Zero sample correlation does not necessarily mean there is no linear relationship - page 57
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
This gave an error in the hedge ratio of 0.07%: it should be 1500.6429 instead of 1499.520. How to go on living?! :(
The purity of the experiment. But that's not the point.
You are in the market as long as the balance is above (below) zero. Based on the chart, that's not a long period of time. Did you open on trend, against, sideways? Profit from where?
Won't you get fat?
No. This is the only way you can understand that there is no mythical cointegration in the market and that CCs are used to build TCs.
Learn and learn again - there is no other way.
If the QC is taken from I(1), it basically shows whether both series are going up or down on average
Here is a graph of sine(x) and sine(x+pi), for example
CC=-1, as you might expect
add a linear trend to both graphs
KK = 0.61 and depends on the slope of this linear trend. And if we increase the number of members of both series by 2 times, keeping the same formulas, the KC becomes 0.88.
That is, for a cumulative series, the reciprocal position of the linear regression is strongly influenced
The formula shows that the QR is affected by the position relative to the mean. Roughly speaking, this is the MA for the last term of the series with a period equal to the length of the series. If the row members are on the same side of their averages, the AC increases, if they are on opposite sides, it decreases.
And of course there is every chance that the 2 SBs will both grow on average on the calculation window, for example, and then the LR will both be up and the QC will have a positive value depending on their angles. False correlation))
Although, assuming that the trend tends to continue on the quotients, the QC will reflect the mutual position of the trends (LR) of these series and may retain its values in the future. That is, the QC for I(1) compares the trends of the 2 series in the calculation window
This is the only way you can understand that there is no mythical cointegration in the market
Do you understand the difference between the fact of cointegration itself and strategy? You don't need a strategy to confirm the presence of cointegration. I have statistically confirmed to you the existence of cointegration. If you argue that there is no cointegration - show me where the regression residuals have a unit root.
The strategy, in its simplest form, is to calculate the Bollinger Bands from the regression residuals and trade on a return to the mean.
and use CCs to build TCs.
After losing an account they cry on forums that "statistical arbitrage doesn't work".
Wouldn't that be too much for you?
Then, when you try to create a TS and understand what's what and why the spread, etc., do the following - expand the analyzed time frame. Because there are smart people that do not understand that prices on financial markets can be cointegrated for a short period of time. Moreover, the beginning and the end of this interval cannot be predicted. Non-stationarity, you see, sets in.
And then again - read, read, read. And study!
Anyway, you have enough tasks for now.
Then, when you try to create a TS and understand what's what and why the spread, etc., do the following - expand the analyzed time frame. Because there are smart people that do not understand that prices on financial markets can be cointegrated for a short period of time. Moreover, the beginning and the end of this interval cannot be predicted. Unsteadiness, you see, sets in.
And then the owl - read, read, read.
Anyway, you've had enough tasks for now.
I don't get it.
Why read a statistics specialist? A sea of books, articles .....
About integration he writes just nonsense.
A sea of books, articles .....
The ones about integration he writes are just silly.
Post the ones about financial markets.
Write the same article, but better. It's all right to criticize..........
And, moreover, the onset and end of this interval cannot be predicted. Unsteadiness, you see, sets in.
Where does that come from? From non-stationarity? So cointegration for non-stationary series only makes sense.
Where does that come from? From non-stationarity? So cointegration for non-stationary series only makes sense.
The stationarity of their linear combination is violated, that is also the point of cointegration
The stationarity of their linear combination is broken, which is also the point of cointegration
How does it follow that these moments are unpredictable? In fact, that they need to be predicted in order to earn