Menu
Sister Sites

Options

Bookmark Thread

First Page First Unread Last Page Last Post

Print Thread

Similar Threads

Triangular Arbitrage 16 replies

Triangular Arbitrage 251 replies

Triangular Arbitrage and carry trade 6 replies

Triangular Arbitrage 7 replies

Question: Trade Mechanics of Triangular Arbitrage 6 replies

Attachments: Simple High Profit low Drawdown Triangular Arbitrage
Exit Attachments

Simple High Profit low Drawdown Triangular Arbitrage

@jesing
I am sorry I took so long to reply.

Was a rough couple of weeks.

I replied to your pm now.



@EVERYONE
I am still here, but the project is still on hold until I can scrape together funds for the development and new pc hardware for the development.
I will take this up again one day, but it might take many months as I am busy doing as much coding work as possible to build funds so that I can take off and work on this again.

I was even playing with the idea of adapting this strategy to the crypto market and have already learned some of the api of some of the crypto brokers out there so my application can be adapted to it later.
This strategy benefits a lot from volatility, instability and price lag.
Forex does work for this, but in all 3 cases crypto currencies are king for this.
I even played with multiple broker api single crypto currency latency arbitrage, which seem to be huge as unlike forex most broker contain their own trading platforms with minimal price synchronization, basically relying from outside trading the regulate the price at a great profit to those just playing the inefficiencies, but you need a lot of fund on the brokers to trade this profitable.

Still.
I will get back to the project one day.
here is this abitrage trading in mt4 terminal:
https://www.mql5.com/ru/signals/394408
@SartSart

Nice.
Are you using the strategy in this thread or something else ?

What broker are you using, what is your ping to the broker and what slippage are you getting ?
What pairs are you using this on ?
Quoting SartSart
here is this abitrage trading in mt4 terminal: https://www.mql5.com/ru/signals/394408
thats useless with this demo account in best case, i would say its useless 100 times.
arbitrage trading in demo has always extrem wins, anyone can do it, but why you should do it when it shows nothing! and in real account then no wins or cant pay out.

at least use a cent account, even there the results are not same, but not this extrem like in a demoaccount where all is wrong whats important for this stategy: volume in market, slippage, server connection, data feed,...
1. i use some modification of triangle arbitrage...
2. some guys say anybody can get such profit as mine on demo, well please try and i'll see...

please look at this signal once more at now time status:
https://www.mql5.com/ru/signals/394408
Quoting jvbJacques
@SartSart Nice. Are you using the strategy in this thread or something else ? What broker are you using, what is your ping to the broker and what slippage are you getting ? What pairs are you using this on ?
Hello JVB,

I have been reading the entire forum for a while now and I cant seem to find the most up to date version and also have some confusion over the position sizing.

On the other hand, if you are still looking for the financial backing for your EA, please PM me. thanks
Quoting SartSart
1. i use some modification of triangle arbitrage... 2. some guys say anybody can get such profit as mine on demo, well please try and i'll see... please look at this signal once more at now time status: https://www.mql5.com/ru/signals/394408
ok, somebody get much more win then you have on demo, but why you are so proud to fake wins much more big then others or smaller then other, near nobody trades your bullshit garbage any more, because its usless waste of time and only beginners and scammers do so? never ever compete your fake wins (and nothing more you have, only fake wins in demo account with arbitrage trading, what would be not this wins and often even losses in real account trading) with other fake wins, thats stupid. and dont ask some people to show same fake wins like you, thats even double stupid. noone has this time to invest to show this garbage fake wins for nothing, at maximum when you want sell this garbage bullshit to idiots later when you work as a scammer.

it really looks like you know its big fake wins you have never ever in real account (even a big lose same time) or you dont know anything about trading? whats the reason to fake this wins and present a demo where arbitrage experts have always extrem wins?
you want other also lose time with this worthless trading style and later want sell other idiots and beginners this garbage expert? or whats your reason behind testing useless garbage arbitrage expert in demaccount?? i see this coming with yourt style of writing that you want sell really this as a big scam.
Hi,

I am looking to do this 2 symbols only... I am looking for a EA/script that places buy and sell market orders on two FX pairs simultaneously - for example buy EUR/USD and sell GBP/USD.

