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

[San1ty]

Pablo & company, I salute you in your efforts. 

Some pretty counter-intuitive suggesting being brought back up here.  I am going to make my comments brief, because this is all very worn territory.

Profit/loss:
"Magic" does not exist in trading.  Only solid, proven systems.  _No_ system is without loss.

Stop-loss:
EMA cross-over trading _is_ stop loss.  It is designed to keep you in a bull market and out of a bear market.  Attempting to "add" a stop-loss to this system creates a double-negative and short-circuits the system.  You will lose.  If you try to bail on your position as the crash occurs, you are selling against the pressure and you will lose much more then if you wait for the snap-back.

"Flash-crash" protection:
When trading manually, the best protection in a flash-crash is to walk away & go eat dinner.  If you join the panic buy/sell, you lose.  You are _always_ far ahead when you wait and catch the rebound caused by the heavy market pressure.  I have had great success programming the bot to do this automatically.  It often means slowing the trade down; you need to wait for the snap-back.

This bot is wildly profitable, and very configurable.  Most everything you are asking for is already here; you just need to understand how the bot works and how to set it up.

Your stop loss remark is incorrect, a trailing stop-loss can be used perfectly in conjunction with ema-crossovers.

Example:
Price goes on a steep climp from 20 dollars to 260 dollars and then severely corrects to 80 dollars in a couple of minutes.
Your EMA crossovers would react when the action has already happend (sell at 80)

while a trailing stop-loss of 20% would try and cut your losses instanteniously. After that your bot can enter the market again based on it's own indicators.

[fible1]

Stop-loss:
EMA cross-over trading _is_ stop loss.  It is designed to keep you in a bull market and out of a bear market.  Attempting to "add" a stop-loss to this system creates a double-negative and short-circuits the system.  You will lose.  If you try to bail on your position as the crash occurs, you are selling against the pressure and you will lose much more then if you wait for the snap-back.

I am using reverse EMA strategy so I am unsure if it applies.

" If you try to bail on your position as the crash occurs, you are selling against the pressure and you will lose much more then if you wait for the snap-back."

I'd also like to point out that some of the same arguments are being used in reverse on either side of the issue.

If a flash crash happens says side A, then EMA will keep get you out of it. There is no guarantee the market will go up again we are told.  People pointed out that there is no guarantee that the price will recover. Alright, that is probably true.

Now the same point is being used (the snap back "will happen") as-if it were guaranteed....

Perhaps we need to all get on the same page.

===========================

Either way, I understand both positions to the best of my ability.

=======[Discount Code for Butter (10% off)]=========

https://butter-bot.com/?r=f5a17f5b

Use my referal code cause' im sooo cute!  



You'll get an extra month free, it's all on me!

A good experiment for you would be to buy Stephen Haas' Simple Trade Bot.  It is versatile, and will let you do most if not all of what you are suggesting.  I have run it side-by-side with Butter for months.  It has never come close to matching Butter's performance.

RSI, "reverse" EMA, MACD all seem like a great idea; until the market runs away from it and strands the bot.

With the different bots I run (if it is available, I buy it), it is the simplest one that constantly makes me the most money.

There are some features I want very much and am looking forward to, but I would be quite happy if they left the trade engine exactly as it is.  I keep coming back to it from the more complex machines.

I have a feeling you don't fully understand those concepts if you just try them out and conclude they aren't profitable.
MACD reacts faster to market movements then the MA crossovers butter-bot uses.

As far as the haasonline bot goes, it's pretty nice and has a lot more options, but the trading systems he currently has implemented for example macd is not according to the spec, so it's hard to judge.

San1ty,
  We will probably be including alternative strategy modules with V3; MACD will likely be a strategy then. We will thoroughly back test all strategies to make sure they are likely to be profitable before we release them, but in principle, I don't see a problem with this; we are all for freedom of choice .

Pablo.

[seanrarey]

Pablo & company, I salute you in your efforts. 

Some pretty counter-intuitive suggesting being brought back up here.  I am going to make my comments brief, because this is all very worn territory.

Profit/loss:
"Magic" does not exist in trading.  Only solid, proven systems.  _No_ system is without loss.

