The philosophy of algotrading - page 13

New comment
 
Logically speaking, there is no parasitism in exchange transactions, because from the very beginning the forex market was created by the uncles to massively involve clients in the general market to maintain liquidity in order to contain sharp and significant fluctuations in exchange rates. Now all forex market participants maintain liquidity, doing, so to speak, a useful thing). And the uncle, who created this market, can easily take as much as necessary without a significant drawdown of prices for the instrument. And we, mostly small fiddlers, do not do any harm. Why then do you repent so much? I do not understand.
 

Today I want to raise a question about the future development of algotrading.

  1. Where is algotrading headed?
  2. What is its future?
  3. What will trading robots be like in 10 years?
If you have any thoughts on this question, please speak up. )
 
Vitalie Postolache:

Speculation used to be punished quite severely. Sometimes by firing squad. Why do you think so?

Forex is speculation. I mean yes, parasitism)))

And the rest of the near-exchange suffering that is not related to trading real resources is also speculation in principle. You don't get your hands on stocks or precious metals or the same petroleum products, you only get numbers. It is like an unsecured dollar, just numbers. Now you have it, but tomorrow you won't even be able to wipe it off, it's not big enough.

as if it's something bad........

e.g. consultancy salaries are also pure parasitism ......

many examples can be found.......

 
Реter Konow:

Today I want to raise a question about the future development of algotrading.

  1. Where is algotrading headed?
  2. What is its future?
  3. What will trading robots be like in 10 years' time?
If you have any thoughts on this question, please speak up. )
Until algorithms learn to analyse the real market situation, nothing will change. Algorithm developers must understand that the real price that has emerged at the moment - is the result of a compromise between the participants on the one hand, and the market situation on the other. If market participants focus on the real price, then, the market situation forms a virtual market price, which actively influences the formation of the real price. But nobody sees this and nobody seems to want to see it. The very fact that no one can conquer the market decisively shows that something essential is not taken into account in market analysis and it has a systemic nature. In my opinion, this essential factor is the consideration of the market situation in the form of a virtual market price. Trend algorithms die at fdet, and fdet algorithms die at trend, though they are one and the same market price flow. Only consideration of the market situation will create a universal algorithm.
 
Yousufkhodja Sultonov:
Trend-following algorithms die when there's a fad, and fad-following algorithms die when there's a trend.

I think there are enough algorithms now to recognise what's happening in the market

 
Yousufkhodja Sultonov:
Until the algorithms learn to analyse the actual market situation, nothing will change. Algorithm developers must understand that the real price that has emerged at the moment - is the result of a compromise between the participants on the one hand, and the market situation on the other. If market participants focus on the real price, then, the market situation forms a virtual market price, which actively influences the formation of the real price. But nobody sees this and nobody seems to want to see it. The very fact that no one can conquer the market decisively shows that something essential is not taken into account in market analysis and it has a systemic nature. In my opinion, this essential factor is the consideration of the market situation in the form of a virtual market price. Trend algorithms die at fdet, and fdet algorithms die at trend, though they are one and the same market price flow. Only consideration of the market situation will allow the creation of a universal algorithm.

Interesting point of view...

As far as I understand your point of view from various posts on various threads, you tend to believe that the future lies with "intelligent" advisors who will easily find different formations?

So to speak - a mastery of technical analysis.

Developers should strive to create smarter and smarter robots?

Or do you have something else in mind?

 
trader781:

I think there are enough algorithms now to recognise what is happening in the market

I don't think that's entirely true... Where are those algorithms?
 
Реter Konow:
I don't think that's entirely true... Where are those algorithms?

there is an advisor like wall street

it doesn't lose for years and doesn't need to be optimised

so there's an algorithm in there.

if you can get it out of there you'll understand how it works.

 
trader781:

there is an advisor like wall street

it doesn't lose for years and doesn't need optimization

so there's an algorithm in there.

If you can get it out of there, you'll understand how it works.

Interesting... I've never heard of such an Expert Advisor.

What algorithms do you think it uses?

 
Реter Konow:

Interesting point of view...

As far as I understand your point of view from various posts on various threads, you tend to believe that the future lies with "intelligent" advisors who will easily find different formations?

So to speak - a mastery of technical analysis.

Should developers strive to create ever smarter and smarter robots?

Or do you mean something different?

Quite right. Now the technical analysis covers half of the information, because the other half in the form of the market price is hidden from the observer. Observations have shown that the change of the trend direction is due to the influence of the virtual market price. As for creation of "smart" robots, we can say that they have their brains in the algorithm beyond which they will not be able to get clever. Unless they try to introduce an additional logic of behaviour depending on the market situation through a series of conditional transitions. In the meantime, it has proven sufficient to take into account the relationship or interaction of MAP and IDA, respectively the average price and the market price. For example, on D1 from 2009 to present. where there are both trending and flat areas, which the algorithm is equally successful or unsuccessful at:

Bars in history 3197

Modelled ticks 5394

Simulation quality n/a

Chart mismatch errors 0

Initial deposit 10000.00

Spread Current (2)

Net profit 34936.93

Total profit 52840.75

Total Loss -17903.82

Profitability 2.95

Expected payoff 16.96

Absolute drawdown 90.33

Maximum drawdown 6614.29 (23.89%)

Relative drawdown 23.89% (6614.29)

Total trades 2060

Short positions (% of winners) 1163 (63.11%)

Long positions (% win) 897 (60.98%)

Profitable trades (% of all) 1281 (62.18%)

Profitable trades (% of all) 779 (37.82%)

Largest

profitable transaction 51.58

Deal with a loss -53.00

Average

profitable deal 41.25

Loss trade -22.98

Maximum number

continuous wins (profit) 239 (11809.68)

continuous losses (loss) 102 (-1937.06)

Max.

continuous profit (number of wins) 11809.68 (239)

Continuous loss (number of losses) -1937.06 (102)

Average

continuous winnings 19

continuous loss 12

1...678910111213141516
New comment