Topic: (Closed) Butter Bot!: Premier Bitstamp, and BTC-E EMA Trading Platform (Closed) - page 77. (Read 274784 times)

IMO a method of tracking profits & losses is sorely needed. I recently had to give up on using MY-BTC.info to track mine because the manual entry of all the little buy/sell orders every time the bot triggered was just too much to bear. And now I have no tracking method at all, which is very unfortunate.

I just wanted to chime in & mention that I would like to see the developers spend time on this.

Hi Pablo, has there been any talk about making a trade log ala backtesting format for real performance review? My bitstamp transaction history is a mess and I'd have to break out a calculator and an afternoon just to make sense of it all.

Hi Pablo.
Thanks kindly!

Hi Kweks ,

   1. You can start with Fiat or BTC or both, it doesn't make a difference.

   2. Yes, you can trade on one exchange, or on all 3. For the time being we only have global settings, but the next release should allow you to set each exchange individually.

 Please let me know if you need anything at all .

Pablo.

Hi flyboy. That sounds good in theory, but we can't always pick the direction the market will go in. For example, if you sold after a $5 price increase, sure you lock in a small profit, but what if the price kept going up? What if it was a huge movement, like a $15 increase? What if a new bubble forms? You wouldn't want to be left behind.

In the same way going down, how do we call the bottom? What if it drops $5 quickly, you buy, and then moves back upwards even higher (called a bear trap) and leaves you on the wrong side of the trade? Now you must either buy in at a higher price, or hold the cash, hoping it eventually goes back down.

The bot watches 24/7 and attempts to mathematically put us at an advantage to profit long-term using exponential moving averages. When the market ranges between values, say $170 and $175, most trades won't be profitable and it's frustrating to watch. But when the price makes a move and begins to move strongly upwards or down, the bot ensures we are on the correct side of the trade to profit from the trend.

For example, my bot bought in correctly yesterday and rode the $10 price rise. The three trades before, however, were a small loss. The hope is that gains on the larger trends will make up for the smaller losses.

I'm a recent investor in the bot, let me share my situation which is similar to yours. I tried it out a few times and was like, what the hell I'm losing BTC, but I had to remind myself it's EMA, it's long term--it's not simply "buy low sell high" (even though that's the general premise).

Basically, short term typically won't result in profits, so trying it out for a few days you won't see a profit (unless you happen to start the bot when the price is high and suddenly drops, which is rare). If you could try it out for, say 2 or 3 weeks, you would see the benefit. The tryout period is more to show how it works and that it would eventually work if you hit one of those magical trade points.

I have had about 3 weeks of trades (a few trades, not many) where most were losses by a single percentage, however, recently when there was a landslide drop in price the bot grabbed some more BTC. I knew the price would then go up and eventually sell high, it was a matter of waiting out and let the bot do its job. I funded my trade accounts with an extra BTC just to make sure to maximize my profit.

Sure enough eventually it grew and again dropped like 20-30$, and the bot sold.

The bot was holding some btc at $115 and some at $120, the peak value was $224/205 (mtgox/btce), and with my settings it sold at $208/190. Instantly made some USD which went back into those trade accounts.

About a day or 2 later the bot picked up on the next EMA rise in price and bought BTC for $190/178, and the price is slowly ticking upward again.

So I made BTC. Not a lot, but it's already paid for itself, so from here on out unless there's an epic failure of BTC, I'm going to be gaining BTC.

You have to wait it out, it's not a magical coin fountain, and you need to find what settings work for you. Mine are not the default settings, but even the default settings work well.

Here's a link to a read which might be helpful, the OP was exploring some settings similar to what I was. Mine aren't exactly what he has but they're close enough, and I find the settings work well for my "tastes": https://bitcointalksearch.org/topic/evaluating-parameters-for-ema-crossover-trading-a-rigorous-approach-kind-of-290961

Some folks don't like waiting things out and get too edgy, so it may not be a game for them to play; what I tell people is to simply setup the bot and forget about it. Check on it every  now and then to make sure it hasn't crashed, but don't worry yourself with if it's going to work. The defaults alone will work.

