Options

Search
Search
Search

Subscribe to Thread

Bookmark Thread

First Page First Unread Last Page Last Post

Printable Version

Similar Threads

Determining randomness in a trading system 35 replies

Best book on randomness in markets and market probabilities? 4 replies

MM - Using randomness against Asymmetric Leverage 12 replies

Thoughts on randomness and systems 40 replies

Fooled by randomness 13 replies

Market Randomness

Quoting OldQuant
BTW, "fat tails" isn't a good search string; the hits will be more porn sites than discussions of generalized characteristic functions. But by the time you get through that list, maybe some of you would be ready for a little of that anyway.
Nice scam, scammer. I just googled "fat tails" hoping to perfect my knowledge of human nature and of course not a single result is related to porn. I never really believed it, but if FF commenters are as bad at trading as they are at porn, it does explain the 95% losers ratio.
This tiny comment about "fat tails" shows how useless some comments can be and how people hide their ignorance by using some obscure vocabulary hoping to impress young traders. I am not referring to anyone in particular.

Oh well, I guess that's what the over-educated would refer to as "bullshititis", also known as random people posting random non-ideas on random boards on the internet.
I'm in a bit of a hurry but just want to quickly say thank you OldQuant for the valuable (and amusing) input. Same to you Adal.
Quoting OldQuant
The question then becomes: "do fat tails produce tradable opportunities?" My opinion is: yes, they do.
I second that.

In fact we got one profitable system in FF which is taking advantage of fat tail.
Market randomness, Santa does really exist.

Learn how to trade and markets will seem less random. But that takes a little effort...
Cheers!
E
Just to come into the discussion and lay down my thoughts regarding the subject.

I'm not trying to be a smart ass here, but my principles on how I capitalize on the market action may look to one to simple.

I don't really want to get into whether or not markets are random or not but I believe though in treating them as random we are losing any edge we could get.

do not take any offense here anybody is just my outlook.

They may appear random but I would say they certainly are not. You will see why, I'll demonstrate my thought process, if you stay with my reasoning.

I bealive that it can often appear that they are random, but is just a confusion. Every time price move up or down,even by one pip,then someone for any reason is participated in the market. So therefor reason opposes randomness. Price behave like it does because of some reasons, the move is reasoned, not random.

However,the movement of price might be perceived as random by anyone who does not know why price just acted as it did. Markets are about price and supply and demand.

So, my view is that markets are random if you do not know why price is behaving the way it is.This concept applies to other areas as well,take football for example.If I had never know anything about sport,never played or watched,and was taken to a game, I might make a conclusion that the players are acting out of randomness.

For me it whould look like they are just throwing the ball for no reason,running in circles. I would not be able to make any predictions about future behavior of the players at this time. If, however, someone next to me explained the rules of the game and why people were behaving as they were, then it would make much more sense to me and no longer be random.

It would be reasoned, and I could make assumptions on what players are going to do next, based on this reasoning.

So, to rationalize the market, it is my belief that you must know why price is moving, otherwise you are trading randomness. If you can understand even one reason why price will behave a certain way at a certain point or at a certain time, then you can profit from this, as it is no longer considered random.

Leave the mathematical randomness model to the math purpose.

There are people out there who don't know nothing about this and yet they make money day in day out.

We are selfdestructive by our own human nature, everything we touch we tend to complicate and forget that simple basic stuff can guide us to resolve the most complicated things we face.

all the best!!
Quoting OldQuant
To frame a discussion strictly in "randomness" terms is too broad brush....
wow, I just wanted to thank you for your valuable input. most insightful reply! I was initially quite undecided but your comment on the fat tail issue really nailed it for me. Interesting.
Quoting triger88990
I bealive that it can often appear that they are random, but is just a confusion. Every time price move up or down,even by one pip,then someone for any reason is participated in the market. So therefor reason opposes randomness. Price behave like it does because of some reasons, the move is reasoned, not random.
I understand your point, but ask you to think about it this way. Go to any of the currency pair threads while trading and traders are active (say morning). In a hour of posts, count how many traders are posting opinion of "long", backed with their carefully done charts, PA analysis or fundamental views. Then count how many are posting opinion of "short."