Stop-loss:
EMA cross-over trading _is_ stop loss.  It is designed to keep you in a bull market and out of a bear market.  Attempting to "add" a stop-loss to this system creates a double-negative and short-circuits the system.  You will lose.  If you try to bail on your position as the crash occurs, you are selling against the pressure and you will lose much more then if you wait for the snap-back.

"Flash-crash" protection:
When trading manually, the best protection in a flash-crash is to walk away & go eat dinner.  If you join the panic buy/sell, you lose.  You are _always_ far ahead when you wait and catch the rebound caused by the heavy market pressure.  I have had great success programming the bot to do this automatically.  It often means slowing the trade down; you need to wait for the snap-back.

This bot is wildly profitable, and very configurable.  Most everything you are asking for is already here; you just need to understand how the bot works and how to set it up.

Your stop loss remark is incorrect, a trailing stop-loss can be used perfectly in conjunction with ema-crossovers.

Example:
Price goes on a steep climp from 20 dollars to 260 dollars and then severely corrects to 80 dollars in a couple of minutes.
Your EMA crossovers would react when the action has already happend (sell at 80)

while a trailing stop-loss of 20% would try and cut your losses instanteniously. After that your bot can enter the market again based on it's own indicators.

More likely, your trailing stop-loss would jump you out of the market on a sharp correction that is contrary to the actual trend.

Sir, you don't have to explain your position to me, after 30 years as a professional trader, I understand what you are thinking.  I just don't agree with it.  Long-term I have enjoyed less success with trying to catch every spike and bounce.  In fact, it has led to frustration and loss.

That does not mean that others have not, or cannot have better success.

I will tell you that after tallying the month of October, Butter returned a 48.7% profit on the funds it was controlling.  That far exceeds any other bot I am testing, and beat my manual trades even.

This bot ain't broke! 

[MicroFi]

Pablo & company, I salute you in your efforts. 

Some pretty counter-intuitive suggesting being brought back up here.  I am going to make my comments brief, because this is all very worn territory.

Profit/loss:
"Magic" does not exist in trading.  Only solid, proven systems.  _No_ system is without loss.

Stop-loss:
EMA cross-over trading _is_ stop loss.  It is designed to keep you in a bull market and out of a bear market.  Attempting to "add" a stop-loss to this system creates a double-negative and short-circuits the system.  You will lose.  If you try to bail on your position as the crash occurs, you are selling against the pressure and you will lose much more then if you wait for the snap-back.

"Flash-crash" protection:
When trading manually, the best protection in a flash-crash is to walk away & go eat dinner.  If you join the panic buy/sell, you lose.  You are _always_ far ahead when you wait and catch the rebound caused by the heavy market pressure.  I have had great success programming the bot to do this automatically.  It often means slowing the trade down; you need to wait for the snap-back.

This bot is wildly profitable, and very configurable.  Most everything you are asking for is already here; you just need to understand how the bot works and how to set it up.

Your stop loss remark is incorrect, a trailing stop-loss can be used perfectly in conjunction with ema-crossovers.

Example:
Price goes on a steep climp from 20 dollars to 260 dollars and then severely corrects to 80 dollars in a couple of minutes.
Your EMA crossovers would react when the action has already happend (sell at 80)

while a trailing stop-loss of 20% would try and cut your losses instanteniously. After that your bot can enter the market again based on it's own indicators.

That's exactly my sentiment, San1ty.

@seanrarey:
What if there's a DDOS attack on mtGox and the market tumbles down to 100$ in 10min time?
Will you just wait x weeks / months for the market to recover, or would you rather buy >2x the amount of bitcoins back, because you were able to sell your bitcoins at 220$?

[seanrarey]

@seanrarey:
What if there's a DDOS attack on mtGox and the market tumbles down to 100$ in 10min time?
Will you just wait x weeks / months for the market to recover, or would you rather buy >2x the amount of bitcoins back, because you were able to sell your bitcoins at 220$?

Nope.   Do this:

Go back to every single "tumble" you can find.  The most recent good one is the "silk road crash".

See the hard downward pressure?  Sell here, you sell too low.  "knee jerk" sell reactions and auto stop-loses get you caught in that trap.