I do not want any SL or TP, just to be able to specify the particular symbol to buy and the symbol to sell, and their relevant volume. I am quite new to that, please help me.
@EVERYONE

140% profit in less than 50 minutes.
(Unrealistic I know based on a lot of factors, but still a nice prove of concept.)

I could not stop myself so I created a little program to connect to Binance, just to see if it was even possible to make profit using triangular arbitrage on the real crypto market seeing that I understand the theory and have done successful tests on forex and manual tests on crypto, but never on the live crypto market.

Understand that this was only a test and therefore I ignored the trading volume.
You can only trade whatever volume is available and not more, therefore the profit is based on what is there and not your total account size.

During my older tests I found that the crypto market is much more volatile and slow, compared to forex.
Therefore it only made sense to me that a triangular arbitrage strategy that loves volatile markets and the slower the better, that it will work much better on the crypto market.

Looking at the logs you can calculate the arbitrage manually to verify the results.
There are 3 lines indicating which cryptos were used in the arbitrage. Symbol1, symbol2, symbol3.
Each one also shows the ASK end BID prices at that time, which you must use in your arbitrage triangle calculation.
Remember to subtract 0.05% after each of the 3 calculations for trading cost.
The last *** line contains the total running profit if 1 bitcoin was traded every time, by only using the biggest profit during that round and not all.
Also keep in mind the program only ran once every 5 seconds, so it missed a LOT of arbitrage signals in-between as it can run once every 0.001 seconds, depending on the exchange rules and connection speed, so it can potentially make a lot more profit.

Basically the program checks the ASK and BID prices of all available currencies in the exchange for arbitrage opportunities once every 5 seconds.
If multiple arbitrage opportunities are found then it only use the highest profit potential one for its profit calculation and ignores the rest.

I ran it a couple of times, but here is the last part of my last run :
Obviously you will trade with less volume so the profit will decrease a lot, but still it proves that there is a lot of profit to be made.

Inserted Code
2018-03-14 10h30s55m289 Starting
Parameters *******
 Profit is dependent on :
     1) Server connection speed to exchange.
     2) Volume available for trading as you can not trade bigger volume than what is available.
     3) How often profit potential is checked. This test only does it once every 5 seconds
        , but can be increased to once every 0.001 seconds for much greater profit, depending on exchange rules and connection limitations.
     4) Persentage profit displayed is based on no slippage (1) and an unlimited volume (2).
     5) Exchanges with bigger trading volumes can make bigger profit so long as their trading costs are not too high.
     6) This test is run on Binance and performance will be different for other exchanges. (Better or worse), but can be tested before trading for real.
 Trading Fee=0.0500%
 Check for arbitrage potential every 5 seconds
.
.
. **** logs cut out to fit. See attached log file for full log.****
.