As to funding, I would recommend having at least half a BTC (or cash equivalent) per trade account to really see the system work. I started out with .2BTC but it was only profiting by fractions. So to visually realize the gains, .5 is a minimum, I would really say 1BTC to get the feel for it.

Good luck to us all, cheers.

Hi there.
Grabbed a copy of the bot over the weekend.

After seeing in the backtesting that it would have made trades that it didn't, I had some questions:

1. Do the accounts (MtGox, etc) initially need to have funds in USD, or BTC, or both (or either?) - Currently I have my funds in BTC
2. Is it possible to trade with multiple exchanges, and if so: how do I set the parameters for each specific exchange?

Thanks kindly.

Hey ,
  Yes, for now you need to keep Chrome open for the bot to work; we are working on a full hosting solution that should be available in a few weeks.

I would suggest that you trade with a small amount of money, if you are not in it for the long haul you really should avoid trading; you will very likely see losses.

That said, thank you for giving our bot a try, please let me know if you need anything at all .

Pablo.

I will keep watching the bot.

Do I have to keep crome open for it to run or just the computer on?

At a loss of 8.8% per trade, I will just let it ride and see what happens. Thanks for the free demo.

I would much rather a bot that does what you said. Buys at a set decrease and sells at a set increase.

Hi Flyboy ,
  I explained above how mean deviating trades happen, it's not about this bot or a bot, all trading systems are subject to mean deviating trades because when looked at independently each trade can be seen as the result of randomness; only when you aggregate them and look at them as averages over the long term can you make an accurate observation about the trading system. In other words, you can't judge any trading system at all by a single trade or a small group of trades.

Please take a look a deeper look into trading systems and the mathematical principles I have explained to you and you will understand will see what I mean .

Please let me know if you need anything at all :=).

Pablo.

Hey ,
  Do you mean a bot that will buy if the price halves and sell if the price doubles?

There is a lot that can for wrong with that, we have been discussing reverse EMA for a while that does something similar (but better) with EMA indicators. Maybe something we can pursue when we get some free development time. Got a lot going on right now .

Pablo.

It doesn't make since to have a bot that buys or sells at a loss. It's math. If the bot sold at 188 it should know not to buy back at 205...

I hope we get some feedback from others but here goes me:

Bought in around 120's I believe, Held past the 170's plateau when everyone wanted to sell (see previous posts), then went to the peak and sold in the upper 180's.

I don't have the exact figures on me but that's more or less it. I'm test running really slow settings so others may have fared slightly differently.

Pablo.

P.S. I got bought in at the upper 170's again and now we're back to the 190's so it's been a good couple of days .

Can u make a bot that starts at say X and if the value drops 2$ you buy and if it increases 4$ you sell?

i was just wondering, how did the bot go doing the "booom" when bitcoin went from 130 to 220?

Hi Flyboy ,
   With EMA about 70% of your trades will be small loss trades and the 30% profit trades should cover any losses plus a decent profit. Now when there is extreme volatility you get what in mathematics are called "mean deviators" profit or loss trades that are way out of the expected average range. It is because mean deviators are inevitable in any market, especially one as volatile as Bitcoin, that we say EMA is a long term strategy; it takes a large number of trades (the more the better) to bury mean deviators and for the math behind EMA to kick in (Think: The Law of Large Numbers).

We have had this discussion in the thread before and people using our bot tend to come out ahead over the long term. In the short term, we acknowledge that trades like the one you experienced may happen because a small number of trades will tend to produce random results (be they profit or loss). You have to either decide to stick with it long term or not do it at all; this is not a "Get Rich Quick" scheme.

Please let me know if you need anything at all .

Pablo.

by the way, thats an 8.8% loss in one trade...

sounds to me like I just need to hold BTC.....

Sounds like you need to read more about ema trading. Many, if not most, of the trades will be unprofitable.

So far I'm going to so no to butterbot

Started it on Friday with default settings for gox

Started with 7.9 @ 195ish

First sale 188$ the lowest point this weekend.

First buy 205$ very close to the high for the weekend.

I went from 7.9 to 7.2 in two days.

At this rate I will be broke in a month.