Now, move forward in time 1 - 2 hours.  See the rebound?  That much market pressure _always_ results in a correction, or rebound... even when the market then continues down.  _Sell there_.  You will be far ahead, every time.

Discipline makes good traders, and twitchy bots with stop-loss triggers are a favorite target of market manipulators.

[San1ty]

Pablo & company, I salute you in your efforts.  

Some pretty counter-intuitive suggesting being brought back up here.  I am going to make my comments brief, because this is all very worn territory.

Profit/loss:
"Magic" does not exist in trading.  Only solid, proven systems.  _No_ system is without loss.

Stop-loss:
EMA cross-over trading _is_ stop loss.  It is designed to keep you in a bull market and out of a bear market.  Attempting to "add" a stop-loss to this system creates a double-negative and short-circuits the system.  You will lose.  If you try to bail on your position as the crash occurs, you are selling against the pressure and you will lose much more then if you wait for the snap-back.

"Flash-crash" protection:
When trading manually, the best protection in a flash-crash is to walk away & go eat dinner.  If you join the panic buy/sell, you lose.  You are _always_ far ahead when you wait and catch the rebound caused by the heavy market pressure.  I have had great success programming the bot to do this automatically.  It often means slowing the trade down; you need to wait for the snap-back.

This bot is wildly profitable, and very configurable.  Most everything you are asking for is already here; you just need to understand how the bot works and how to set it up.

Your stop loss remark is incorrect, a trailing stop-loss can be used perfectly in conjunction with ema-crossovers.

Example:
Price goes on a steep climp from 20 dollars to 260 dollars and then severely corrects to 80 dollars in a couple of minutes.
Your EMA crossovers would react when the action has already happend (sell at 80)

while a trailing stop-loss of 20% would try and cut your losses instanteniously. After that your bot can enter the market again based on it's own indicators.

More likely, your trailing stop-loss would jump you out of the market on a sharp correction that is contrary to the actual trend.

Sir, you don't have to explain your position to me, after 30 years as a professional trader, I understand what you are thinking.  I just don't agree with it.  Long-term I have enjoyed less success with trying to catch every spike and bounce.  In fact, it has led to frustration and loss.

That does not mean that others have not, or cannot have better success.

I will tell you that after tallying the month of October, Butter returned a 48.7% profit on the funds it was controlling.  That far exceeds any other bot I am testing, and beat my manual trades even.

This bot ain't broke!  

I assumed you weren't very experienced, my mistake. Let's have a healthy discussion then:

You know that MA Crossovers (any kind) work great in a trending market, and I think we can both agree bitcoin has currently known only one trend and that's one big bull rally. So that this bot was profitable should be no surprise to both of us.

That however doesn't mean you can adopt a better strategy.

Let's go back to the trailing stop-loss proposal:
You state that it's possible that the trailing stop loss kicks me out on a sharp correction. And that is a fair argument, but I'd argue that it isn't such a bad thing. Back to my original example:

Price goes on a steep climp from 20 dollars to 250 dollars and then severely corrects to 80 dollars in a couple of minutes.
In this example a trailing stop loss of 20% would have closed my positions at 200 dollar. After that as soon as there is a positive cross-over again the bot can go ahead and buy back in.

The point I'm trying to make is that a 20% correction is pretty significant and probably will have an impact for a while. Your bot can buy back in at 80 dollars and make a significant profit.

Without the trailing stop-loss you wouldn't be able to make such a profit.

Ofcourse buy and hold would also be a winning strategy in this example, but then I ask you, why use a bot in the first place?

In reply to your other reply:
A twitchy stop-loss bot will get manipulated to no end (I agree) but a 20% drop is significant and even if there is a rebound, the first couple of minutes / hours will have intense selling pressure.

Also keep in mind that I'm not proposing a regular stop-loss but a trailing one, so incase of the current trend, and I get manipulated out of my position, I'd still be taking home a healthy profit.

[seanrarey]

Price goes on a steep climp from 20 dollars to 250 dollars and then severely corrects to 80 dollars in a couple of minutes.
In this example a trailing stop loss of 20% would have closed my positions at 200 dollar. After that as soon as there is a positive cross-over again the bot can go ahead and buy back in.