2018-03-14 11h18s48m434 IS PROFIT POSSIBLE cc= 473 Profit = 1
2018-03-14 11h18s50m239 Get all prices available.
2018-03-14 11h18s50m740
symbol1=BNBBTC ASK=0.00113300 Volume=315.88000000 BID=0.00113170 Volume=4.41000000
symbol2=BCPTBNB ASK=0.06280000 Volume=0.53000000 BID=0.06249000 Volume=150.00000000
symbol3=BCPTBTC ASK=0.00007135 Volume=984.00000000 BID=0.00007133 Volume=2123.00000000
2018-03-14 11h18s50m740 symbol1=882.17122683135105 symbol2=14040.28887289701920 symbol3=1.00099305840109250734623200 result=1
2018-03-14 11h18s50m740 symbol3=14008.40925017517275 symbol2=874.94780129642482187492625000 symbol1=0.98968333751380038893039611010643750000 result=-1
2018-03-14 11h18s50m740
symbol1=ETHBTC ASK=0.07599500 Volume=0.07000000 BID=0.07591300 Volume=0.75900000
symbol2=LUNETH ASK=0.02197200 Volume=16.76000000 BID=0.02190000 Volume=0.01000000
symbol3=LUNBTC ASK=0.00165640 Volume=44.31000000 BID=0.00165280 Volume=2.00000000
2018-03-14 11h18s50m740 symbol1=13.15218106447765 symbol2=598.28895748879270 symbol3=0.98835756294300783627272000 result=-1
2018-03-14 11h18s50m740 symbol3=603.41704902196190 symbol2=13.20822595689417512719500000 symbol1=1.00217471903717466267253865798250000000 result=1
2018-03-14 11h18s50m742 Profit check done************************************************ totalProfit=1.40569746874650179693713299799745550000 Percentage Total Profit=140.56974687465017969371329979974555000000 %    biggestProfitDuringRound=1.00217471903717466267253865798250000000
 
 

2018-03-14 11h18s55m742 IS PROFIT POSSIBLE cc= 474 Profit = 1
2018-03-14 11h18s56m268 Get all prices available.
2018-03-14 11h18s56m770
symbol1=BNBBTC ASK=0.00113300 Volume=327.52000000 BID=0.00113290 Volume=0.01000000
symbol2=BCPTBNB ASK=0.06280000 Volume=0.53000000 BID=0.06250000 Volume=92.90000000
symbol3=BCPTBTC ASK=0.00007135 Volume=984.00000000 BID=0.00007133 Volume=2120.00000000
2018-03-14 11h18s56m770 symbol1=882.17122683135105 symbol2=14040.28887289701920 symbol3=1.00099305840109250734623200 result=1
2018-03-14 11h18s56m770 symbol3=14008.40925017517275 symbol2=875.08781534688032272656250000 symbol1=0.99089129251347747725811419492187500000 result=-1
2018-03-14 11h18s56m771
symbol1=ETHBTC ASK=0.07600400 Volume=0.61700000 BID=0.07592100 Volume=0.33800000
symbol2=LUNETH ASK=0.02197200 Volume=16.76000000 BID=0.02190000 Volume=0.01000000
symbol3=LUNBTC ASK=0.00165640 Volume=44.31000000 BID=0.00165280 Volume=2.00000000
2018-03-14 11h18s56m771 symbol1=13.15062365137375 symbol2=598.21811121185160 symbol3=0.98824052676384285031776000 result=-1
2018-03-14 11h18s56m771 symbol3=603.41704902196190 symbol2=13.20822595689417512719500000 symbol1=1.00228033201192598849685570920250000000 result=1
2018-03-14 11h18s56m772 Profit check done************************************************ totalProfit=1.40797780075842778543398870719995550000 Percentage Total Profit=140.79778007584277854339887071999555000000 %    biggestProfitDuringRound=1.00228033201192598849685570920250000000
Attached File(s)
File Type: txt CryTri2018-03-14 10h30s54m250.txt   649 KB | 543 downloads
Some more logs.

I left the program running in a continues loop.
So far it ran 6735 times, once every 5 seconds for a demo profit of 1507%.

Unfortunately My connection is to slow for live trading testing, but the crypto works and the way I adapted the strategy to crypto, you can trade with as little as you want on the exchange.
$1 if you want, but I think the exchange limit might be $10 or more, depending on the transaction cost of bitcoin.
Attached File(s)
File Type: zip CryTri2018-03-14 10h30s54m250.txt.zip   1.1 MB | 997 downloads
try to use it on MT5 but i get some error

2018.03.14 21:33:53.490 2017-09-18 09h00 JVB Arbitrage Auto Profit19 (USDJPY,H1) **** ERROR1=lastError=**** ERROR at=Time=2018-3-14 21:33 newError=4302 in_Str=Start1 lastStr=

can you help?

thanks
The biggest drawdown is ahead of you
@amando77

I disabled all EAs as it stands now for anything beyond 2017.
There are many bugs in the code.

