Is it better not to trade at 11 p.m.? - page 2
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
Well, then the statistics need to be interpreted differently.
What does that figure even mean? And if 100 times less trades were opened in that 23rd hour (by volume) than in any other hour, what difference does it make that 85% are losing? What does that figure even tell us?
zaskok3:
I'm monitoring all the glasses a little bit. The toolkit is weak so far. But the same limit orders do not trade much (lot < 0.5 mainly). However, in the evening more professionals become active; among them there are a lot of PAMM and ATS traders. Limiters are considerably enlarged. For example, since midnight someone reduces ~150 lots on many symbols. It seems that due to the lack of margin they remove the limit (when the price is far away from it) and put it back (when it is close to it). This manipulation allows opening for a larger lot.
Made a little bit of progress in analysing the trading of the trading floor clients. For many hours (and still is) one guy was scraping EURCHF up to 200 lots per side, breaking the net position a bit on several price levels.
This was the most active pair. Then AUDCHF, there was also scalping there, but a few people. Basically, it is quite good to see where their limiters are triggered.
The scalper on EURCHF may be primitively shown this way:
Of course, more clearly - visualise in the form of price levels on history. Reengineering in action, in short.
Reengineering is not even necessary. It is enough to identify the profitable trading system of someone else, and then there is no need to understand how it works. It is possible to copy trades of that type, setting limits on the same price levels.
That is, we write TS, which finds in the market someone else's TS and blatantly copies it, sharing the profit with the exact sign of the profit. On the exchange, such identification would be more difficult.
For many hours (and even now) one guy has been scraping EURCHF up to 200 lots per side, breaking the net position a bit over several price levels.
We need each ECN/STP-platform must be presented in the form of separate units: requests of PL (STP) and limit orders of clients (ECN).
And understand, to whom (PL or client) belongs each of these summands.
This cannot be done based only on the value of stakes at one moment as there is not enough information. But taking into account the dynamics of changes in the cup - it's quite possible.
For example, I set a Limiter at 24.56 lots far from the current price. Now I can determine at once that it is the client's limit. But when the limit is approaching to some critical distance from the current price, there are LP-applications that add more volumes to my bank.
I see something like 37.06. I can see that my limit order at 24.56 and 13.5 lots are in this volume. I.e. speculatively I can make such a division. These 13.5 lots most likely belong to some LPs and possibly partially to other clients.
All in all, what is needed is an algorithm on how to break down the gangs into summands and identify their ownership.
So far I'm digging into the fact that client limiters, unlike LPs, hang around for a long time. Hence it would seem so far intuitively, that one can find out the additional information. Visually it is as if it can be seen.
Maybe someone knows, where the solution can be found. Exchangers will hardly bother with this, because they will debit each band to separate summands according to warrantlog data. But there are no warrantlogs in darkpools, so the solution can only be approximate and algorithmic.
There are no orderers in the darkpools, so the solution can only be approximate and algorithmic.
What is the value of such approximate data?
Any one of the applications could be both client and PL, right?
And longevity can only be analysed based on the assumption that many clients hold applications for a long time. But do they?
What is the value of such approximate data?
Any one of the applications can be both client and PL, right?
And longevity can only be analysed based on the assumption that many clients hold applications for a long time. But do they?
An excellent and easy way to algorithmize the partitioning of requests into two non-overlapping subsets: LPs with clients and clients without LPs. See figure.
It is well understood that it does not make much sense to focus on LP bids, because in the darkpool some sides of LPs are stupidly turned off by darkpool itself to level out the resulting position distortions.
However, it does make sense to look at client bids, and not a small one at that.
Most of the time it's the strong guys who deal in limiters. The big limiters are the pros. And they will not go to the limit for nothing, especially if they are not PAMM traders. When I see a 200 lot channeler trading around the clock making about 1 billion trades a day (by a rough estimate), it can't help but arouse at least some interest.
Obviously, this client earns an average of >> $10K per day at that rate. He shines like a Christmas tree at night as the trading is frantic compared to the rest of the pairs. It's profitable (lots slowly but growing) in a rather difficult current market for a channeler.
Great to see when other pros are activated. Which pair and how they trade. I.e. I can see where the probability of profitable patterns is high. I can see not only incoming/outgoing positions, but also intentions - I can see the channels. And therefore I can reengineer.
If I can programmatically identify my clients' Limits when they are very close to the current price, then I can even impudently (without their knowledge) copy their trades. At the same time, no Signals will be able to copy like that, as they are not able to work with limiters.
The word "can" is not 100%, as LPs sometimes spoil the raspberry on identification. But trying algorithms to be able to separate the flies from the cutlets. In terms of lifetime of client gangs - unambiguously, client gangs live orders of magnitude longer. I monitor it all.
I have not yet written an indicator for displaying trading history of profitable traders. Slowly glued, as he seems to be a pioneer in this business. The bicycle completely has to be invented by myself. Google says that forex stacks, if ever they tried to analyse it, not publicly.
Sometimes I see 300-lot Limiters placed very far away in some exotic symbol CADJPY. If it is possible to analyze these guys, they need a very large history. But channelers are more simple.
There are scalpers, but they are not interesting, because they are cheaters. The newsmakers are not detectable, because there are almost no limiters.
I have a code, that shows all tumblers and writes statuses on them. But the main thing is that it is a working kernel/bicycle from which further research can be done. The topic is unploughed, from the word "nothing". I cannot see trading of profitable large high-turnover traders indifferently. Motivating is not a bad thing.
Now coming up with algorithms on a piece of paper on how to break down into summands and identify them by affiliation. It is really interesting, because you understand very well that you are not writing another bullshit indicator, but the necessary toolkit for market analysis. And it will definitely help. It has already helped me.
To do this you need to know:
1. What was the profit of the last operation (lots and entry points are known from the cup).
2. How much the lot has changed after the rollover (also known).
I.e. in real-time and on the history, you can see not only the channels themselves, but also equity.
It sounds simple, but it's a bit more complicated to program... I didn't think it was possible to get that kind of information out of the stack.
In fact, if you trade large channel lots via limiters, you are in the palm of your hand (there is more information about your account than on monitoring services).
So far I have found two options: one and two.
I want to use normal third party visualisations to look at my data. Who else knows, throw me a link.
About what, about the time. At that moment, history was making its way, the price was choosing its direction.)