Sounds great... but what ends up happening when I tried this is that too often a sharp correction such as we had a few hours ago would trigger the stop-loss in a scenario, like today, that ultimately did not trigger a dead cross.

BTC is still climbing, but I would be out.  Not a win.  

EDIT:

You are absolutely right regarding the trending nature of the market.  Once that changes, so will the trading strategy.  I think we have a bit of trending ahead of us tho, and intend to capitalize on it.

[San1ty]

Price goes on a steep climp from 20 dollars to 250 dollars and then severely corrects to 80 dollars in a couple of minutes.
In this example a trailing stop loss of 20% would have closed my positions at 200 dollar. After that as soon as there is a positive cross-over again the bot can go ahead and buy back in.

Sounds great... but what ends up happening when I tried this is that too often a sharp correction such as we had a few hours ago would trigger the stop-loss in a scenario, like today, that ultimately did not trigger a dead cross.

BTC is still climbing, but I would be out.  Not a win.  

Indeed, but don't forget that after the stop-loss is triggered, the bot just continues trading.
So if your EMA Diff is positive it'll just take back the position (and take a small point difference loss + transaction fees)

[seanrarey]

Indeed, but don't forget that after the stop-loss is triggered, the bot just continues trading.
So if your EMA Diff is positive it'll just take back the position (and take a small point difference loss + transaction fees)

I like the theory, really I do.

In practice, however, I always seem to lose when I try it.  You are fighting heavy pressure and fast order action, and by the time I am out & back in, I have lost more then I care too.

Price goes on a steep climp from 20 dollars to 250 dollars and then severely corrects to 80 dollars in a couple of minutes.
In this example a trailing stop loss of 20% would have closed my positions at 200 dollar. After that as soon as there is a positive cross-over again the bot can go ahead and buy back in.

Sounds great... but what ends up happening when I tried this is that too often a sharp correction such as we had a few hours ago would trigger the stop-loss in a scenario, like today, that ultimately did not trigger a dead cross.

BTC is still climbing, but I would be out.  Not a win. 

Indeed, but don't forget that after the stop-loss is triggered, the bot just continues trading.
So if your EMA Diff is positive it'll just take back the position (and take a small point difference loss + transaction fees)

Not so small in my experience.  Too much trade traffic and market pressure.  I always lost way more then I cared to, and inevitably, the math showed that I would have been better off going for coffee until the "panic" buying/selling was over, come calmly back in and catch the rebound (providing a trend change is identified).  If on the other hand the spike was not indicative of a trend change, I avoided the trap.

I Found the same with the STB.  I now disable the stoploss in the bot, because it hurts my returns more often then not.

But again, all of that plays into a very methodical approach to trading, and plenty of money has been made with different methods.

I will tell you, however, that how I arrive at my strategy and settings is by spending hundreds of hours with charts testing ideas and theories.  One of the things that is hard to table-test is a stop-loss.  You may have it set for %5, but more often then not it will sell far below that because you are at that point "chasing the trade"; this amplifies with larger amounts of money.

I am also going to come back to a trend that you might be overlooking.  In this volatile market, manipulating fast-acting bots with bull & bear traps is common practice.  One of the things that makes Butter better is it's relative immunity to these market games.

FWIW

[senor_coconut]

Very interesting discussion these last few posts, especially now that we are so high that a correction is ever more likely.
Perhaps things are different now vs. April but still my gut feeling is that this just sounds too good to be true (no need waste keyb strokes grilling me for posting a personal view, eh?!).

I'm quite curious and have setup a poll: http://micropoll.com/a/mpview/1146756-3814481
(cannot figure out how to embed nicely in the post...)