All have their reasons. Now pretend that they are each analysts for large funds. (The real analysts, as a group, are probably not any better than those posters are as a group), and pretend that each opinion will cause an order of some millions EUR buy or sell. The NET (say it is "buy") of their opinions/orders will move the market up if there isn't enough size being offered at the last asked price to match that net long.

Now ask yourself two questions: 1. "Could I have predicted the net position that was implied by the traders who posted?", and 2. "Having observed these opinions this hour, can I predict what the traders' net opinion will be next hour?"

If you have to answer no to either or both of those questions, then you must accept that their opinions would induce a random change in price. Even though those traders are rational and intelligent, they are reaching different conclusions, maybe using systems they are keeping secret, so we couldn't "front run" their analysis. Since you or I have no way to predict what a group's individuals will conclude, there will be no way to predict the market effect. And in that sense, reasoned actions translate to random result.
Quoting OldQuant
If you have to answer no to either or both of those questions, then you must accept that their opinions would induce a random change in price.
Not random. Chaotic.
Quoting Adal
Weather looks random, but some are able to predict it with more or less success.
Weather is a great comparison, as like the market, there is nothing random about it. Every outcome is the sum of its parts. Our ability to predict the weather is only limited by our understanding of how it works (and the forecasters do a far better job than many seem to realise).

Quote
What I'm trying to say is that if we, retail traders can't find order in the market, it doesn't mean that it is random.
Precisely, it just means we lack the ability to process the information, or lack access to the information.

Quoting Marv
Yes, if I for instance single you out and track your trading behavior, it would be anything but random. But when you add up the influence of too many people (and systems) whose actions are based on all kinds of sentiments, information, methods, etc. etc., the end result would pretty much be random.
It still wouldn't be random, it may just appear random. The outcome would be the result of all actions by all participants. Like I said, a human does not act randomly, so a group of humans cannot produce a random result. In fact, a group of humans is far easier to predict due to the influence the group has on individuals.
Hmmmm... That stable chart animation and commentary seem familiar.

Since a few posters want to explore this subject in a bit more detail, I invite you to add one more term and its common synonym to your vernacular:
bias and drift. I think that will help clear up some of the disinformation being mistakenly propagated in this thread regarding markets requiring zero long term mean in order to be declared 'random.'

As you said, however, there are many different classifications and interpretations to properly describe randomness. First, one must begin to familiarize themselves with the very simple concepts to even begin to understand why more complex models are uttered: pareto-levy, chaotic, stable, etc..

Regarding chaotic, one must be very careful with precise definitions as well.
A truly chaotic system has an underlying deterministic order embedded within. Markets are closer to quasi-chaotic in my experience.

As others have noted, fat tail distributions (which are in many ways far worse than gaussian) have certain expectancy properties that make them quite useful (depending on where you
happened to have entered and how you scale your betting).

Regarding the old argument about 'people' moving markets in a non-random way, I'd respond that turbulent fluid flow yields to the laws of physics, yet
the casual observer would have quite a difficult time predicting even the next few particle trajectories. Ask yourself, does that make it non-random to the participant/observer?
------------------------------------ ------------------------------------ -

Quoting Marv
The charts themselves were simply meant as something to get people to think about this. I want them to forget about patterns and pictures in the clouds for a second, and just think about probabilities, expectation, and central tendency with large numbers.

As for actually empirically proving that market prices are a stochastic (random) process, there are better ways to do that.



You are right. The markets are not Gaussian.

In reality, the markets have varying stable distributions, which is even worse than a simple Normal/Gaussian distribution...
Quoting Mr J
Our ability to predict the weather is only limited by our understanding of how it works...
...it just means we lack the ability to process the information, or lack access to the information.
But for a trader, what's the operational difference? Unless a sequence can be predicted, we can say 'random' as a synonym for 'unpredictable'.

