Discussion of high-frequency trading on MT5 - page 5
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
To launch such a start-up alone, you need at least $15-20m.
First and foremost, you need brains, ideas and strategies. And the price of the issue and the entry threshold are the next (although important) question. HFT on the CME is one thing, but it is quite another, for example, on the FORTS.
And in general, I do not quite understand how you can organize HFT in the "classic stock" version of Forex? If I'm not a big bank - liquidity provider or some ICAP, then I can use only the "consumer" software and standard services of Forex brokers. But even if I take ECN account (others are rejected right away), it is still no single platform, no classic market depth, no possibility to expose my orders in market depth fast enough (especially limit orders with guaranteed execution) because there is no analogues of high-speed connection, spread (even on ECN) is big enough, etc.
If someone has experience - advise how to overcome these "infrastructural" obstacles of forex? And is there any advantage of HFT over HFT on exchanges here?
If anyone has experience - tell me how to overcome these "infrastructural" obstacles of forex? And are there any advantages of HFT here over HFT on the exchanges?
The advantage of FX-HFT over stock exchanges is that it is poorly developed. That is, there is almost zero competition. Moreover, it does not require any wild speed in microseconds, a dozen or two milliseconds is sufficient.
If you delve into the subject of the current state of FOREX in retail, you can find a great option Home-FX-HFT. There is almost always only one thing holding you back - laziness.
EvMir: вам совсем поплохеет если узнаете, про то что некоторые инвестируют в исследования квантовой запутанности, потому как им мешает скорость света и надеятся за счет квантовой спутанности решить проблему мгновенной передачи данных...
If these some people suddenly found out that quantum entanglement does not in principle allow information to be transmitted faster than light, they would be sickened by the sudden realisation of the mediocrity of the money spent.
First and foremost, you need brains, ideas and strategies. Price and entry threshold are the next (albeit important) question. HFT on the CME is one thing, but on FORTS, for example, is another.
And in general, I do not quite understand how you can organise HFT in the "classic stock" version of forex? If I am not a big bank - liquidity provider or some ICAP, then I can use only the "consumer" software and standard services of Forex brokers. But even if I take ECN account (others are rejected right away), it is still no single platform, no classic market depth, no possibility to expose my orders in market depth fast enough (especially limit orders with guaranteed execution) because there is no analogues of high-speed connection, spread (even on ECN) is big enough, etc.
If anybody has experience - advise how to overcome these "infrastructural" obstacles of forex? And is there any advantage of HFT over HFT on exchanges here?
"For example, if we take a person with brains and knowledge of Einstein and give him some basic literature about trading and strategy logic, then 0-vaya probability he will invent a profitable HFT system for Goldmansax, only in his head, without any tests on super expensive equipment. And to test such TS we need equipment which is thousands times more powerful than home PCs, otherwise it's all just intellectual masturbation and demagogy.
There has to be a stage in the progression to minimise information dead ends. If HFT is the only beacon on the path of algorithmic trading, then we need to understand what the stage of progression to this goal is. For example, is it important to test conventional trading strategies on mass trading platforms, intraday, scalper strategies, and are algorithmic patterns of such strategies applicable to HFT? In other words, won't it just be a waste of time? I admit that there is no zero-track, and one has to invest a lot of money in education and equipment, and people have to "partner" on scams, price-action, solve candlestick patterns, listen to Mr. Safin.
I therefore wrote about 15-20, this information from a man who is closely connected with the HFT business at the stock exchange.
Of course I will be glad to hear that you can do it from 0 on Forex, using MT5. Because I like this software, but I need to understand if I am not addicted to another toy...
The advantage of FX-HFT over stock exchanges is weakness. I.e. the competition is almost zero. Moreover, it does not require any wild speed in microseconds, a dozen or two milliseconds is enough.
If you delve into the subject of the current state of FOREX in retail, you can find a great option Home-FX-HFT. There is almost always only one thing holding it back - laziness.
That sounds strange, it can't be. Where's the good old-fashioned struggle for existence? I think that if there seems to be no competition, it is more likely that you have lost.
If these some people suddenly found out that quantum entanglement does not in principle allow information to be transmitted faster than light, they would be sickened by the sudden realisation of the mediocrity of the money spent.