I've also been emailing the developer regarding a workaround for those who e.g. have to go to work and would like to have Butter Bot installed in two computers but actively running alternately in one of them.  Running several instances simultaneously is NOT recommended!
Sharing it below until a better solution such as hosting is offered, it has been confirmed as OK by the developer.
Take care

**************
Workaround for 2 chrome instances running Butterbot:
To be able to ensure Butterbot is running 24/7 I'll have to run it at work on one pc and then close it at the end of the day, reopening it at home on another pc.
I have butterbot installed on both Chromes (work / home) but configured with different API key/secret sets.
I will be activating/deactivating the API sets manually from the exchange dashboard (possible in Bitstamp, you would have to check the other exchanges) to have either one or the other instance of Butterbot actually actively trading, thus avoiding having them active simultaneously while being able to have both pcs on all the time with their respective Chromes and butterbot instances running.

[San1ty]

Indeed, but don't forget that after the stop-loss is triggered, the bot just continues trading.
So if your EMA Diff is positive it'll just take back the position (and take a small point difference loss + transaction fees)

I like the theory, really I do.

In practice, however, I always seem to lose when I try it.  You are fighting heavy pressure and fast order action, and by the time I am out & back in, I have lost more then I care too.

Price goes on a steep climp from 20 dollars to 250 dollars and then severely corrects to 80 dollars in a couple of minutes.
In this example a trailing stop loss of 20% would have closed my positions at 200 dollar. After that as soon as there is a positive cross-over again the bot can go ahead and buy back in.

Sounds great... but what ends up happening when I tried this is that too often a sharp correction such as we had a few hours ago would trigger the stop-loss in a scenario, like today, that ultimately did not trigger a dead cross.

BTC is still climbing, but I would be out.  Not a win. 

Indeed, but don't forget that after the stop-loss is triggered, the bot just continues trading.
So if your EMA Diff is positive it'll just take back the position (and take a small point difference loss + transaction fees)

Not so small in my experience.  Too much trade traffic and market pressure.  I always lost way more then I cared to, and inevitably, the math showed that I would have been better off going for coffee until the "panic" buying/selling was over, come calmly back in and catch the rebound (providing a trend change is identified).  If on the other hand the spike was not indicative of a trend change, I avoided the trap.

I Found the same with the STB.  I now disable the stoploss in the bot, because it hurts my returns more often then not.

But again, all of that plays into a very methodical approach to trading, and plenty of money has been made with different methods.

I will tell you, however, that how I arrive at my strategy and settings is by spending hundreds of hours with charts testing ideas and theories.  One of the things that is hard to table-test is a stop-loss.  You may have it set for %5, but more often then not it will sell far below that because you are at that point "chasing the trade"; this amplifies with larger amounts of money.

I am also going to come back to a trend that you might be overlooking.  In this volatile market, manipulating fast-acting bots with bull & bear traps is common practice.  One of the things that makes Butter better is it's relative immunity to these market games.

FWIW

I appreciate your point of view, and completely agree that you are chasing the trade and will most likely not be filled (big issue with stop-losses). I'm actually agreeing with you right now, there isn't much use for a stop loss during a big and fast crash like that.

Thanks for sharing your thoughts, nice chat!

[seanrarey]

Very interesting discussion these last few posts, especially now that we are so high that a correction is ever more likely.
Perhaps things are different now vs. April but still my gut feeling is that this just sounds too good to be true (no need waste keyb strokes grilling me for posting a personal view, eh?!).

I'm quite curious and have setup a poll: http://micropoll.com/a/mpview/1146756-3814481
(cannot figure out how to embed nicely in the post...)

I've also been emailing the developer regarding a workaround for those who e.g. have to go to work and would like to have Butter Bot installed in two computers but actively running alternately in one of them.  Running several instances simultaneously is NOT recommended!
Sharing it below until a better solution such as hosting is offered, it has been confirmed as OK by the developer.
Take care

**************
Workaround for 2 chrome instances running Butterbot:
To be able to ensure Butterbot is running 24/7 I'll have to run it at work on one pc and then close it at the end of the day, reopening it at home on another pc.
I have butterbot installed on both Chromes (work / home) but configured with different API key/secret sets.
I will be activating/deactivating the API sets manually from the exchange dashboard (possible in Bitstamp, you would have to check the other exchanges) to have either one or the other instance of Butterbot actually actively trading, thus avoiding having them active simultaneously while being able to have both pcs on all the time with their respective Chromes and butterbot instances running.

