An effective trading strategy based on multi-currency analysis of multiple DCs

 

Participating in the discussion of the topic 'Multi Currency Pair Analysis, your opinion, can it be used?' About multicurrency analysis I saw the emergence of interest in the creation of trading systems based on this principle. And also seeing in different forum threads that most of the participants are programmers, but not all of them found their strategies and are trying to experiment with less effective ones, I decided to propose an experimentally tested idea of an effective trading strategy to a wide public.

I have been doing research in the direction of multicurrency analysis for building expert systems for a long time. In brief, the idea is that an optimally selected group of different instruments is used as input parameters for the expert system. Analysis and decision making is performed for the whole group. The optimal group is selected according to the criterion of maximum positive and negative correlation in relation to those symbols which are supposed to be traded. Lagging arguments are not used, so as not to decrease the dynamic characteristics of the system. For increasing informativeness of input parameters - covariances of input data are used, besides, there may be different methods of combining, from the simplest multiplication to reduction of some parameters into some non-linear polynomial.

Last

year, based on these researches, I got the idea to use in the multicurrency analysis the quotes of not one, but several brokerage companies

.

It was triggered by the fact that while observing price dynamics of one and the same instruments in charts of different brokerage companies you can see significant differences in dynamics. And it was clear that some brokerage companies would react earlier to some market changes, and the reaction was considerably delayed in other ones.

I decided to check it experimentally, trading manually on a demo account

.

I opened three terminals, one of a Russian brokerage company, another of a European one, and another of an American one. I opened 4 charts in each terminal, including gold, on which I was going to trade. All I worked on is one-minute timeframe. I worked with two other brokerage companies that had the most dynamic quotes I could find. I did not use any additional indicators, expert advisors, made my decisions following the price change dynamics only. The experiment proved my assumptions. Even by eye I could foresee an anticipatory reaction of European and American brokerage companies and even in a manual mode managed to react to reversals and reversals of market trends, close positions in time and predict the trend direction to open an order more efficiently.

After three weeks of trading I reached the level when it was clear enough that this strategy was very effective and further trading would not cause any significant corrections in the result

But I have to say this strategy is very difficult for hand trading, I had to watch every minute of it without distractions and watch closely several charts during several hours trying to catch and remember reciprocal price trends

.

After two or three hours I was exhausted, so I limited myself with one or two deals per day. These results may not seem too much to someone, but it must be said that this is not about testing a trading system, where you need to check its performance on different historical data, but about testing a method, where you can quickly understand if it works or not, and you just were lucky in trading for three weeks without any additional information.

The conclusion is that the strategy is very effective

.

And I am sure that with technical implementation of this method, when the so-called human factor will be eliminated, the trading results will be 5-7 times better than on my statement.

Unfortunately, I do not have enough resources to check this strategy in expert system

at the moment.

I made an Expert Advisor system for multicurrency analysis from one brokerage company and my laptop is already running on steam and it hardly pulls it, but for the implementation of this strategy the resources need to be much higher.

