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

[vesperwillow]

I know two weeks isn't a good indication for trades based on ema indicators, but using one of the weeks I used the bot for back testing, the bot says I should have made 46% using the same settings I did during live trading.
Thanks.

I echo what the guy above me said, but I would point out it's likely you just have bum settings if it's selling is that far off the mark. I spent hours last week finding settings which worked for each individual exchange to ensure something like that didn't happen.

[1714714562]

1714714562

[1714714562]

1714714562

[1714714562]

1714714562

[1714714562]

1714714562

[madpoet]

Heh, wish BB had been smart enough to execute the sale at $900!  Ah well, I wish I had been smart enough to also

[simiane]

So, the bot didn't do very well for me in this latest bit of volatility... I'm on BTCe with the stock settings.
The high point was recorded at over $800, but the crossover didn't occur until many hours later. It eventually sold me out at $498.
At the time of writing, price is rising back up pretty quickly, but I'm in fiat now! Presumably it'll buy me back in soon, but I will have missed out on about $70 and will have less BTC than I had before.

It's fairly understandable that the bot would be blindsided by such violent swings in price, but I'm not sure why it took so much longer to sell me out. Obviously, I'm not saying it's the bot's fault, it's just going on the EMA settings I have. I don't really understand how EMA is calculated so I just have to trust it to do it's thing.

I've been trying to experiment with settings, but the backtesting doesn't seem to correlate with live trades. I have no idea why that is - according to the backtester, I should still be in BTC. Perhaps it's delayed?

Plenty of settings have been posted on the forum, but it seems that the volatility of the market plays quite a huge role. In this current market, I've missed a lot of opportunities.
Does anyone have any suggestions for profiting while BTC is up and down in this yo-yo style manner? Other than 'stop using the bot', obviously.
Any settings that are likely to take advantage of quicker dips and rises?
I'm not too sure what I'm asking for here, but some guidance would be appreciated.

[San1ty]

So, the bot didn't do very well for me in this latest bit of volatility... I'm on BTCe with the stock settings.
The high point was recorded at over $800, but the crossover didn't occur until many hours later. It eventually sold me out at $498.
At the time of writing, price is rising back up pretty quickly, but I'm in fiat now! Presumably it'll buy me back in soon, but I will have missed out on about $70 and will have less BTC than I had before.

It's fairly understandable that the bot would be blindsided by such violent swings in price, but I'm not sure why it took so much longer to sell me out. Obviously, I'm not saying it's the bot's fault, it's just going on the EMA settings I have. I don't really understand how EMA is calculated so I just have to trust it to do it's thing.

I've been trying to experiment with settings, but the backtesting doesn't seem to correlate with live trades. I have no idea why that is - according to the backtester, I should still be in BTC. Perhaps it's delayed?

Plenty of settings have been posted on the forum, but it seems that the volatility of the market plays quite a huge role. In this current market, I've missed a lot of opportunities.
Does anyone have any suggestions for profiting while BTC is up and down in this yo-yo style manner? Other than 'stop using the bot', obviously.
Any settings that are likely to take advantage of quicker dips and rises?
I'm not too sure what I'm asking for here, but some guidance would be appreciated.

It's sad that you are using a bot without understanding it...

MA is designed to give a smoothing effect therefor filtering out noise like this insane volatility.

But anyway, just shorten your MA Periods.

[123orangebud]

then drag and drop the file attached to this
email into the same page, you are all set to go!?

No Robot attached...

[NginUS]

So, the bot didn't do very well for me in this latest bit of volatility... I'm on BTCe with the stock settings.
The high point was recorded at over $800, but the crossover didn't occur until many hours later. It eventually sold me out at $498.
At the time of writing, price is rising back up pretty quickly, but I'm in fiat now! Presumably it'll buy me back in soon, but I will have missed out on about $70 and will have less BTC than I had before.

It's fairly understandable that the bot would be blindsided by such violent swings in price, but I'm not sure why it took so much longer to sell me out. Obviously, I'm not saying it's the bot's fault, it's just going on the EMA settings I have. I don't really understand how EMA is calculated so I just have to trust it to do it's thing.

I've been trying to experiment with settings, but the backtesting doesn't seem to correlate with live trades. I have no idea why that is - according to the backtester, I should still be in BTC. Perhaps it's delayed?

Plenty of settings have been posted on the forum, but it seems that the volatility of the market plays quite a huge role. In this current market, I've missed a lot of opportunities.
Does anyone have any suggestions for profiting while BTC is up and down in this yo-yo style manner? Other than 'stop using the bot', obviously.
Any settings that are likely to take advantage of quicker dips and rises?
I'm not too sure what I'm asking for here, but some guidance would be appreciated.

