(Closed) Butter Bot!: Premier Bitstamp, and BTC-E EMA Trading Platform (Closed)

[fible1]

I am looking into buying butter. Right now I am using the demo. Please explain this to me:

EMA diff is at -0.296% (1 BTC = 118.166 USD) at btce --- Sept 6
EMA diff is at  0.256% (1 BTC = 118.2 USD) at btce -- Sept 8

How does this make money?

If I had 1 BTC and it sold for 118.116 and then I bought at 118.2, would I not have less than I had before?

Hi Bruce ,
   I'm guessing you posted the trade prices of your last two trades?

Like has already been stated, EMA will likely generate about 70% small loss trade (like the one above) and 30% profit trades that would account for all the small losses plus a healthy profit. Also, as back testing will show, you need to run the bot for at least a couple of weeks on uninterrupted, un-interfered with trading to let the math kick in and get rid of mean deviators. Think "The Law of Large Numbers" and you will get a good idea of what I mean.

Back testing and community response is a better gauge of profitability than short term testing .

Please do not hesitate to let me know if you need anything, I am here to help .

Pablo.

[Pangia]

In regards to your request for feedback --- while an arbitrage bot could be very profitable, the issue that we face is the extraction of the funds to place them back into another exchange to permit the arbitrage to work.  If we look at MTGOX for example, the time frame for withdrawing funds could be forever and recently there have been several posts about not being able to withdraw BTC's for a variety of reasons.  The arbitrage bot would be an alert system and therefore the actual moving of BTC's and funds in an out of exchanges would be up to the customer (time consuming). The profits would also be eaten up (factoring in trading fees and other fees associated with the depositing of funds into financial entities to get the funds into a BTC exchange of course a great arbitrage spread could make up for that, but how often would those opportunities surface).

Not true.  As someone who does it manually, there is more profit in arbitrage then there is in trading right now; it is just time consuming and a lot of work.  A bot would change this.

Alts are what makes arbitrage fast & easy.  You will notice that exchanges that support alts stay pretty close.  As they drift apart, there is too much profit to be made through arbitrage using alts for them to stay that way long.  This is healthy for the market, and very profitable for those with arbitrage bots.  _I want one!_

Based on GOX's announcements, their next generation trade engine will support at least LTC.  This opens a massive opportunity for arbitrage profit, and will cause GOX's price to quickly equalize with BTC-e & the other alt-supporting exchanges and stay there.  Those with a bot ready & waiting for this stand to profit nicely, and the volume difference between GOX and the others will mean profits will continue as arbitrage brings the rest of the exchanges into step with GOX as it moves up & down.

At last check, there were literally hundreds of stock-bots... and have been for a long time.  I have played with stockbots over the years, and can purchase them at will.  What I have not found is a good (read "safe) release of an arbitrage bot.  This so far is a closely guarded profit maker, & I want in.

Hi Sean,

Thank you for sharing your thoughts.  Based on what you've shared I have to reconsider my opinion.  I did not give enough thought about an arbitrage opportunity involving the alt coins. 

Thanks.

[fible1]

In regards to your request for feedback --- while an arbitrage bot could be very profitable, the issue that we face is the extraction of the funds to place them back into another exchange to permit the arbitrage to work.  If we look at MTGOX for example, the time frame for withdrawing funds could be forever and recently there have been several posts about not being able to withdraw BTC's for a variety of reasons.  The arbitrage bot would be an alert system and therefore the actual moving of BTC's and funds in an out of exchanges would be up to the customer (time consuming). The profits would also be eaten up (factoring in trading fees and other fees associated with the depositing of funds into financial entities to get the funds into a BTC exchange of course a great arbitrage spread could make up for that, but how often would those opportunities surface).

Not true.  As someone who does it manually, there is more profit in arbitrage then there is in trading right now; it is just time consuming and a lot of work.  A bot would change this.

Alts are what makes arbitrage fast & easy.  You will notice that exchanges that support alts stay pretty close.  As they drift apart, there is too much profit to be made through arbitrage using alts for them to stay that way long.  This is healthy for the market, and very profitable for those with arbitrage bots.  _I want one!_

Based on GOX's announcements, their next generation trade engine will support at least LTC.  This opens a massive opportunity for arbitrage profit, and will cause GOX's price to quickly equalize with BTC-e & the other alt-supporting exchanges and stay there.  Those with a bot ready & waiting for this stand to profit nicely, and the volume difference between GOX and the others will mean profits will continue as arbitrage brings the rest of the exchanges into step with GOX as it moves up & down.