That's why I wrote about 15-20m, it's information from a man who is closely connected with HFT business at the stock exchange.
Naturally, I will be glad to hear that it is possible to do it from scratch on forex, using MT5. Because I like this software, but I need to understand if I'm not hooked on another toy...
...
Sounds strange, it can't be. Where's the good old-fashioned struggle for existence? I think that if there seems to be no competition, you have probably lost out in it.
Building the infrastructure for FX-HFT also fits into the order of magnitude of the amounts quoted above. However, there is a side other than traders that may be interested in developing this class of strategy. For this reason, the bicycle (HFT-infrastructure) can be once "invented" and made available at no cost to everyone - Home-FX-HFT.
Any platform is an additional pad between the trader and the market. So using MT5 in particular is simply wasteful with respect to HFT:
Regardless of whether you are an algotrader or not, the speed of the platform will have a direct impact on your bottom line. The higher the speed, the greater the profit. This is caused by the fact that the best moments (prices + liquidity) to open are of short duration (10-300 ms). And any additional hundred ms even in manual trading, though imperceptibly, but it decreases the profit.
P.S. That is why it is always more profitable to trade where there is a sharpening for HFT.
Building the infrastructure for FX-HFT also falls within the order of magnitude of the amounts quoted above. However, there is a side other than traders that may be interested in developing this class of strategy. For this reason, the bicycle (HFT-infrastructure) can be once "invented" and made available free of charge to all comers - Home-FX-HFT.
Any platform is an additional pad between the trader and the market. Therefore, using MT5 in particular is simply wasteful with respect to HFT:
P.S. Therefore, it is always more profitable to trade where there is a HFT sharpening.
Where is there one from your point of view? Except inhouse of course...
"For example if we take a person with brains and knowledge of Einstein and give him some basic literature about trading and strategy logic, then 0 probability he will invent a profitable HFT system for Goldmansax in his head only, without any tests on super expensive equipment.
Partly agree. If you want to work seriously - you need to understand the "technique", you need both equipment and tests.
Often people start from the technical capabilities they already have. For example, a certain "bigwig", who has his own server in the same room as the exchange's server and therefore has a huge advantage in speed, is aware of his technical superiority in advance and already adjusts his "strategies" to it. Or this approach may be inherent in people, programmers by profession, who are aware of the current level of technical infrastructure of the market and look at trading solely as an algo-trading (not necessarily HFT).
And there are traders without a detailed knowledge of programming, who first generate ideas and strategies and then easily develop them in practice, using the help of professional programmers.
And by the way, what will a team of good programmers with very expensive equipment do without "a person with the intelligence and knowledge of Einstein", without a profitable strategy? - A pile of metal and money down the drain.
You need a fusion of the brains of a creative trader, a programmer and the hardware.
As for the cost of equipment, it still depends on the tasks that need to be solved within trading and the characteristics of the trading floor. I cannot speak for Forex (the market is new to me, there are white spots regarding detailed information about its participants and trading organization schemes), but the exact cost of HFT in CME (where you have to fight with Goldman and Co) and of HFT in FORTS (where you will soon have nobody to fight with and nothing to do) will be very different (by the way, for FORTS there is already existing software for HFT trading, effective enough for "home" HFT).
Everything you say is right, all the components are important. The thing is, I've heard over and over again that HFT is basically the only way to get rich in algorithmic trading, all other ways are super risky and more like a casino. So the question arises, how to get such skills without such a technique and such software? Is there continuity in development, from small algotrader working with kitchens, to HFT monster that Goldmansax and Gipimorgan can't share?
Or is it a completely different kind of activity, like different typing and OS kernel development like linux... If so, then you should not waste your time and play with kitchens and mass Forex platforms, where you can earn maximum 1-2K$ a month, with big risk to lose your honestly earned 10-20K and to do something else.I`m not going to earn on Forex market as a usual programmer with 3-5K month salary ceiling. Since there is such a risk, there should be no ceiling, and the kitchen will not let me do it, so working with DC is a transitional stage, and if it is senseless, I have to think in another direction.
You need to understand the goals and the risks.