Here's an old example of the difference between apparent random and non-linear generated (and one which also argues, in its way, for the pattern recognizers, and the guys who seem to recognize 'cyclical' behaviors where the analytics won't find them).

Some "random" digits, off the top of my head: 5,9,2,6,5,3,5,8,9. Say odd numbers represent 'long' moves, 1 the smallest, 9 largest; even numbers represent shorts, 0 to 8. What's your prediction of the next move?

Likely, the guesses of a group would be all over, averaging around 5. But then some somebody takes the numbers, removes the commas, pre-pends the digits 3141, adds a comma after the three, gets 3,141592653589, identifies a plausible generating process, and says "7."

Maybe the digits were random, or maybe they were part of the expansion of Pi. Seven would be an odds-on bet, if you recognized the match to that generating function. If not, though, they still will look, feel, and smell like a random sequence to the rest of us, with all the risk that implies.
Quoting OldQuant
But for a trader, what's the operational difference? Unless a sequence can be predicted, we can say 'random' as a synonym for 'unpredictable'.
If logic is consistent I expect there likely would be. I consider there to be a great difference between concluding that something is unpredictable, and concluding that I am not able to predict it.
Quoting jamjamjam
I invite you to add one more term and its common synonym to your vernacular:
bias and drift.
Nail. Head. This is, IMO, why and how trend followers prosper. In market applications, the focus regarding the fat tails is on volatility, not drift. That's probably because the market segment which has the most to lose from the fat tails are the exotic option writers, both currency and interest rate, and their modeling quants use risk-neutral (static drift) models to figure out the hedges. But the trend followers can get in, take their bumps whilst they wait for the drift to turn their way with a vengeance, and then stay there until the black swan lands.


Quoting jamjamjam
...fat tail distributions (which are in many ways far worse than gaussian) have certain expectancy properties that make them quite useful (depending on where you
happened to have entered and how you scale your betting)
I would go so far as to say that one of the main reasons so many individuals lose money, when they first come to retail forex with small accounts, is being encouraged to use pre-defined TP's rather than being taught how to use trailing profit stops or, e.g., PA based exit rules.

I'm not going to say much about this, because I'll get badly flamed for flying in the face of the conventional doctrine of money management.

But the fat tails are the sub-martingale offset (wiki it if need to) to the overly frequent choppy market where your loss stops get hit more often than it seems they should. It doesn't matter if the particular black bird in question flies through the body and extender of a 60 tick, 10M candle, or over enough time to be called a trend. Either way, if you have a relatively tight TP, you don't participate in that bird's flight. And if your order is resting in MT4 where your market maker can see it, he knows the risk he's facing, and probably doesn't even have to hedge you off.

Like Forrest Gump said, that's all I'm gonna say about that.
I get a taste of your reasoning, and it seems to me that you are high educated person. I myself was in love with the math,thats not more true now , I'm an engineer by profession and math was my job, but when it comes to markets I try to forget what I know except the elementary math which we stil need it .

but I'm feeling that each of us are trying to convince others that he has more valid arguments than the other,and if he didn't get our point we bring up more arguments to sustain our own arguments,

(I know is a paradox because thats what I'm trying to do right now,isn't so?)

instead of just leave our rigid mind at the door, and try to forget for a time what we know and to build up into others oppinions.

I think that doing so we can gain a different perspective. I always start not only when it comes to trading but also in my life, with the assumption that I may be wrong. In trading must sound that I'm defensive with my positions, but if market flow is moving with me I can be the opposite, very offensive.

I'm not trying to convince anyone of nothing, is just the way I see it.

You first need to be caution because I'm the worst teacher on planet Earth

When I try to put together some pieces to explain to someone else my process thought, I tend not to be understand, but thats my fault is not because of you.