Too many for me to fix in the little free time I have.

As it stands now I am saving up to fund further development, but at the moment development is standing still.
I have created an indigogo campaign in the hopes of getting some funding for the development and testing process.

When I get back on my feet I will update this forum with a new post on it again.

Error 4302 basically means you did not load the symbol in the market watch window.
Make sure all 3 symbols/currencies are loaded.
Open their charts just to make sure.
Then the error should go away.
@EVERYONE

Another interesting update.

I modified the program to artificially delay the profit calculations.

Basically.
I check the prices and determine if there is an triangular arbitrage opportunity like normal.
I then wait for 1 to 2 seconds and then ask for the prices again and then I force the program to trade the arbitrage opportunity it detected previously, whatever the outcome.
This simulates my pathetic internet connection by saying to only trade the arbitrage 1 to 2 seconds later.

As you can see at the end of the attached logs there are 2 losing trades.
There are many more losing trades in the log because of this massive delay also.

But as explained in post one, even with losing trades because of slippage, in the long run the profitable trades cancels them out and still makes profit seeing that triangular arbitrage by its nature contains extremely low draw down.

Ultimately the logs were started at 2018-03-14 23h49s14m068 and ran until 2018-03-15 12h21s01m795 for a total profit of 1466% over the 13 hours it ran.

Keep in mind that it trades as if there is no limit to the volume of the currency and therefore in reality the profit for the total account will be much less seeing it can only trade what is available.
Also keep in mind that the program only checks for opportunities once every 5 seconds and it can be sped up to twice a second or even more if the exchange allows that many trades.

Speed wise I suspect the exchange limits trades to 8 a second and for triangular arbitrage you require 3 trades to make the round trip, so 2 program runs equals 6 trades and therefore 2 runs a second is possible.

Some information about reading the logs :
In the log CC = indicates the loop number of the program.
There is 2 CCs per loop.
The program starts on a odd number EXAMPLE CC=21;
The program only does the actual profit calculation on an even number EXAMPLE CC=22.

Therefore in the logs odd CC numbers check for opportunities and only even CC numbers contain the profit calculation after some delay.
So the profit is displayed after an even CC number loop.

For easy viewing you will see **** and .... lines after the even CC loop.
**** mean that there was no slippage since the opportunity was detected during the odd CC loop.
.... mean that there was slippage since the opportunity was detected during the odd CC loop.

At the end of each of those lines you will see the biggestProfitDuringRound variable that shows if profit was made or not.
If biggestProfitDuringRound is bigger than 1 then profit was made.
If biggestProfitDuringRound is smaller that 1 then a loss was made.

Percentage Total Profit=The total running profit/loss so far.

It seems like arbitrage opportunities tend to stick around for a while in crypto markets, unlike fiat where it happens in a split second.
Attached File(s)
File Type: zip CryTri2018-03-14 23h49s13m234.txt.zip   1.5 MB | 705 downloads
  • Post #434
  • Quote
  • Edited 12:29am Mar 28, 2018 12:12am | Edited 12:29am
Hello JVB, Im novate in the trading, and interested on your EA Triangular Arbitrage. I ve download the last version 2017-09-18 for mt4, but I dont know how it works.

Could you help me, please.

