For those who have (are) seriously engaged in co-movement analysis of financial instruments (> 2) - page 13

New comment
 
Mathemat:

It will not work. For it is pure empiricism without any theoretical justification.

It doesn't come out stone flower. There is no guarantee that the "current optimum synthetic" will hang around in the channel you assume. And so, too, refills to the "current optimum" may sooner or later turn into trivial dilutions, which are fraught with danger.

not a fact, conventionally, see delta hedging... the essence is the same....
 
Mathemat:

It will not work. For it is pure empiricism without any theoretical justification.

It doesn't come out stone flower. There is no guarantee that the "current optimum synthetic" will hang around in the channel you assume. And so, too, refills to the "current optimum" may sooner or later turn into trivial dilutions, which are fraught with danger.

If you are a trader and know how to trade and handle synthetics (what a clever word they came up with!), you may want to ask for a disguised lock.
 
Mathemat:

It will not work. For it is pure empiricism without any theoretical justification.

It's just unsubstantiated, unfortunately.
 
hrenfx:

And now for a clear definition of cointegration, which for some reason is nowhere to be found.

Would this not work?
 
genro:
Is this not good enough?

It won't work. We are not dealing with infinite BPs. We only have samples of BPs at our disposal.

The need for cointegration to extract a profit is theoretical bullshit, as is the need for stationarity.

Moreover, there is nothing remarkable about my definition of cointegration sampling for us. So you found a cointegrating vector for sampling. Why on earth would such a synthetic use market interrelationships rather than being another contrived stretching of the market to fit your criteria! You can create all the synthetics you want, but there has to be a basis which will use the market interrelationships. Without such a framework, all the work would be for naught.

I am not claiming that Recycle makes use of market interconnections. I only have certain arguments in its favour:

  1. Finds majors and crosses without using the names of financial instruments.
  2. Shows that crosses don't carry any additional information. And it does.
  3. For the two dimensional case it is fully consistent with QC.
  4. Portfolio theories, correlation and regression-dispersion analysis are not used. Books on which are written by innumerable and in vain.
  5. Versatility, out-of-the-box approach and relative simplicity.
  6. The whole market is analysed.
  7. The result does not depend on the order of prices of each instrument. You may mix up the time as you wish, creating gaps, etc. It will not affect the result.
  8. ...
 
hrenfx:

Synthetics can be created all you like, but there has to be a basis that will use the interconnections in the market. And without such a framework, all the work is worthless.

I am not claiming that Recycle makes use of market interconnections. I only have certain arguments in its favour:

  1. For amateurs like me in linear algebra and correlation-regression-dispersion analyses, could you explain popular science or "on your fingers" how weight coefficients are calculated in your Recycle indicator, i.e. what is the minimum eigenvalue of the covariance matrix composed of columns of the source matrix?
  2. What kind of market relationships do you have in mind? I.e. by creating and calculating, synthetic we should choose FIs that have some kind of fundamental economic relationship ?

For example: "It is known that the Australian dollar can be traded in the same way as gold. Australia is the largest producer of gold in the world. The Australian dollar has a very strong positive correlation to this metal and when the price of gold goes up, the Australian dollar also strengthens. The New Zealand dollar, whose economy is closely tied to the Australian economy, is even more strongly correlated to gold than the Australian dollar. This fact can be used in trading."

Or: "There is a clear correlation between oil prices and the Canadian dollar. The price of oil acts as a leading indicator of Canadian dollar price activity. When oil prices rise, the Canadian dollar exchange rate also rises. "

 
Aleksander:
not a fact, conventionally, see delta hedging... the gist is the same....


Delta hedging is a dynamic hedging strategy that uses options and continually adjusts the number of options used as a function of option delta. In delta hedging, the number of options is purchased such that the exchange losses arising from equity trades can be absorbed by the exchange gains from put option trades.
This assessment is based on the delta coefficient for the relevant option. The delta coefficient shows how much the value of a put option increases if the share price declines by 1 point. The delta coefficient is between 0 and 1 and changes with the share price. The delta value of a put option increases as the share price declines and is closer to 1 the further away it is from an option whose strike price is more favourable to the buyer than the current price of the underlying instrument. In a delta hedge, the gain from a put option is equal to the exchange loss from a change in the stock price if the stock price declines. If the stock price declines sharply, the loss is clearly outweighed by the gain in the option. As the stock price declines, the option position is reduced according to a predetermined delta value.

Explain the point for application to the trading options discussed here.

 
genro:

  1. For amateurs like me in linear algebra and correlation-regression-dispersion analyses, could you please explain popular science or "on your fingers" how weight coefficients are calculated in your Recycle indicator, i.e. what is the minimal eigenvalue of the covariance matrix composed of columns of the original matrix?

It is an analytical solution to the previously given matrix problem. You can watch and use TV, but I can't tell you "on your fingers" how it works.

  1. What market interrelationships are you referring to?

"Linear" (inverted commas, so as not to reduce everything to a flub of terminology). The flow of money from one FI to another is not by polynomial laws, trigonometric etc, but "linearly". If there are only two FIs in the market, the growth of one will cause the same decline in the other. But the world market is far from two FIs.

So when creating and calculating a synthetic we must choose FI which have some fundamental economic relations?

When creating a synthetic, we have to analyse the whole market - all FIs.

 
You could look for the truth in a vector of volumes orthogonal to the cointegrating one. But I don't know how to justify it, it's just on intuition level. And it will no longer be statarbitrage, because the equity process of the resulting synthetic will not be stationary.
 

Why even complain about the lack of stationarity?! It's the same as complaining about the lack of an elementary grail.

There are no cointegrating vectors for majors. The number of vectors orthogonal to a given vector is infinite.

1...67891011121314151617181920...38
New comment