Are you doing this so you will have physical control over your bot at all times; i.e. at home & at work?  I have tested multiple bots on the same account, I don't think you have to activate/deactivate your APIs when doing this... just start/stop the different instances of the bot.

I am far from an API or bot expert, but in my experience, Butter does not seem to be a poll-hog, so even if you overlapped, I am thinking it would not really spam the exchange with requests (just thinking aloud here), and I don't see how the bots would ever trade "against each other" with like settings.  Interesting & not a bad solution.

Another, and the one that I use, is Remote Desktop (RDP).  It is super easy & convenient.  As noted prior, I have multiple machines running different bots as well as miners.  They are all set to "listen" on RDP, so I can log into them remotely from any computer, tablet or even my smart phone.  So, like you, I have instances running both at home and office (all on different accounts), but can access any and all of them with a click from a single screen.

[fible1]

Very interesting discussion these last few posts, especially now that we are so high that a correction is ever more likely.
Perhaps things are different now vs. April but still my gut feeling is that this just sounds too good to be true (no need waste keyb strokes grilling me for posting a personal view, eh?!).

I'm quite curious and have setup a poll: http://micropoll.com/a/mpview/1146756-3814481
(cannot figure out how to embed nicely in the post...)

I've also been emailing the developer regarding a workaround for those who e.g. have to go to work and would like to have Butter Bot installed in two computers but actively running alternately in one of them.  Running several instances simultaneously is NOT recommended!
Sharing it below until a better solution such as hosting is offered, it has been confirmed as OK by the developer.
Take care

**************
Workaround for 2 chrome instances running Butterbot:
To be able to ensure Butterbot is running 24/7 I'll have to run it at work on one pc and then close it at the end of the day, reopening it at home on another pc.
I have butterbot installed on both Chromes (work / home) but configured with different API key/secret sets.
I will be activating/deactivating the API sets manually from the exchange dashboard (possible in Bitstamp, you would have to check the other exchanges) to have either one or the other instance of Butterbot actually actively trading, thus avoiding having them active simultaneously while being able to have both pcs on all the time with their respective Chromes and butterbot instances running.

Are you doing this so you will have physical control over your bot at all times; i.e. at home & at work?  I have tested multiple bots on the same account, I don't think you have to activate/deactivate your APIs when doing this... just start/stop the different instances of the bot.

I am far from an API or bot expert, but in my experience, Butter does not seem to be a poll-hog, so even if you overlapped, I am thinking it would not really spam the exchange with requests (just thinking aloud here), and I don't see how the bots would ever trade "against each other" with like settings.  Interesting & not a bad solution.

Another, and the one that I use, is Remote Desktop (RDP).  It is super easy & convenient.  As noted prior, I have multiple machines running different bots as well as miners.  They are all set to "listen" on RDP, so I can log into them remotely from any computer, tablet or even my smart phone.  So, like you, I have instances running both at home and office (all on different accounts), but can access any and all of them with a click from a single screen.

Hey ,
  I can't wait for hosting to go live . Meanwhile, both solutions are correct, the enable and disable is very safe and very efficient to do; this question was geared toward someone who can't have a single computer running all the time.

If you can run a single machine all the time, then you can run multiple instance of the bot on the same machine as long as they use different API keys and different settings .

I hope that helps, please let me know if you guys need anything at all.

Pablo.

[seanrarey]

I appreciate your point of view, and completely agree that you are chasing the trade and will most likely not be filled (big issue with stop-losses). I'm actually agreeing with you right now, there isn't much use for a stop loss during a big and fast crash like that.

Thanks for sharing your thoughts, nice chat!

To continue it... lets make some money.

Earlier this morning we talked a bit about what was working & what was not, but I am interested in using you guys as a sounding board to get deeper into the mechanics of _why_ Bitcoin is a different market then anything I have worked with before, and honing a strategy to maximize profits for our little group here.

As stated before, I do not believe it "trade magic" or secrets, it has all been tried.  Anybody who ever tells you they have a secret system is delusional or trying to sell you something.  Apologies to the very fine folk at Butter, but you don't need their bot to use EMA and profit from it; it is one of the easier systems to use manually.

