From theory to practice. Part 2 - page 51

[Aleksey Nikolayev]

sibirqk:

I also wanted to analyse movements on small TFs - e.g. minute candles, in order to predict bar characteristics on a large TF - e.g. a day. I.e. try to predict Hg, Lw, and direction of Cls-Opn, say, a day candle by previous movement on a minute TF. And then somehow refine the prediction based on analysis of the daily chart. It is clear that one cannot gain a great advantage, but something acceptable may be obtained using a portfolio on several symbols. But I have not got down to it yet.

This is a very interesting question - how exactly large movements are composed of small ones. I studied it a bit on the basis of how a big zigzag is made up of small ones, but did not find anything interesting.

[Aleksey Nikolayev]

Evgeniy Chumakov:

The problem is that small movements are more likely to reverse , and large ones in a trend take away most of the profits of small reversals.

My research (based on zigzag) shows that if compared to SB, small ones are more likely to continue and large ones to reverse. Considering the spread, the difference is not very significant. But since there are quite a few small movements, it is always possible to try to select from them those that are more interesting.

I have already written here before about the movements that may be interesting to enter in their direction. I also wrote that it's just about trying to undercut the drawdown of Shurik's original system (by adding an auxiliary system), which may allow for building up volume.

[secret]

Aleksey Nikolayev:

There is too much to do to try to get Shurik's idea right.

And what is his fruitful debut idea? I, for one, do not see it.)

It's not even a return model, it's just a stationary model. It's a long way from the market.

[denis.eremin]

secret:
And what is its fruitful opening idea? I, for example, do not see it)

It's not even a return model, just a stationary one. It's a long way from the market.

A smooth non-stationarity model would work if the variance and IR varied smoothly. Then we could play with the size of the floating window of the model

[Renat Akhtyamov]

I finally found the fucking arbitration

[denis.eremin]

Renat Akhtyamov:

I found the fucking arbitrage.

You found it again? How many times do I have to find it - it's been a year now.....

[Aleksey Nikolayev]

secret:
And what is his fruitful opening idea? I, for one, don't see it)

It has been very fruitful in the sense of providing an opportunity to scratch tongues on the forum) This is very important in terms of socialising traders)

secret:

It's not even a return model, just a stationary one. It is quite far from the market.

When the market is stationary, the reversion is automatic and it is always resilient, therefore it is looked for in spreads and portfolios.

With non-stationarity, it may well be present too, but it may be in a non-persistent form (as in SB, for example).

[Renat Akhtyamov]

denis.eremin:

You found it again? How many times do we have to find him - it's been a year now.....

if i knew, i'd be living in sochi ;)

[Evgeniy Chumakov]

From theory to practice. Part 2

Everyone sings about something different. Part2

[secret]

Aleksey Nikolayev:

With stationarity, there is automatically a return

I see, I was referring to the correlation of increments)