At last check, there were literally hundreds of stock-bots... and have been for a long time.  I have played with stockbots over the years, and can purchase them at will.  What I have not found is a good (read "safe) release of an arbitrage bot.  This so far is a closely guarded profit maker, & I want in.

Hi Sean,

Thank you for sharing your thoughts.  Based on what you've shared I have to reconsider my opinion.  I did not give enough thought about an arbitrage opportunity involving the alt coins. 

Thanks.

To further develop this, how do you guys visualize the mechanics of such a bot?

As Pangia noted, extracting funds through wire would be very difficult to automate so I imagine we would need to do all our trading in crypto-coins?

Example: Buy BTC in Gox, Sell for LTC in BTCE.
With LTC in BTCE, wait for profit opportunity and sell those for BTC in Exchange X for net profit?

Pablo.

[balanghai]

anyone using this one? And starting bitcoin capital in trading?

[fible1]

anyone using this one? And starting bitcoin capital in trading?

Hi Balanghai ,
   A lot of people using this one, and we recommend 250-300 USD starting capital; of course the more capital you have, the smaller the swing you need to make a profit.

Feel free to peruse the thread or download a demo at Butter-Bot.com .

Pablo.

[seanrarey]

Ok... off the top of my head, but is something to start the building process for an arbitrage bot, using GOX & BTC-e as examples:

BTC value GOX = BG
BTC value BTCe = BB
LTC value ratio between houses = LR
BTC/LTC ratio on GOX = GR
BTC/LTC ratio on BTCe = TR
Console-adjustable variable = X
 
Rule 1:  BTC & LTC accounts in both houses to remain equal (GOX & BTCe)
Rule 2:  If GR > TR by X% buy BTC on BTCe & sell BTC on GOX.  Invoke Rule #1 to equalize.
Rule 3:  Flip side of Rule #2
Rule 4:  If BG -LR>BB by X% sell GOX buy BTCe.  Invoke rule #1 to equalize.
Rule 5: Flip side of Rule #4.
 
A crude place to start, but would give a coder an idea of what we are looking for.  Rule 1 is a key.  The bot would be able to act fast on an imbalance, to the extent we could fund it.  It will take an hour to "recharge" in BTC, less to recharge LTC to be able to "fire" again; but because transactions are locked in at the time of trigger you never risk any loss during recharge... you just can't do any more trades until Rule 1 cycle is complete.

[tralala33]

The problem is that the exchange APIs do not allow for withdrawing funds if I'm not mistaken, so all the transfers between the exchanges would have to be done manually. Or am I missing something?

[seanrarey]

The problem is that the exchange APIs do not allow for withdrawing funds if I'm not mistaken, so all the transfers between the exchanges would have to be done manually. Or am I missing something?

At this point this is true.  If automated transfer were to be negotiated with the exchanges, the bot could make the transfers.  If not a simple re-charge routine ("send x number of LTC to x address") could make "recharging" a one click process.

[fible1]

****************
V2.1.1 Release
****************
Friends,
   We have just released Butter-Bot V2.1.1; which fixes some minor issues with the V2.1 release, particularly as related to Bitstamp trading. The release should be in your inbox, please take a few minutes to upgrade .

Thank you,
   Pablo.

[Grinny]

Please find the release attached and take a few minutes to upgrade .

....your mail has no attachement^^

[edit]there we go [/edit]

[seanrarey]

Further discussion on an arbitrage bot...

The need/desire for a bot to conduct or assist with arbitrage, particularly when alts become involved, is based in the fact that there is a lot more math  involved in these decisions then with simple trading.

When trading, buy low sell high.  Sure, the system (in the case of Butter EMI) involves more math, but a beginner can look at a transaction and make a snap decision because when selling high=good low=bad.

Now lets start an arbitrage transaction using alts.  Keep in mind, the market is moving, moving fast, and will change when you blink your eye.  So make the following calculation taking no more then 4 seconds:

GOX BTC=131 LTC=1.70
BTC-e BTC=128 LTC=1.66

Quickly now, which way do you move the money?  The ratios between the alts must be calculated against the ratios between the mains and the exchanges.  By the time you do this manually the numbers have changed.  A _computer_ on the other hand...