Alpari Ltd

00 0 00 00 00 00 00.00 00 00 00 00 00 00 00 00 0. 00.00.00 00 0 00 0 00 00 00 00 00
Account: 272168Name: KhristiCurrency: USD2006 September 12, 16:21
Closed Transactions:
TicketOpen TimeTypeLotsItemPriceS/LT/PClose TimePriceCommissionesSwapProfit
70578992006.08.22 11:16BalanceDeposit100,000.00
70949412006.08.23 16:25sell25.00gold625.700.000.2006.08.23 16:41624.700.00
.000.6 250.00
71195412006.08.24 16:04sell40.00gold622.700.000.002006.08.24 18:24622.200.000.
0.00 0.
5,000.00
71388002006.08.25 14:02sell40.00gold622.000.000.2006.08.25 16:39621.100.000.000.
9,000.00
71583582006.08.28 12:45sell40.00gold623.000.000
2006.08.28 15:44621.200.000.0.00 0.
18,000.00
71827042006.08.29 14:36sell40.00gold614.700.000.2006.08.29 15:42614.200.000.000.5,000.00
71850262006.08.29 16:04sell40.00gold613.500.000.2006.08.29 16:43608.200.000.
0.53,000.00
72110482006.08.30 14:33buy40.00gold616.500.000.2006.08.30 15:22618.200.00
000.
17,000.00
72325742006.08.31 14:20buy40.00gold624.700.000
2006.08.31 14:45626.200
000.000.0015,000.00
72575672006.09.01 14:21sell0.10gold624.200.000.2006.09.01 16:01622.200.00.000.0050.00
72576092006.09.01 14:22sell40.00gold623.700.000.002006.09.01 15:12623.200.000.000.5,000.00
72624922006.09.01 15:20sell40.00gold622.200.000.002006.09.01 15:44620.700.00
.000.15,000.00
73041682006.09.05 14:21buy40.00gold633.700.000.
2006.09.05 14:43635.500.000.000.000.18,000.00
73051652006.09.05 15:03buy40.00gold637.600.000.002006.09.05 17:46638.600.000.000.10,000.00
73276802006.09.06 13:16sell40.00gold635.500.000.
2006.09.06 16:18636.200.00
0
00 00 00 00
.
00-7,000.00
73564292006.09.07 13:07sell40.00gold633.200.000.2006.09.07 13:49632.700.000.000.5,000.00
73881222006.09.08 13:14sell40.00gold612.100.000.2006.09.08 15:32610.60
0.
00
0
00,000.00 0 00.00..
.
15
00
73944992006.09.08 16:15sell40.00gold610.100.000.002006.09.08 19:01610.500.00.000.
4 000.00
74169762006.09.11 14:37sell40.00gold592.700.000
2006.09.11 16:24589.000.000
0037 000.00
74418402006
09.12 14:43sell40.00gold593.000.00
0
00 0.00.
.2006.09.12 16:19592.500.00
0
005 000.00
0.00 0
.00
0.00235 300.00
Closed P/L:235 300.00
Open Trades:
TicketOpen TimeTypeLotsItemPriceS / LT / PPriceCommissionTaxesSwapProfit
No transactions
00 0.00
0.000.000.
0.00
:
Floating P/L:0.00
Working Orders:
TicketOpen TimeTypeLotsItemPriceS / LT / PMarket Price
No transactions
Summary:
Deposit/Withdrawal:100 000.00Credit Facility:0.00
Closed Trade P/L:235 300.00Floating P/L:0.00Margin:0.00
Balance:335 300.00Equity:335 300.00Free Margin:335 300.00
Details:

Gross Profit:242 300.00Gross Loss:7 000.00Total Net Profit:235,300.00
Profit Factor:34.61Expected Payoff:12384.21
Absolute Drawdown:0.00Maximal Drawdown (%):7 000.00 (2.5%)
Total Trades:19Short Positions (won %):14 (92. 86%)Long Positions (won %)
5 (100.00%)
Profit Trades (% of total):18 (94.74%)Loss trades (% of total):1 (5.26%)
Largestprofit trade:53,000.00loss trade:-7,000.00
Averageprofit trade:13,461.11loss trade:-7,000.00
Maximumconsecutive wins ($):13 (176,300.00)consecutive losses ($):1 (-7,000.00)
Maximalconsecutive gains (count):176,300.00 (13)consecutive losses (count):-7,000.00 (1)
Averageconsecutive wins:9consecutive losses:1

 
I think this kind of shortcoming of some Russian DTs will not last long. I know one case when guys sitting in the hall of one brokerage company were watching the quotes of a news agency on TV and found that the brokerage company's quotes were almost a minute late. Having got used to it, they made almost 100K GEL in a short time. Of course, by that time they were found out. The profit was paid, because the company was not cheap. But the delay was eliminated.

The only advantage of this approach, other than earning money, is the possibility to detect the elimination of quotes lag by lagging broker even by eye. As for the rest I cannot assert anything yet.
 
If a demo account lags, it does not mean that the real account will also lag. In addition, they may not open or may open with a large slippage.
 
DrawDown:
With a bit of skill, they made almost 100K green in a short period of time. Of course, by this time they had been tracked down.
DrawDown, how did they figure it out? Obviously, it was a pips, because the quotes will not run too far in a minute. Maybe that was the formal reason for the DC to pay attention to them?
 
Mathemat:
DrawDown:
With a knack for it, they made almost 100K green in a short period of time. Of course, by this point, they'd been identified.
DrawDown, how did they figure it out? It is clear that it was a pips, because quotes will not run away from the market for a minute. Maybe that was the formal reason for the DC to pay attention to them?

Maybe it was. I was not acquainted with these guys. The director of one of the branches of brokerage companies where the story took place told me about this case. As the saying goes: "The music did not play long... the freeman did not dance long..." The dealing room managers paid attention to these guys because of the abrupt change of trade results. And this change was out of the usual scope. If they had not been greedy, they would have hardly noticed lagging of quotes :))

I think the results shown in Pilgrim's statement are also out of the ordinary, for one can work with such volumes, if not with a heavy depo, then with a considerable confidence in the result.
 