My email is

alejandro
.
vielma
@
gmail
.
com


Best regards

From Chile
Quoting jvbJacques
@EVERYONE Another interesting update. I modified the program to artificially delay the profit calculations. Basically. I check the prices and determine if there is an triangular arbitrage opportunity like normal. I then wait for 1 to 2 seconds and then ask for the prices again and then I force the program to trade the arbitrage opportunity it detected previously, whatever the outcome. This simulates my pathetic internet connection by saying to only trade the arbitrage 1 to 2 seconds later. As you can see at the end of the attached logs there are...
Your algo looks very interesting, im currently working on something similar on binance aswell.
Do you use skype or discord?
@avch1202
The EA will not work past 2017 until I can get around to update it again.
My email jvbccc at gmail

@k0ngoz
Nice.
Hope your binance program work good.
I am having a lot of fun on binance and discovered some nice functions and strategies.
I mostly use email and whatsapp.
PM me for email and whatsapp information.
My email jvbccc at gmail
Dear JVB
Can you post last updated virgin also what is recommended setting and time frame too?
Hi JVB,

Concerning your Binance Bot. IN your logfile I see that you did the trades always with 1 BTC, but there is not always enough liquidity in the order book.

Here the example: you have not enough volume in the order book to fill the orders:

symbol1=BNBBTC ASK=0.00107670 Volume=85.23000000 BID=0.00107520 Volume=32.07000000
symbol2=VENBNB ASK=0.43640000 Volume=0.30000000 BID=0.43560000 Volume=470.13000000
symbol3=VENBTC ASK=0.00047216 Volume=10.00000000 BID=0.00047150 Volume=2006.00000000
2018-03-14 23h49s31m036 symbol1=928.29943345401915 symbol2=2126.11201589655080 symbol3=1.00196058458747609034890000 result=1
2018-03-14 23h49s31m036 symbol3=2116.86716367323170 symbol2=921.64628282781169865574000000 symbol1=0.99045860625481490682545432217600000000 result=-1
2018-03-14 23h49s31m075 Profit check done************************************************ Quick ***


Can I ask you which profit percentage you are getting in real?
Thanks for your reply
Best regards,
Polenergy
Quoting jvbJacques
@Everyone I did not include the modifiable parameters into the EAs. The newly attached ones contains the parameters so you can make sure the symbols match your broker's symbols. Keep in mind that the EAs is set to not use a lot of your account balance. If you back-test you will see the maximum draw-down is quite low. A 3rd parameter allows you to set the amount in % .. 0.5 = 50% of the account balance not to use. If you want to make it aggressive then try 0.1 or 0.2 on a leverage of 1:200. Problem is not that you will bottom out your account, but...
It does not make sense to my brain. Why can you not get out of your losses at break even then? There will always be a day when price goes the other way, no matter how strong of a trend. If we are talking 10 -20 pips. (take a trade wait for +- 10 pips, THEN hedge). Then why can't you always get out of your losses at break even? You may have to wait a week, but so what? I'm calling my idea a straddled triangular hedge. My crazy brain invented it. LOL What am i missing?
@EVERYONE
Long time no see.

I made a lot of test EAs.
One triangular arbitrage is even running live on Binance.
Some good profits at about 0.2% profit a day.
Then again it is not very aggressive, but wow there is a lot of slippage.
Mostly because there is a large delay from the time your market order completes, till the time the new funds of that market order reflects in my account and in the mean time the market slips away.

That and there is not nearly the same amount of volume as on forex
Still.
Not bad.


I also found a good way to make some peanuts by applying a grid onto the TUSD/USDT symbol pair for some nice profits, without worrying about slippage much as these are limit orders and set long ahead of time.
Stable coins seem to be a very nice place to make some good profit for their natural arbitrage opportunities on the same exchange even.



@polenergy
You are correct.
That time I was not running it on a real account yet.

Now I am.
Profit is very low, but still good for me.
Started with 19.63298312 BNB and ended with 19.66216005 BNB.
That is 0.148611802 % profit in 7 hours.

You can see it in the logs by searching for "checkArb start=accountBalance=".

In the attached logs you will see a 7 hour snippet before the EA crashed because BINANCE did not update my account balance in time after the last order.
I inserted more checks since then and it does not crash anymore, but wow sometimes it can take ages for the account balance to reflect and I am still looking for a solution to that, because it eats a LOT of profit.