... do you guys see what a valuable money-maker an alt-driven arbitrage bot could be?

[faetos]

I tried the bot out for the 3 day trial period and ended up losing money on it. Has anyone had any success with it? I'm not sure to buy it or not. I have it on btc-e with 5 min frequency, short 10, long 25, buy 0.1, sell 0.25.

[fible1]

I tried the bot out for the 3 day trial period and ended up losing money on it. Has anyone had any success with it? I'm not sure to buy it or not. I have it on btc-e with 5 min frequency, short 10, long 25, buy 0.1, sell 0.25.

Hi Faetos ,

    Those are MASSIVELY bad settings. At 5 minutes, you are really using the bot as a high frequency bot, which it is not intended to be, it is actually a trend bot. The fees probably ate any profit you may have generated. Back testing will show you that the higher frequencies are much better, we are currently recommending 2 hours. The lower frequencies exist because some customers asked for them for testing purposes, they are not recommended. Please see the delivery email or the bot manual you got with your demo for the current recommended settings.

I will let others answer as to profitability (or suggest you peruse recent and past comments). Please feel free to play with back testing and you can see these results for yourself .

Here are the results for back testing with your settings vs. the recommended settings for the same time period:

Recommended Settings:



Your Settings:



Please do not hesitate to let me know if you need anything .

Pablo.

[spikejnz]

I tried the bot out for the 3 day trial period and ended up losing money on it. Has anyone had any success with it? I'm not sure to buy it or not. I have it on btc-e with 5 min frequency, short 10, long 25, buy 0.1, sell 0.25.

This has been covered over, and over, and over. This is not a SHORT TERM trading application. A good chunk of your trades will be losing trades (that's just how EMA Trading goes), but overall you will be up percentage wise. I have about BTC5.5 in my MTGox account, and that number just keeps going up over the past two months. I won't be withdrawing money for QUITE some time.

[seanrarey]

I tried the bot out for the 3 day trial period and ended up losing money on it. Has anyone had any success with it? I'm not sure to buy it or not. I have it on btc-e with 5 min frequency, short 10, long 25, buy 0.1, sell 0.25.

The demo version is good for getting a feel for the bot's controls and look, but you are _not_ going to make money with it in 3 days.  Moreover, your settings are way outside the recommended settings.

I have been using it for a few weeks, and have made very good returns.  The bot also has a backtest feature that will show you what you would have made over the past x time period with any given settings.  Punch in the recommended "beginner" settings in the bot, and click "test settings".  Now go back two months and see what your profits would have been if you had let the bot run.

Um, that should really speak for itself.

[spikejnz]

Did MTGox change their fee schedule and not update their page? The default fee for back-testing in BB is .55, which would be for BTC100 to <BTC200. I doubt most people using this will be trading that many coins ... for now.

 

Just FYI, everyone may want to adjust your fee according to this schedule: https://www.mtgox.com/fee-schedule

[fible1]

Did MTGox change their fee schedule and not update their page? The default fee for back-testing in BB is .55, which would be for BTC100 to <BTC200. I doubt most people using this will be trading that many coins ... for now.

 

Just FYI, everyone may want to adjust your fee according to this schedule: https://www.mtgox.com/fee-schedule

Solid point, we will correct this for the next release, we used the fee on hand, we should have double checked that that was the highest starting fee .

Pablo.

[Grinny]

Add the the list of wonderful additions to butter that I am hoping for are the ability to test against data from different exchanges (not just GOX), and the ability to tune the settings on the different exchanges individually based on these tests.

+10

[faetos]

Thanks for answering my question even though it may have been covered before. I appreciate the advice and the detailed descriptions on the settings.

[chanberg]

I am on day two, and I dont really understand the logistics of it, but I will give it a few more days (will be applying for another trial key).

So far, it seems to have sold low and bought high. I am really not sure , but from what i can tell it takes the differences in percentage of the price of atleast two ticks, before it does a buy or sell... But from what I can tell, i dont understand why it does a sell even when the percentage difference is in the negative.. Shouldn't it only initiate an order when the percentages are positive?

But then again i dont really know what the hell is going on, so i will let it hang out for a bit while it does its thing..

I deposited about 5 BTC, and its hanging around 4.8 now from the past 2 days... so we will see

Danny