I'm using the "recommended" settings for BTC-e, and my bot sold me out @ $499, similar to you, now the market's back up, & I'm with fiat. I'm afraid I'll end up with less BTC than I had before after this now once it buys in at a higher price eventually.

Pablo, these "recommended" settings are no good if this is what they're going to do! I need some better settings.

Could someone recommend some settings that would sell closer to the peak of $800 than $499, and still buy low, please? I don't understand EMA enough to come up with them myself.

Pablo, this is the kind of bad trading I hope the devs can code a better way to protect against. Maybe they don't need to code anything- maybe I just need better settings, I don't know.

All I know is I just lost a bunch of money & I don't like it.

[bassclef]

The bot doesn't exist to make you money. It auto trades EMA lines with parameters you set.

If you don't understand this method of trading DON'T USE THE BOT. EMA trading is a long-term strategy.

In fact there are smart traders who know that noobs use this bot on default settings and trade against it. I'm sure they are grateful for your extra coins. Be careful.

[vesperwillow]

So, the bot didn't do very well for me in this latest bit of volatility... I'm on BTCe with the stock settings.
The high point was recorded at over $800, but the crossover didn't occur until many hours later. It eventually sold me out at $498.
At the time of writing, price is rising back up pretty quickly, but I'm in fiat now! Presumably it'll buy me back in soon, but I will have missed out on about $70 and will have less BTC than I had before.

It's fairly understandable that the bot would be blindsided by such violent swings in price, but I'm not sure why it took so much longer to sell me out. Obviously, I'm not saying it's the bot's fault, it's just going on the EMA settings I have. I don't really understand how EMA is calculated so I just have to trust it to do it's thing.

I've been trying to experiment with settings, but the backtesting doesn't seem to correlate with live trades. I have no idea why that is - according to the backtester, I should still be in BTC. Perhaps it's delayed?

Plenty of settings have been posted on the forum, but it seems that the volatility of the market plays quite a huge role. In this current market, I've missed a lot of opportunities.
Does anyone have any suggestions for profiting while BTC is up and down in this yo-yo style manner? Other than 'stop using the bot', obviously.
Any settings that are likely to take advantage of quicker dips and rises?
I'm not too sure what I'm asking for here, but some guidance would be appreciated.

I'm using the "recommended" settings for BTC-e, and my bot sold me out @ $499, similar to you, now the market's back up, & I'm with fiat. I'm afraid I'll end up with less BTC than I had before after this now once it buys in at a higher price eventually.

Pablo, these "recommended" settings are no good if this is what they're going to do! I need some better settings.

Could someone recommend some settings that would sell closer to the peak of $800 than $499, and still buy low, please? I don't understand EMA enough to come up with them myself.

Pablo, this is the kind of bad trading I hope the devs can code a better way to protect against. Maybe they don't need to code anything- maybe I just need better settings, I don't know.

All I know is I just lost a bunch of money & I don't like it.

I don't have any of my settings on hand but if I can get to them later I'll give you some to see if they might help. Best thing to do is to lower your EMA's and play with the timing to make sure you're not real far behind the curve--that's if you're going to let your bot manage it entirely.

Example, when I saw the volatility today, I killed my BTCE bot but let the MTGox one go (as it had settings geared more toward the flux). BTCE had already sold high for me, so I was sitting nice when I killed it as I was holding very low. I then purchased manually on a second downswing and am holding until I can get time later to work on new BTCE settings.

Sometimes you need to keep an eye out on this stuff.. hope your luck changes. The good news folks, is that BTC is going to keep going up.

[NginUS]

I changed my settings to one with much lower EMA values, and which still does well in backtesting. A modified version of the one Velacreations posted which got 7000% several pages back.

I hope it'll buy at a good price should the market take another swing down. Otherwise, I'm going to be stuck with fiat for a long time it would seem.

[vesperwillow]

This situation is the very reason I had mentioned a bot-idea to Pablo in the past, regarding backtesting--being able to select same-day testing (only), or a time frame  so you could specify the previous day or 2 up to present.

This way if things didn't go how you would've liked, you could twiddle with settings to find what would've been better, and then make a note of those settings for future flux.

Hopefully it is a feature Pablo can make happen

[simiane]

So the consensus is that you need to adapt your settings depending on what the market's doing over any given timeframe?
I can play with the EMA settings and get the graph to more closely match the market over the last couple of days and indeed it results in a much higher profit in back testing for the last week. Backtest for the last 6 months however and it results in a drastically lower profit than the stock settings.
Presumably then, when the market starts going crazy I should either have a play with settings until I get lines that are closer to the market movement, or switch off the bot entirely and trade manually - then switch to a longer term setting like the suggested ones for the day-to-day?
I've not had much luck trading manually though... which is why I like the idea of the bot taking control!

[madpoet]

I changed my settings to one with much lower EMA values, and which still does well in backtesting. A modified version of the one Velacreations posted which got 7000% several pages back.