That said, Butter is the most valuable single tool I now have in my arsenal.  Not because of the trade bot (which I have already stated is the best one I have found), but because of the back-testing engine.

In the past, I would spend weeks with a grease-board charts designing and testing systems, trade points and theories.  I would come up with a better trading strategy in my mind and have to spend dozens of hours proving or disproving it's worth using historical data.  Butter's ability to do this in seconds is like finding the Thor's Hammer of trading.

So, back to the hot topic of the day.  If there are still those out there that are still trying to jump out of the market the instant they see trouble coming, it is because you have not tried it enough times in the real world.  Especially in the Bitcoin market, by the time you see an actual flash-crash (sure, let's call it that), it is too late to sell.  You are chasing the landslide and you _will_ lose a ton of money trying to catch it.

Here is an exercise to sharpen your skills.  With a good charting system, go back one year with GOX trades.  Identify and mark all the flash-crashes.  Now, go back and count off all the crashes that telegraphed a true trend reversal verses those that did not.  67% of the time, a downward trend followed the FC.  That means 43% of the time the market corrected and continued upward.

If you sold during the slide, you lost your shirt 43% of the time, because the market corrected, continued up, and you had to buy back in higher then you sold; often a lot higher.  This because you sold while chasing the market down, and then had to buy while chasing the market back up.

Now, back at your chart, let's look at the 67% of the time the crash did actually telegraph a trend reversal and selling was appropriate.  Take the lowest value 7 minutes after the first long red candle.  In my experience, this is where a quick acting trader or bot can "catch up" to a fast falling market.  Remember, just because set a bot to dump @ 5% does not mean that is where it will trade.  You (or your bot) have to find buyers in a panicking market.  This is always going to be well below where you would like it to be.  Mark this number.

Now, move forward on the chart 45 minutes to an hour.  See the bounce back?  There are very few things that "always" happen in trading, but this is one.  There is always a correction after unusual pressure.  I tell students in to think of these extreme downward movements as having a rubber band attached.  Because this pressure is finite, there is always a snap-back before the trend continues.  _This_ is where you want to sell, _if_ you have identified a downward trend.  Now, instead of chasing a market that is running away from you, you are selling into a market that is actually pushing up into your sale.  Timed right, you will get a much better price and a quick sale.

The larger the amounts you are trading, the more important this technique becomes.  Good luck selling >$20,000 worth of points to a fleeing market at a price you can live with.  You want that market moving hard into your sale to efficiently unload such a position.

If you are accomplishing these trades manually, timing and patience is the key.  It is the hardest thing in the world to watch your stake fall through the floor, and resist selling.  That is why everyone else is chasing the trade down.  But you did your homework.  You know that _every_ true panic-sell results in a hard, if brief, correction.  _Wait   for    it..._ lol.

If you are trying to program a bot to catch the snap-back, we get to hone down the requirements to just timing (because computer can be programmed to be patient).  We can do some very simple math and find the average timing of all the snap-backs in the last year.  Remarkably, I am finding this timing to be very consistent.  I have found that selling 1 - 2 hours _after_ the crash is identified virtually always results in getting a better price in the sale then if I had tried chasing the slide.

Finally, there is the element of bots.  There are a _lot_ of bots in this market.  It is estimated that because of the techie nature of this market, the bots actually number in the tens of thousands.  Scalping bots, arbitrage bots, RSI bots and limit bots are all reacting to increasingly slighter twitches of the market.  Bots are trying to fool other bots into reacting in order to profit off of the swings it causes.  Yes, I am playing this game with them (mostly with STB right now), but although it is fun, there is much less actual profit in it when compared to good, old-fashioned solid trading.

So, full circle as to what we really want to see added to this bot.

Stop-loss?  For me, no. it defies the trading strategy that by nature is the best stop-loss already; if set right.

Profit checking?  Some have asked that the bot make sure it is not making a losing trade.  Again,this defies the principals of this type of trading.  This system will make trades for loses specifically because it then always has you in the most favorable market position.  If it failed to get me into the market because the trade would have resulted in a small loss, and the market then gains 100 points, I would be very unhappy indeed.  In fact, the most recent huge gain was immediately predicated by a trade resulting in a small loss... to get me back into the market.  The "profit checking" suggested would have negated this gain.  No place for that in this type of trading.

