Machine learning in trading: theory, models, practice and algo-trading - page 300
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
Technical Insider and MO should not be confused.
What is this "technical insider"? The proverbial "flash orders"? Something consperological?
You are not only talking to the clients of the Forex kitchens, there are those who have repeatedly read the HFT A Practical Guide to Algorithmic Strategies and Trading Systems and even understood everything)))) Of course there is a lot of "technical" in HFT, but without advanced analytics, which is in MO (old-style econometric models), if you don't know where/how the price will move at the next tick/second/minute/hour, you will still sell out and in case of HFT very quickly. HFT is not only "ultra", stupid nanosecond MM, which is not very interesting and less and less profitable, HFT is also an intraday, as long as you don't leave it for the night, in fact the border between HFT and algotrading is quite wide.
Do you think that Renaissance with Tesa mainly on ultra-HFT earn their profits? I assure you no! And evenultra-HFT also use MO, since we need to find the optimal algorithms for calculating a " fair" price for a given instrument, depending on thousands of others, these algorithms are stupid and inefficient to find by hand, of course the algorithms are simpler there, due to known technical limitations, but still.
I can't do without MO nowadays. But MO is not simple, more complicated than wizards and martin.
What is this "technical insider"? The proverbial "flash orders"? Something consperological?
You are not only talking to the clients of the Forex kitchens, there are those who have repeatedly read the HFT A Practical Guide to Algorithmic Strategies and Trading Systems and even understood everything)))) Of course there is a lot of "technical" in HFT, but without advanced analytics, which is the MO (econometric models in the old fashioned way), if you do not know which way/how the price moves at the next tick/second/minute/hour, you will still sell out very quickly in case of HFT. HFT is not only "ultra", stupid nano-second MM, which is not very interesting and less and less profitable, HFT is also intraday, if only not to leave deals for the night, in fact the border between HFT and algotrading is quite wide.
Do you think that Renaissance with Tesa mainly on ultra-HFT earn their profits? I assure you no! And evenultra-HFT uses MO too, since we need to find the optimal algorithms for calculating a " fair" price for a given instrument, depending on thousands of others, these algorithms are stupid and inefficient to find by hand, of course the algorithms are simpler there, due to known technical limitations, but still.
I can't do without MO nowadays. But MO is not simple, it is more complicated than mash-ups and martin.
Well, these are probabilistic models, not MO. Talk to the same high-frequency LPIs, there is no MO there. There are probabilistic models in the form of their own bikes.
Now for some reason it has become fashionable to call everything MO. However, HFT worked even before the term appeared.
HFT is also an intraday, if only not to leave trades overnight, in fact, the boundary between HFT and algotrading is quite wide.
So one may agree that scalping is an hft. As for me hft is any trade, the main risks of which are quality and speed of execution. everything else is pipsing.
This way we can agree that pipsqueak is hft. The hft is any trade whose main risks are quality and speed of execution. everything else is pipsqueak.
Yes, "scalping" is hft, but "scalping" is a kitchen-forex term, it is associated with "no one likes scalping", beaten out client's stops, regression, affiliate and other absurdity, so it is better not to use it, scalping is better, algo-sculpting is hft))
In general, I will not argue about the definition, as far as I know some regulatory commission in the states for 2 years thought and did not give birth to a definition of HFT, the smart books refer to HFT all that does not leave positions overnight.
Seems to me if an algorithm digests an orderlog or a tape/stack to make trading decisions, it's HFT.
Well, these are probabilistic models, not MO. Talk to the same LFI high-frequency people, there is no MO there. There are probabilistic models in the form of their own bicycles.
Now for some reason it has become fashionable to call everything MO. However, HFT worked even before the term appeared.
Again the argument about definitions...
Let "probabilistic models", ARMA, GARCH... When I saw a video about forecasting the guy at the end of his video told me that all this old econometric stuff is rarely used by anybody except students because they are inferior to MO.
https://youtu.be/U5kIdtMJGc8
The guru said: "In fact we were doing machine learning all the time" And you can't argue with a guru))
The guru said: "In fact, we were doing machine learning all the time" And you can not argue with the guru))
Well, of course, if you used a calculator, then MO immediately!
What is this "technical insider"? The proverbial "flash orders"? Something consperological?
You are not only talking to the clients of the Forex kitchens, there are those who have repeatedly read the HFT A Practical Guide to Algorithmic Strategies and Trading Systems and even understood everything)))) Of course there is a lot of "technical" in HFT, but without advanced analytics, which is in MO (old-style econometric models), if you don't know where/how the price will move at the next tick/second/minute/hour, you will still sell out and in case of HFT very quickly. HFT is not only "ultra", stupid nanosecond MM, which is not very interesting and less and less profitable, HFT is also an intraday, as long as you don't leave it for the night, in fact the border between HFT and algotrading is quite wide.
Do you think that Renaissance with Tesa mainly on ultra-HFT earn their profits? I assure you no! And evenultra-HFT also use MO, since we need to find the optimal algorithms for calculating a " fair" price for a given instrument, depending on thousands of others, these algorithms are stupid and inefficient to find by hand, of course the algorithms are simpler there, due to known technical limitations, but still.
I can't do without MO nowadays. But MO is not simple, more complicated than wizards and martin.
where's the thing you need to read in there? empty folders) Can I take a personal note?
Do the same for the random BP.
My teacher is ZZ, i.e., the patterns that bring profit are selected, not patterns in general.
What is the teacher of the random one? What meaning will the selected patterns have?
And the random one has what teacher? What meaning will the selected patterns have?
My teacher is ZZ, i.e., the patterns selected are those that bring profit, not patterns in general.
And the random one has what teacher? What meaning will the selected patterns have?