I hope it'll buy at a good price should the market take another swing down. Otherwise, I'm going to be stuck with fiat for a long time it would seem.

It's a long term strategy, not a day trader bot.  You have to accept that.  You can look at the charts and see clearly why its not going to trade.

[NginUS]

Here are my new settings, already its trading, not at a loss thankfully.

What do people think of these? Could they be improved?

[bruter]

Pablo

I would like to see 'don't sell at a loss' feature.

Please

[NginUS]

Hey Pablo,

I think it would be nice if the bot would provide a chart with EMA lines so we could visualize exactly what different tweaks in our engine settings are doing. It would help noobs like me understand the product.

[fible1]

Hey Guys ,
  Lot's of activity .

About the "Do Not Trade at a Loss" feature, results are not yet conclusive, it shouldn't be overly hard to code, we just don't want to include a feature that looks "good" but that could end up creating losses; especially given that a good portion of our users are new to trading.

About losses or profits in the last market movement, I will say again: you can not account for success or failure with individual market movements you have to look at them in aggregate over a period of time when using EMA. I have been getting emails from people who have had quite some success these past two weeks by tweaking their settings to this particular market (lower time frames, alternate thresholds, etc.).

The issue with this is that it can often work well. Why is this an issue? Because although Butter may generate some losses in a given market, and tweaking Butter to what you as a user perceive the market to be can generate some profits in this market, the math tells us that if we were to look at 50 or 100 markets over a longer span of time the EMA strategy would win out in the end.

I was much more militant about this in the beginning because we designed Butter with the Mathematics of Goomboo's quite in depth discussions and theories in mind but I can understand the lure of immediate profit, in the end it is a personal choice of how you wish to use Butter, it will work either way, you just have to decide what time frame you are looking at.

The "Starter Settings" will stay as they are with the long term view, but please keep in mind you are meant to tweak these as you gain understanding of Butter and according to your own risk profile.

We are also working on improving the trading engine with some other indicators, but it's quite premature for me to promote this as we are still trying to figure out what works and what doesn't; we want to make sure that whatever we release it is not only an improvement over the current engine, but an important improvement over the current engine.

For any issues of a one to one nature (tech support, refunds, etc.) please drop me an email at Pablo@Butter-Bot.com .

[uck]

Hey Pablo,

I think it would be nice if the bot would provide a chart with EMA lines so we could visualize exactly what different tweaks in our engine settings are doing. It would help noobs like me understand the product.

+100  Can someone retrofit the backtest data output to produce something if Pablo can't get it into the Roadmap?

[boozezela]

Hi Pablo,

I think it would be very useful to put a warning that says "WARNING: THE BOT WILL BUY AT THE CURRENT PRICE, WHATEVER THE MARKET CONDITION IS AS SOON AS YOU ENABLE TRADING. (I UNDERSTAND/STOP THE BOT)".

In the future you should add an additional per-market setting like "Initial buy price: XXX".
Then when you click on the "enable trading" link or button the BOT should show an alert that says: "Do you want to wait until a price of XXX has been reached before allowing the BOT to buy? (YES/NO/CHANGE THRESHOLD)"

Of course the BOT should use the threshold only when trading goes from disabled to enabled and not during its normal operation.

This could help with users that start to try the BOT in a day that shows a lot of fluctuation and avoid a situation where the BOT buys at 700 and sells 30 seconds later at 500.

[NginUS]

In my case today it was selling at whatever the market price was as soon as I enabled trading, too. It's causing enormous problems.

Now I'm waiting for a manual BTC-e order to go through & replenish my BTC so I can go do it again this time catching it in the logs and sending a proper bug report so it can be fixed.

But since I'm trying to buy real low on the exchange in order to recoup my bot losses today it may be quite awhile until my order gets filled. *Sigh

[boozezela]

OK, so the BOT should also have an initial sell threshold price when you enable it (it could be the same threshold price for both buy/sell actually, but I am sure users would like the extra flexibility).

If the BOT in going to buy the alert should be:

Code:

Do you want to wait until a price of XXX has been reached before allowing the BOT to buy?
YES(checkbox)/NO(checkbox)/CHANGE THRESHOLD(link)

WARNING: If you select NO, the BOT will buy immediately at $CURRENTPRICE. [this could come out in the alert when you check "NO"]

I AM ALL SET, ENABLE TRADING (button).

If the BOT in going to sell the alert should be:

Code:

Do you want to wait until a price of YYY has been reached before allowing the BOT to sell?
YES(checkbox)/NO(checkbox)/CHANGE THRESHOLD(link)

WARNING: If you select NO, the BOT will sell immediately at $CURRENTPRICE.  [this could come out in the alert when you check "NO"]

I AM ALL SET, ENABLE TRADING (button).

So the user will know what is going to happen.