will do just fine.

 

[Shakone]

You might find this amusing

bookofjoe: A coin flip isn't random

The reliability of the random generator isn't really that much of an issue, a coin will do just fine.

Am I the only person that gets annoyed by people mis-using the word random when they mean something else, like equally likely, or uniformly distributed or whatever. A coin can be biased and still be random. I'm struggling to think of things that aren't random.

P.S. I recommend using ran2() for your random number generator (see numerical recipes)

 

[Shakone]

the hare you may find this blog interesting

Intelligent Trading: Can one beat a Random Walk-- IMPOSSIBLE (you say?)

lets not get bogged down in the mistakes of the assertion...

 

[the hare]

the hare you may find this blog interesting..

Cheers, I hadn't seen that one previously. Looks promising.

 

[Liquid validity]

Am I the only person that gets annoyed by people mis-using the word random when they mean something else, like equally likely, or uniformly distributed or whatever. A coin can be biased and still be random. I'm struggling to think of things that aren't random.

True

P.S. I recommend using ran2() for your random number generator (see numerical recipes)

I don't actually use an RNG myself, I use price itself.
No particular reason, if I'm honest it was easier thats all.

 

[Liquid validity]

To point to one form of randomness and say that's a better randomness than another randomness is just pure arbitrary and random, and a sign of grasping at straws.

Purely for convenience and automation.
I don't use a coin or RNG, I use price.

 

[Pat494]

Purely for convenience and automation.
I don't use a coin or RNG, I use price.

seems like that is chaotically random at times

 

[scose-no-doubt]

drift factor

 

[NVP]

:-0 Some studies have shown that a high percentage of chart patterns may be simply random, squiggly lines on a piece of paper. Burton Malkiel in A Random Walk Down Wall Street described an experiment his students participated in, using a hypothetical stock trading at $50 a share. Each day they flipped a coin, plotting heads as a 1/2-point daily gain and tails as a 1/2-point daily decline. The resulting stock chart from these random coin flips displayed all the classical chart patterns such as head-and-shoulder formations, flags, pennants, triangles, tops, and bottoms. There were even indications of cycles in the random tosses.

so point proved and dont trade ever again my friend ..........youve saved yourself a fortune in time and wasted energy over the next few years

walk away as you cannot trade a random market ....

cheers
N

PS - is the world still Flat ?

 

[random12345]

so point proved and dont trade ever again my friend ..........youve saved yourself a fortune in time and wasted energy over the next few years

walk away as you cannot trade a random market ....

cheers
N

PS - is the world still Flat ?

I think anyone who has really traded live with their own account knows that the market isn't 'random' in the same way a coin flip is because over a billion coin tosses, there will be little deviation from 50/50 for any individual, with only a few small standard devs. Anyone who has designed an automated system for the market knows that they'd cut off a nut for a 50% HFT winner using equal entry and exit strats over 10 years, which would make 100,000s of trades. You can only get a 50% winner if your parabolics or MAs are skewed to the overall direction of the market over those years, which is not a system that can trade any move intraday (i.e. you would optimise a very short 'short entry' signal for EURJPY over the past 20 years and it would have made big losses this year, but profitable overall). All this nonsense lifted from Fooled By Randomness and others about Russian Roulette etc just doesn't hold true about people with their OWN accounts. Perhaps institutional crackheads firing off bets might get lucky due to the depth of their roll (and thus ability to Martingale, but frankly this the only ability that would explain the idea of their trades being random yet successful) and Nassim Taleb's contempt for these individuals might be somewhat justified (though if only the supposed New Jersey uncivilised goon ('John' in his book) was truly as stupid as he wished). No individual trader without inside information got so lucky because there are not just two outcomes to each bet. There is no way you can run up 20 grand to 7 figures by luck with a 1% margin and 100:1 leverage even if you lived a million times over.

There are stories about people betting the farm on their very first bet and going on to do ok from there. I don't know, I often doubt these stories as being loose with the facts. I won my first bet. Then my second. Then went big on the third. Won that too. Then I lost it all twice from there. If I had made a million or a billion on my third bet, I still would have lost that right quick, the size of the win would have been irrelevant. I was humbled by my own confidence within weeks, not years, as I imagine most people were here. And even despite that, I was then humbled again and again and again and again and again........

 

[Pat494]

I I was humbled by my own confidence within weeks, not years, as I imagine most people were here. And even despite that, I was then humbled again and again and again and again and again........

The market does humble even the mighty imho.
One has to bear that inevitable strain on a daily basis to play the markets.

 

[random12345]

The market does humble even the mighty imho.
One has to bear that inevitable strain on a daily basis to play the markets.

Or learn how to backtest tick by tick instead of with OHLC

 

[Liquid validity]

It doesn't matter in the slightest whether markets are random or not.
Sometimes they are random, other times they most certainly are not.
Both camps can cite evidence proving their own point.
None of it matters.

The distribution of winning / losing trades is random.
Individual trade outcome is random.

 

[BeginnerJoe]

The markets are distinctively non-random. 90%+ of retail gambler joes loose, and the rest couldn't produce a statement. How can anyone point to a 90%+ certainty and call it random, or psychology, or any other convoluted exotic explanations ?

 

[the hare]

The markets are distinctively non-random. 90%+ of retail gambler joes loose, and the rest couldn't produce a statement. How can anyone point to a 90%+ certainty and call it random ?

Quite easily

Ray the random robot operating in "moron mode" can quite easily achieve a 90% win rate.

I wouldn't waste my time even thinking about doing that (other than 4 d lulz) unless you have some money you want to donate to your broker, then I can sort you out with the details.

Have you even tried applying any kind of statistical tests of randomness on time series data ?

Not that it matters, you are looking in the wrong direction anyway. I'm sure the penny will drop eventually

 

[BeginnerJoe]

Ray the random robot operating in "moron mode" can quite easily achieve a 90% win rate.

Who's Ray ? Can you demonstrate this robot ? But no need if you use this robot to trade demo or imaginary pips.

Have you even tried applying any kind of statistical tests of randomness on time series data ?

No. Can't say I am any good with stats, or wants to be in a game like this. Although there are many random things in the game, the game is not random.