The logs contain multiple strategies running together, but only the triangular one is running on live money.
1) Triangular arbitrage
Search for "CryTri{checkIfProfitPossible(resultZero)} retestCC=2"
"retestCC=1" is failed tests and not traded for multiple reasons, but mostly because of the binance API problems and to somehow bypass those problems.
If binance can sort out there API delays then you can see that the triangular EA have got multiple times more hits it can process, but until then "retestCC=2" will do for now

There is a lot of A.I code in there so maybe one day I will explain, but it is running well.
Can use a lot more TLC to improve the code to run faster and more profitable.


2) Grid EAs
There are multiple grid EAs running.
Search for:
GridSpikeHalfStrategy{} thisName=GridSpikeHalfStrategyName=isIncreaseAccountUsage=false isSwapCurrencies=false gridPriceStepPercentage=0.0004 gridAccountPercentage=0.1 GridSpikeHalfStrategy{} thisName=GridSpikeHalfStrategyName=isIncreaseAccountUsage=true isSwapCurrencies=false gridPriceStepPercentage=0.0004 gridAccountPercentage=0.1 GridSpikeHalfStrategy{} thisName=GridSpikeHalfStrategyName=isIncreaseAccountUsage=false isSwapCurrencies=true gridPriceStepPercentage=0.0004 gridAccountPercentage=0.1 GridSpikeHalfStrategy{} thisName=GridSpikeHalfStrategyName=isIncreaseAccountUsage=true isSwapCurrencies=true gridPriceStepPercentage=0.0004 gridAccountPercentage=0.1 GridPreCalcStrategy{} thisName=GridPreCalcStrategyName=isSwapCurrencies=false gridPriceSizePerc=0.0001 in_GridMaxLevels=20
GridPreCalcStrategy{} thisName=GridPreCalcStrategyName=isSwapCurrencies=false gridPriceSizePerc=0.0002 in_GridMaxLevels=20
GridPreCalcStrategy{} thisName=GridPreCalcStrategyName=isSwapCurrencies=false gridPriceSizePerc=0.0005 in_GridMaxLevels=20
GridPreCalcStrategy{} thisName=GridPreCalcStrategyName=isSwapCurrencies=false gridPriceSizePerc=0.001 in_GridMaxLevels=20
GridPreCalcStrategy{} thisName=GridPreCalcStrategyName=isSwapCurrencies=false gridPriceSizePerc=0.002 in_GridMaxLevels=20
GridPreCalcStrategy{} thisName=GridPreCalcStrategyName=isSwapCurrencies=false gridPriceSizePerc=0.005 in_GridMaxLevels=20

3)
The best grid strategy I just started testing now so it still need to prove itself, but it looks good in theory and runs on the above mentioned TUSD/USDT grid arbitrage strategy.
Looks like it will even run better than the triangular arbitrage.
Not because the triangular does not work as it has proven itself, but binance seem to have problems with their API interface and that interferes less with the grid EA than it does to the triangular EA.



@Dagoods
It also does not make sense to my brain.
Maybe if you can explain your question in another way I can answer you.
Maybe I am just to tired to understand at the moment and for that I am sorry. Not you fault. My brain is week at the moment need zzzzzz

If you meant to ask why I do not keep the triangle open all the time and way for profit.
Short answer is . "SWAP"
Swap is a killer so I want to be in and out as fast as possible.

As for the crypto version of the triangle.
My crypto EA works differently in that it transfers the money to actual live wallets, but as I am typing I realise that I probably can buy and hold.
Problem is that then you have to home all the potential arbitrage wallets as I play in all of them using the current method.
Not only is that very expensive, bit will also decrease the profit as your account is not split between a hundred wallets and normally only one hit on one wallet happens at a time and a very small hit.
Binance have got no good leverage system in place for this and then only good leverage is bitmex and they only have pairs connected to BTC, so no triangle is possible.

Still there is a lot to do.
Attached File(s)
File Type: zip CryTri2018-06-10 10h56s42m820.zip   52 KB | 842 downloads
0 traders viewing now