Econometrics: bibliography - page 10

 
anonymous:


The algorithm is elementaryprovided that a cointegrated portfolio is found. The Johansen method should beused to find the portfolio.

2. To have a sufficient number of trades. If the threshold is too high - you will rarely receive large profits. If the threshold is too low, you will often make small profits. If the threshold istoo low, you will very often make a loss in the amount of transaction costs.

There are more ways, but they are moretypical for HFT bots, which are usually not written in MQL :P


it's not even about the threshold... in fact, such a threshold may continue to grow and grow - for a long time, even months and years :-)

what is more important here is the algorithm for lot management in one way or another... and also, why exclude the possibility to earn and on the Expansion itself ...

 
Aleksander:

Why don't you use Leonid's tool? Here's a triple index spread (spread of threeinstruments) for example

the description of the indicator - see page 67 of the Leprechaun magazine (10th issue of 2010)...

The basic principle of spreads over 2 pairs - see Quasi Arbitrage in Short Term Trading

http:// www. procapital. ru/showthread.php?t=28081

well, you can put more legs in the indicator using the same principle...


Let's not confuse science with the shamanism of the Chukchi who sing what they see.
 

Why don't youlike the methodology proposed by Leonid?

it takes into account volatility of legs, selects market-neutral positionof instruments... well, it's easily formalized for practical implementation...


and the fact that you're approaching the problem in a standard way :-) as in the first pages of this thread... you'll just get an unsteady series - and it's already visible in the drawings - lingering trends on the synthetics

absence of a long flat area (where it is most effectively to trade by arbitrage)

Well, I repeat - no algorithm for lot management - to bring the total equity of instruments in the portfolio to a "stationary" view ...


and without that... how can I put it... your trading won't be any different than trading one pair... with all the ensuing consequences :-)

 
Aleksander:

Why don't you like the methodology proposed by Leonid?

it takes into account volatility of legs, selects market-neutral position of instruments... and is easily formalized for practical implementation...

I don't like the whole TA
 
Aleksander:

in reality, such a threshold can continue to expand and expand - for a long time, even months and years :-)


This means that you have not built the portfolio correctly :P

it's more important the Lot Management algorithm for this or that development... and also, why exclude the possibility to earn and on the Expansion itself ...

And that means that your strategy will not be difficult to drain, because in this case it is no different from intuitive single instrument trading. If the value of the portfolio fluctuates in a small range - you will gain a small position and take a small profit. When the value of the portfolio starts to fluctuate very strongly in one direction - you will enter into a much larger position and incur a larger loss.

 

What is the Correctness Criterion? Once you have built a stationary form of equity on a certain period (window) of data, your stationarity will immediately disappear once you go outside that window...

see Recycle by Crenfix...

 
Aleksander:

What is the Correctness Criterion? Once you have built a stationary form of equity on a certain period (window) of data, your stationarity will immediately disappear once you go outside that window...

see Recycle by Crenfix...

Yes. It is named above and the literature is given.
 
anonymous: And thatmeans that your strategy will not be difficult to lose, because in this caseit is no different from intuitive single instrument trading.

why wouldyou lose :-)

like in this picture - where at a certain bifurcation we buy the bottom pair / sell the top pair, and when they close, the trades are closed...



at


if you don't close when instruments converge, juststart trawling of instruments - just like on "Split" - you'll make profit...
 
Aleksander:

why would you lose :-)

like in this picture - where at a certain bifurcation we buy the bottom pair / sell the top pair, and when they close, the trades are closed...



at


if you don't close when instruments converge, just start trawling of instruments - just like on "Split" - you'll make profit...

Why do you always put this picture everywhere? Show me a segment without any pronounced movements, like..... where a fluttering flat is. How many trades will be opened and closed in this segment and what will be the spread loss?
 
Freud:

Why do you always put this picture everywhere, show me a section without any pronounced movements like..... where there is a rattling flat, for example. how many trades will be opened and closed in it? and whatwill be thelosses on the spread .

please... here's a rattling flat....