Magic wand to tell us exactly when we are at the top or bottom of a trend?  Yes please.  Lol.  But since there is no magic in trading, I am very satisfied with a good, solid machine that based on good numbers and research returned a damn-near 50% profit in the last 30 days.

[MicroFi]

I appreciate your point of view, and completely agree that you are chasing the trade and will most likely not be filled (big issue with stop-losses). I'm actually agreeing with you right now, there isn't much use for a stop loss during a big and fast crash like that.

Thanks for sharing your thoughts, nice chat!

...

Magic wand to tell us exactly when we are at the top or bottom of a trend?  Yes please.  Lol.  But since there is no magic in trading, I am very satisfied with a good, solid machine that based on good numbers and research returned a damn-near 50% profit in the last 30 days.

Thank you for your insights, seanrarey. When you're right, you're right.
Maybe I just need to find my 'zen' and start relying on proven trading algorithms ...

Your posts are extremely interesting to read and I'm sure there's a lot more to learn from you.
If you ever happen to be in Brussels one day, please let me know so I can pick your brain over a couple of beers

[seanrarey]

Ah, "Zen, and the Art of Trading"; good book, a recommended read.



Don't be surprised if I take you up on that beer.

[NginUS]

So, full circle as to what we really want to see added to this bot.

Stop-loss?  For me, no. it defies the trading strategy that by nature is the best stop-loss already; if set right.

Profit checking?  Some have asked that the bot make sure it is not making a losing trade.  Again,this defies the principals of this type of trading.  This system will make trades for loses specifically because it then always has you in the most favorable market position.  If it failed to get me into the market because the trade would have resulted in a small loss, and the market then gains 100 points, I would be very unhappy indeed.  In fact, the most recent huge gain was immediately predicated by a trade resulting in a small loss... to get me back into the market.  The "profit checking" suggested would have negated this gain.  No place for that in this type of trading.

Magic wand to tell us exactly when we are at the top or bottom of a trend?  Yes please.  Lol.  But since there is no magic in trading, I am very satisfied with a good, solid machine that based on good numbers and research returned a damn-near 50% profit in the last 30 days.

Hey Pablo,

So with this being said, after the explanation which preceded it, then I want what he wants. Which would be none of the additions we were talking about recently except maybe the new trading engine improvements should they pan out. Plus profit/loss tracking, please.

[MicroFi]

Ah, "Zen, and the Art of Trading"; good book, a recommended read.



Don't be surprised if I take you up on that beer.

Please do, that's exactly why I made my offer

[ghostshirt]

Best thing with butter-bot, I guess, it comes with insights from experienced traders. I love it and learn much from you. Thank you all.

[fible1]

So, full circle as to what we really want to see added to this bot.

Stop-loss?  For me, no. it defies the trading strategy that by nature is the best stop-loss already; if set right.

Profit checking?  Some have asked that the bot make sure it is not making a losing trade.  Again,this defies the principals of this type of trading.  This system will make trades for loses specifically because it then always has you in the most favorable market position.  If it failed to get me into the market because the trade would have resulted in a small loss, and the market then gains 100 points, I would be very unhappy indeed.  In fact, the most recent huge gain was immediately predicated by a trade resulting in a small loss... to get me back into the market.  The "profit checking" suggested would have negated this gain.  No place for that in this type of trading.

Magic wand to tell us exactly when we are at the top or bottom of a trend?  Yes please.  Lol.  But since there is no magic in trading, I am very satisfied with a good, solid machine that based on good numbers and research returned a damn-near 50% profit in the last 30 days.

Hey Pablo,

So with this being said, after the explanation which preceded it, then I want what he wants. Which would be none of the additions we were talking about recently except maybe the new trading engine improvements should they pan out. Plus profit/loss tracking, please.

Sounds Fair ,
    We will still improve stop loss but the new trading algorithm is taking precedence. We'll see .

On an unrelated matter, we may have some other improvements you guys have been wanting out pretty soon .

Stay tuned.

Pablo.