Quoting OldQuant
I understand your point, but ask you to think about it this way. Go to any of the currency pair threads while trading and traders are active (say morning). In a hour of posts, count how many traders are posting opinion of "long", backed with their carefully done charts, PA analysis or fundamental views. Then count how many are posting opinion of "short."

Now to come back at your post, firstly:

I think that you start this wrongly, markets doesn't care of opinions, it does what it does, you must take into acount other factors here such as market structure.

{ I'm in opinion that you must get the concept of market structure right to understand the charts,and this will be also the key for achieving the necessary mental skills everyone would need. }

As i say in my previous post that markets may seem random if you do not know why price is behaving the way it is.


Quoting OldQuant
All have their reasons. Now pretend that they are each analysts for large funds. (The real analysts, as a group, are probably not any better than those posters are as a group), and pretend that each opinion will cause an order of some millions EUR buy or sell. The NET (say it is "buy") of their opinions/orders will move the market up if there isn't enough size being offered at the last asked price to match that net long.

Now ask yourself two questions: 1. "Could I have predicted the net position that was implied by the traders who posted?",...
"Random" is a word invented by people to describe irrational and unpredictable behaviors. Randomness doesn’t really exist in nature. Every thing can be explained even if we don’t know the explanation for it.

Its true that some phenomenon’s are impossible to explain with the tools we have but it doesn’t make them random.

I think that because of what drives the markets, which is people and not the roll of a dice, predicting them it can be done. Although one could argue that such an amount of different strategies trading the markets can produce a randomness, I will say that people are predictable and the herd phenomenon created by a crowed is even more rule based.

The problem will be with the coming of the automatic trading machines(based on mathematical models). They don’t act as humans and there for harder to predict. When they will significantly out number humans in trading, that’s were troubles will begin.


all the best!
Quoting triger88990

but I'm feeling that each of us are trying to convince others that he has more valid arguments than the other,and if he didn't get our point we bring up more arguments to sustain our own arguments,
haha, like I've been saying.. were all guilty of misusing the word "random" and taking it way to literal.. people die for pointless reasons taking biblical text to literal

when I say "theres nothing on tv" I don't literally mean that every channel is off the air

"she cheated on me.. no literally I woke up and she was having sex with another guy right on top of me" lol

I wish humans would just learn to understand one another for the concepts but yet we are all guilty of taking things way to literal

I think we are all talking the same concepts here people
Quoting triger88990
The problem will be with the coming of the automatic trading machines(based on mathematical models). They don’t act as humans and there for harder to predict. When they will significantly out number humans in trading, that’s were troubles will begin.
second that

taking profits in a zero-sum or a negative-sum environment means exploiting inefficiency in market behavior better then the competitor, humans are far less efficient then machines...
one way or another as science progress brokers fees will be the killer for small time traders as market becomes less exploitable
i personally think that flipping the coin is not random. from what i read random.org bases it's randomness from atmospheric noise. is that really random? the thing which puzzles me is that i manage to find transactions on the charts which you presented. i found 3 and they all were successfull. if you can generate candlesticks i think that i can find more.
Something for the 'impossible' to beat a random walk camps (Hanover?)=)

http://intelligenttradingtech.blogsp...sible-you.html
One thing missing from the random charts is time scale. Based on the type of currency one could easily distinguish a randomly generated 1 hr chart from the actual by looking in at the volatility (bar size). For example EURGBP goes into very tiny ranges between 22 - 0 GMT most days. So volatility seems to be the edge we could exploit.
Quoting yesbee1
One thing missing from the random charts is time scale
that is a good observation. It has me thinking

if you take the random charts and measure them with time from the first tick to the last, and created a candle out of each one (open, high, low, close), each candle will be different height, size, range, ect... and put them together to make a "higher time frame", maybe even break these random charts down and group them into multiple time frames, do you think that important support/resistance levels will come into play? (just a question, to be honest it kind of hurts my brain to think about it, lol)
Thread Tools Search this Thread
Search this Thread:

Advanced Search

0 traders viewing now