Correlation, allocation in a portfolio. Calculation methods

 

Good afternoon dear forum users!

I have reached a stage where it is necessary to objectively determine the degree of asset correlation for competent portfolio diversification and mismatching signals in different markets and sectors (not only forex), not a market neutral strategy, just different phases and market conditions.

I have no doubts about the effect of the correlation analysis on the accuracy of the input signal.

I don't know how to check if the market is falling apart or not, but I don't know how to analyse it. I would love to hear your opinion!

 
Well if you have shares, each one is a separate market)

With a link to the real world, of course.
See how promising the company is and buy it.)

Better still, start your own company - it will be a more effective investment, imho.

For example, I suggested hiring a copywriter for marketeers - a real business.

I can sell aktsul - to 'expand' business, and the client does not go) From hand to hand to sell, without intermediaries)
 
Account_:
Well, if you have stocks, then each stock is a separate market).

With a link to the real world, of course.
See how promising the company is and buy it.)

Better still, start your own company - it will be a more effective investment, imho.

For example, I suggested hiring a copywriter for marketeers - a real business.

I can sell aktsul - to 'expand' business, and the client does not go) From hand to hand to sell, without intermediaries)
Correlation is correlation. A degree of repeatability of assets. Fundamental is not interesting

Are you saying that there is no correlation between stocks at all?
What about the correlation by branch?
 
CHINGIZ MUSTAFAEV:
Correlation is correlation. A degree of repeatability of assets. Fundamental is not interesting.

Are you saying that there is no correlation at all between stocks?
What about the correlation by branch?
I think there is no correlation) Because not all companies go down in times of crisis) And some even grow despite 'crises'. For example this one:

RU000A0DJ9B4, INGR

(Not an advertisement or a guide to action, just as an example).

Depends on the asset structure and a bunch of other clever words.

Only you need to buy them in advance)) and sell them in advance. Not exactly them, it's like a principle and approach to 'rascollation'.

 
Account_:
I think there is no correlation) As not all companies go down in times of crisis) And some even grow despite 'crises'. For example this one:

RU000A0DJ9B4, INGR

(Not an advertisement or a guide to action, just as an example).

It depends on the asset structure and a lot of other clever words.

Thanks for your answers. I meant a trivial method of calculating the correlation coefficient for a certain period. Let's say not with standard indicators, but with candlestick methods.

The other influencing factors can be a bunch, it is meaningless to take them into account in quantitative analysis, not fundamental in short.

Let's say growth in a crisis can be inversely correlated.

Suppose companies like Uber and Yandex would move as correlatively as possible in a calm environment.

 
CHINGIZ MUSTAFAEV:
Thanks for your answers. Really I meant a trivial method of calculating the correlation coefficient for a certain period. Suppose not with standard indicators, but methods based on candlesticks.

There may be a bunch of other influencing factors, it is meaningless to take them into account in quantitative analysis, not fundamental in short.

Let's say growth in a crisis could be due to inverse correlation.

Suppose companies like uber and yandex would move as correlatively as possible in a calm environment

In short, one should invest before the 'rise' and exit before the 'fall'. Charts will not help. Imho. Better to invest before the IPO in me (just kidding).
 

Account_:
Короче вкладываться надо до ‘взлёта’ и выходить до ‘падения’. Графики не помогут. Имхо. Лучше вложиться до IPO в меня (шутка).

Delete it, don't embarrass yourself.

 
CHINGIZ MUSTAFAEV:

Delete it, don't embarrass yourself.

That's fine.
Well, don't hit me.)
The point doesn't change.
You asked for an opinion and I answered ;)

 
Account_:
That's fine.
You don't have to be me.)
The point doesn't change.
You asked for an opinion - I replied ;)

Thank you for your answers.

 
CHINGIZ MUSTAFAEV:
Thank you for your answers. But I meant the trivial method of calculation of correlation coefficient for a certain period. Suppose not with standard indicators, but methods using candlesticks.

For the "trivial" calculation of the correlation coefficient, no indicators are used. It is calculated directly from the numerical values of the time series.
 
There is a problem with the trivial correlation calculation. Because of the non-stationarity inherent in price increments, it gives wrong (often inflated) results. This is why econometrics usually takes the difficult route of constructing an autoregressive model for the series.