Trends in random data
- Thread starter Ciatronic
- Start date
You cannot prove that prices move randomly. It is not scientifically possible to prove this statement. It is the same as trying to prove that life does not exist in the Universe except on Earth, or that there are no black swans. The only way to prove such statements is to check an infinite number of possibilities. There is probably a term for such statements in logic, but I can't think of it right now. It is also very difficult to define what is random, think about it 
In the real word prices are not random.
They are never arbitrarily scattered.
They are always grouped around the current price.
More buyers than sellers and that group is weighted toward the top end of the current price.
More sellers than buyers and that group is weighted toward the bottom end of the current price.
They are never arbitrarily scattered.
They are always grouped around the current price.
More buyers than sellers and that group is weighted toward the top end of the current price.
More sellers than buyers and that group is weighted toward the bottom end of the current price.
You made an inference about some data. You argued that the data are random. That is fair enough although, as someone noticed already, there can be no robust test of randomness. You can have only tests of the significance of the hypothesis that there is no evidence against randomness in the data. Make a note of this, it is a huge difference from claiming that the data are random.
Then, you jumped to make a conclusion about TA. Your conclusion is disconnected in a seriously fallacious way from your original premises.
(1) You have to prove that market data are random
or that
(2) TA methods do not result in systems with an expected return greater than 0.
You have not proved either of those. There have been many studies in the past. Many claim that TA does not provide an edge and many that it does. It depends on assumptions and understanding of the market. One thing is certain, that trading on fundamentals does result in zero mean returns unless you are an insider and you can get the information before anyone else or you can act faster than anyone else. Both of these assumptions are questionable, especially for retail traders. If there is an edge for trading the market, only TA can provided it. I am not saying there is. But if there is, fundamental analysis is useless unless you rely on asymmetric information.
Not weird at all. In fact, because it is a "random" function your graph could literally break off your supposed trend at any moment with no need for a catalyst to trigger it.
Not true of equities. Strong trends are often there for a reason and more often then not they continue until a new catalyst arrives to change investor sentiment. Weak trends however could signal a range bound stock.
Any serious trader can see important technical levels and understand the moves that occur at those levels will not be random at all. In your exercise, there are no important levels.
Your next result is totally random.
This is a chart of the close price for EURUSD 10/17/1994 till 7/3/2001
The Deutsche Mark is used as a proxy for the years before 1999
How much would you believe this trend to make a prediction for
7/4/2001?????
If you offset the line up you can make it a trend line or a channel if you also offset down. How much would you put if these lines gets approached or broken?
I think that this model is completely wrong for the data presented.
No, this was not the message that I wanted to send.
First of all, let's clarify:
I don't say that the market is random all the time. Actually I believe that it does have moments when is not random. I trade based on macroeconomic analysis. (So, I don't trade stocks > I don't need inside informations, as someone indirectly defined the FA). I won't enter to much in details of my style of trading.
I just want to say that actually I believe the market could be predicted(when trade based on macro, you rely on the so called "inertia" of the economy). Here is how I see the market: in the short term the market really become very random (high noise if I can say so) but on a longer term, the market tend to have a bias. Cuz it's common sense, the market adjust based on the "need". Let's say the inflation is heating(and inflation is quite cyclic, it doesn't change from having a bias up or down instantly, theoreticaly) then through different mechanisms/effects, the local currency gain in strength and the inflation is suppressed(this is just a plain theoretical example, we know what is in real time ) . So, the market have a bias in that moment because of the stage of the economy, the market is moving from point A to B, BUT between this, the market is quite random with a lot of noise.
And one more thing, if someone believe that the market is 100% random then, how could he expect to make profit? Because after you enter in the market you don't have a 50/50 chance of winning losing (with a 1:1 RR) , because of the spreads/commissions and whatever tax you are charged, so the ratio is let's say 49/50 W/Loss. You already have to pass this hurdle, so you need a edge. And you start from the premise that you can predict to some extend the market and cover that cost and on top of this, make a profit.
Basically, the edge should be extracted starting from the idea that not everything is random.
Finally, I just wanted to send a wake up signal for those who rely solely on silly TA patters. I went through this type of analysis about 3 years ago so I know pretty much what it does involve. Just look outside the box.
I just wanted to help, I apologize if I offended someone.
Regards.
First of all, let's clarify:
I don't say that the market is random all the time. Actually I believe that it does have moments when is not random. I trade based on macroeconomic analysis. (So, I don't trade stocks > I don't need inside informations, as someone indirectly defined the FA). I won't enter to much in details of my style of trading.
I just want to say that actually I believe the market could be predicted(when trade based on macro, you rely on the so called "inertia" of the economy). Here is how I see the market: in the short term the market really become very random (high noise if I can say so) but on a longer term, the market tend to have a bias. Cuz it's common sense, the market adjust based on the "need". Let's say the inflation is heating(and inflation is quite cyclic, it doesn't change from having a bias up or down instantly, theoreticaly) then through different mechanisms/effects, the local currency gain in strength and the inflation is suppressed(this is just a plain theoretical example, we know what is in real time
) . So, the market have a bias in that moment because of the stage of the economy, the market is moving from point A to B, BUT between this, the market is quite random with a lot of noise.And one more thing, if someone believe that the market is 100% random then, how could he expect to make profit? Because after you enter in the market you don't have a 50/50 chance of winning losing (with a 1:1 RR) , because of the spreads/commissions and whatever tax you are charged, so the ratio is let's say 49/50 W/Loss. You already have to pass this hurdle, so you need a edge. And you start from the premise that you can predict to some extend the market and cover that cost and on top of this, make a profit.
Basically, the edge should be extracted starting from the idea that not everything is random.
Finally, I just wanted to send a wake up signal for those who rely solely on silly TA patters. I went through this type of analysis about 3 years ago so I know pretty much what it does involve. Just look outside the box.
I just wanted to help, I apologize if I offended someone.
Regards.