Mathemat:
DrawDown:
With a knack for it, they made almost 100K green in a short period of time. Of course, by this point, they'd been identified.
DrawDown, how did they figure it out? It is clear that it was a pips, because quotes will not run away from the market for a minute. Maybe that was the formal reason for the DC to pay attention to them?

Most likely, there weren't any losers at all.
 
stateman some old, can't you trade like that nowadays?
 

In principle, technically there is nothing illegal about this strategy, it's just a shortcoming of the DC itself. And instead of sitting in the hall, you can do the same at home. Another thing is that it would be pipsing... So, alas, it does not interest me.

 
I don't think you quite understand what we are talking about and what I am suggesting in this strategy.
DrawDown is not a shortcoming of a particular broker, and it's not just about Russian brokerage companies. And this is not only about Russian ones. Different brokerage companies get quotes from different sources, and some ones get them through intermediaries. And it is not by chance that I have chosen one European and one American brokerage companies as additional ones.
Belford - the same situation on the real account.
DrawDown "I think the results shown in Pilgrim's statement are also beyond the usual limits, because you can work with such volumes if not with large depot" - the depot is irrelevant, you could have 1k and trade with 0.4 lots, the question is not about the total profit, but about stability and speed of multiplication.
"I'm not sure about the total profit, but the stability and the speed at which it increases. " - in this you are absolutely right, after three weeks of trading such confidence was absolute.
brOOt - I gave a statement of the time when I got this idea and I decided to check it, to double-check it again neither now nor later, I do not think it is necessary. Besides, as I wrote, for manual trading this strategy is very heavy.
Mathemat- we are not dealing with pipsing here at all. When an EA system is trained based on multicurrency analysis, its algorithm is based on comparing the dynamics of changes in various symbols, and if these changes are stable and one instrument is always faster than the other in its movement, even by fractions of a second, it will be enough for the expert system to detect and remember this fact. Due to this, the accuracy of both short-term and medium-term forecasts increases. And when using quotes from several brokerage companies the differences in dynamics become even more significant.
 
Piligrimm:
I don't think you quite understand what we are talking about and what I am suggesting in this strategy.

1 - the picture is the same in real life.

2 - "So with great confidence in the result. "You are absolutely right, after three weeks of trading it was an absolute certainty.

3 - We are not talking about pipsing here at all. When an Expert Advisor is trained on the basis of multicurrency analysis, its algorithm is based on comparing the dynamics of changes in various symbols, and if these changes are stable and one instrument is constantly ahead of the other in its movement, even a fraction of a second, it will be enough for the expert system to understand and remember this fact. Due to this, the accuracy of both short-term and medium-term forecasts increases. And when using quotes from multiple brokerage companies the differences in dynamics become even more significant, and I tried to draw your attention to it.

1 - Have you tested the method on a real account?

2 - Why was the absolute certainty?

3 - If we are not talking about pipsing, then tell me how the accuracy of predictions, both short-term and medium-term, may be affected by changes in the dynamics of various instruments from several brokerage companies, even for a fraction of a second?

Don't take it as a criticism, I'm just trying to understand what this strategy is about and what you are suggesting :)
 
DrawDown:
Piligrimm:
1 - it's the same picture in the real world.

2 - "so with great confidence in the result. "You are absolutely right, after three weeks of trading it was an absolute certainty.

3 - We are not talking about pipsing here at all. When an Expert Advisor is trained on the basis of multicurrency analysis, its algorithm is based on comparing the dynamics of changes in various symbols, and if these changes are stable and one instrument is constantly ahead of the other in its movement, even a fraction of a second, it will be enough for the expert system to understand and remember this fact. Due to this, the accuracy of both short-term and medium-term forecasts increases. And when using quotes from several brokerage companies the differences in dynamics become even more significant.

1 - Have you tested the method on a real account?

2 - Why was the absolute certainty?

3 - If we are not talking about pipsing, then tell me how the accuracy of predictions, both short-term and medium-term, may be affected by changes in the dynamics of various instruments of different brokerage companies, even for a fraction of a second?

Don't think of it as a criticism, I'm just trying to understand what you're talking about and what you propose in this strategy :)

I understand the author has created an indices of different pairs
Then he took some DTs
and uses lagged quotes

this strategy needs a good quality internet connection to get quotes for several instruments
from different dealers


Piligrimm right ?
Reason: