[Trader's Handbook] Draft articles, "out of pocket" discussions - page 21
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
Probably a little late with the questions. But I will ask them anyway, and I will be very grateful for the answers.
After "satisfactory" results on the tester the next steps are: demo account, and real cent account.
What is the model of such emulated accounts? In particular I am interested in ECN/STP accounts. How does the brokerage company use e.g. large order placement/execution on the demo account, when the real account at this time is "calm"? To what extent does the sensing mentioned, for example on the cent account, correlate with the same operation on a standard (1:100) account? In short, what is common (especially when working with limiters) between the execution of orders on these accounts and execution on regular ones?
I've looked at smartlab (the mouthpiece of stock market communication) now. The impression is that stock traders are no more literate than forex traders. They demonise HFT, talk about puppets, etc. Almost all clickers who are convinced that the stock market is right and forex sucks.
If they were slipped such a libel like this, it would almost be an insult to the feelings of the faithful. MT5 will 100% kill a lot of people on the Russian fund, but I wish it was also competent.
I've looked at smartlab (the mouthpiece of stock market communication) now. The impression is that stock traders are no more literate than forex traders. They demonise HFT, talk about puppets, etc. Almost all clickers who are convinced that the stock market is right and forex sucks.
If they were slipped such a libel like this, it would almost be an insult to the feelings of the faithful. MT5 will 100% kill a lot of people on the russian fund, but I wish it was also literate.
it is stupid to train meat unless it is a circus )
i have three pigs in my flock, i don't teach them the alphabet because i know i will eat them.
As an aside:
Fraud option: PAMM pyramiding.
Forum on trading, automated trading systems and trading strategy testing
Traders Guide: Orders, Prices, Deposits, Funds, Currencies
hrenfx, 2013.06.10 14:37
A simple HFT MM algorithm.
Obviously an ECN/STP aggregator has some insider information: who trades through it and how. A competent owner of such information creates a classifier of clients' profitability. For example, as a result of the analysis he will know that aggregator's clients steadily (and at a proper speed) lose on GBPJPY from 12:00 to 17:00 every Wednesday. What it means they are draining is that their money is flowing into LPs which had the most competitive prices at the time.
IMHO questionable analysis results. In other words, the conclusion noted can be stated as follows:
- from 12 to 17 every Wednesday, market behaviour will be such that clients will drain, i.e., in this case it is a market prediction,
or another way of putting it:
- From 12:00 to 17:00 every Wednesday, the behaviour of the broker will be such that clients will drain, which in this case is a prediction of the broker's fraudulent actions.
It turns out that insider and/or client profitability classifier is not really needed, especially if you agree with your opinion about total illiteracy of traders.
Being able to classify anything is a big deal. For example, you are a super-literate algotrader. Try to classify the results of your TS. It is quite possible that you will find that your TS is losing in some situations, and it is quite possible to isolate and filter them out.
The HFT MM algorithm classifier will do this for the super-literate trader. And if it succeeds, it will use this filter to its profit: to receive statistically predicted algotrader's loss and decrease it (loss) by more profitable prices for the algotrader. That is, the process will be mutually beneficial for both the algotrader and the HFT MM algorithm.
And all just because someone is better at classification, or someone is too lazy to do it.
Being able to classify anything is a big deal. For example, you are a super-literate algotrader. Try to classify the results of your TS. It's quite possible that you'll find that your TS is losing in some situations, and they may well be identified and filtered out.
The HFT MM algorithm classifier will do this for the super-literate trader. And if it succeeds, it will use this filter to its profit: to receive statistically predicted algotrader's loss and decrease it (loss) by prices more profitable for the algotrader. That is, the process will be mutually beneficial for both the algotrader and the HFT MM algorithm.
And all just because someone is better at classification, or someone is too lazy to do it.
I will insert my 5 kopeck. Your statement is correct if 1) there is a statistical pattern between trading time and trade performance, and 2) you can accurately determine the period of time during which the position will be held. And even this information may not be enough, because we do not know the volume of entry for each trade and MM of the trading algorithm. Also, the attempt to find a pattern between time and trade results will eventually lead to finding a pattern between market conditions (trend, flat) and trade results. It is also clear that everyone is losing in a trend, but trades can be made when the market was still in a flat.
In my opinion, the netting of client positions by the HFT MM algorithm should be done according to two features of the TS: the level of risk (leverage) and the profit/loan ratio.
The entry volume of each trade is unimportant, as is the MM of the trading algorithms. Equity_All BP is assessed - all PriceTakers of the ECN/STP aggregator: whose orders go to the outside (other than LP0 sources).
The example of classification by time is chosen on purpose in order to make it as simple as possible on the fingers to tell the point. Classification can be by many parameters.
After all, this is a reference book, there is no need to use algotrading tricks.