Author

Topic: margin terminal - over 25 spot and futures exchanges! Bots and more - page 229. (Read 268553 times)

hero member
Activity: 770
Merit: 506
Is there extra fees for the exchanges or are the included?

and is there a possibility of adding Yobit and Cryptopia in the future? Your bot is a decent price for newbies who tend to trade on exchanges I mention.

Edit
Disregard. I see one is free and charges for extra additions
member
Activity: 64
Merit: 10
Hello,

can u pls calrify me somethings?

For the "Static Ping Pong" strategy, what's the stop loss policy for it (if there's one)? I mean if I'm having loss with a buy order how can I stop this loss and close my position (sell) automatically? Is there anywhere I can set this?

The same qustion applies to "Margin Maker". I'm using the demo version, and I was in loss, but no stop loss was applied to my buy order. How can I fix this? Also can u pls explain me better how this "Margin Maker" strategy works? what's buy/sell margin, and "min efferctive gain" option? What's the stop loss policy for it? what's take profit policy? I could see it adapts the "limit sell" order automatically sometimes. so how does it really work?

Thnx.

Hi leomedina01,

Thanks for checking out leonArdo.

There is currently no stop loss functionality in the bot - we are adding it soon.

We do support stop loss and trailing stop on the bitfinex exchange as they provide for it in their API. But we realise we cannot rely on all exchanges to add this into their API and there will have to be a leonArdo specific solution.

The main difference between the static ping pong and mArgin maker bots is that the ping pong bot is static and the mArgin maker in dynamic. What that means is that with the ping pong bot you explicitly set a buy and sell price and that is all the bot does. If the course bounces between those pre-determined prices the bot will cycle through buy and sell orders. In contrast to this you don't set a buy and sell price for the mArgin maker bot, but rather you set a relative margin which adapts based on the history window - which continually adapts to what has happened. This means that it allows for prices to be dynamically chosen. There are other parameters you can set to ensure the bot behaves like you want it to such as: don’t sell below a certain price or ensure a min gain has been achieved.

The minimum effective gain allows you set the amount you want to make on each trade cycle. This is the amount after exchange fees have been subtracted. Setting a negative value allows the bot to make a loss in the hope that it can continue trading and make a gain. Otherwise, the bot would stop until the course came back to it.

The take profit mechanism has nothing to do with the mArgin maker bot, but rather it is a leonArdo margin trading feature. I discuss it in the margin trading video you can find here: http://www.marginsoftware.de/tutorial.html

Best wishes,
Jonathan


Ok. Thx for the prompt reply.

What I called of "take profit" is in fact the "limit sell" order the bot creates after a buy order is executed. I saw this "limit sell" order was not executed sometimes although the target price has been hit.

Like here: Why it didn't sell here > http://prntscr.com/ehfzo6 ?

You can see there's a "limit sell" order to 0.00009823, but price hit 0.00009994, but sell order was not executed

Another thing I noticed is for example selected timeframe is 3D, but candle displayed is for 1H timeframe > http://prntscr.com/ehglr6. Pls let me know if this is the correct behaviour.

Thanks.

Hi leomedina01,

The demo market tries to emulate what would happen on the real market, but seems like here it misfired? Without the logs I couldn't tell you why, but my guess would be that you had a network out issue at the point the course went through your sell order and that's why it didn't execute.

If you were using the real version and that was a real limit order on the exchange it would have been filled.

As for the candles, I am not sure I understand. The buttons at the top of the chart show the time frame and this then determines the time frame of the individual candles. For example in the 3 month view the candles are 1 day, in the 1D view the candles are 30 min and in the 30 minute view the candles are 1 minute candles.

All the best,
Jonathan


Thnx for all prompt replies Jonathan.

I'm about to purchase Leonardo, I'm only concerned about the fact it does not have a way to set a stop-loss for poloniex orders. Do you think this is something I really need to be concerned about? Do you think it's safe trading without a stop-loss?

I mean suppose I cannot be in front of my pc all the time looking in the chart. Do you think it's safe being out for some hours (ex. for sleep) and leave the bot go alone, without any stop-loss? do you think it's safe? Just need an honest answer.

Regards.
legendary
Activity: 1988
Merit: 1008
Hello,

can u pls calrify me somethings?

For the "Static Ping Pong" strategy, what's the stop loss policy for it (if there's one)? I mean if I'm having loss with a buy order how can I stop this loss and close my position (sell) automatically? Is there anywhere I can set this?

The same qustion applies to "Margin Maker". I'm using the demo version, and I was in loss, but no stop loss was applied to my buy order. How can I fix this? Also can u pls explain me better how this "Margin Maker" strategy works? what's buy/sell margin, and "min efferctive gain" option? What's the stop loss policy for it? what's take profit policy? I could see it adapts the "limit sell" order automatically sometimes. so how does it really work?

Thnx.

Hi leomedina01,

Thanks for checking out leonArdo.

There is currently no stop loss functionality in the bot - we are adding it soon.

We do support stop loss and trailing stop on the bitfinex exchange as they provide for it in their API. But we realise we cannot rely on all exchanges to add this into their API and there will have to be a leonArdo specific solution.

The main difference between the static ping pong and mArgin maker bots is that the ping pong bot is static and the mArgin maker in dynamic. What that means is that with the ping pong bot you explicitly set a buy and sell price and that is all the bot does. If the course bounces between those pre-determined prices the bot will cycle through buy and sell orders. In contrast to this you don't set a buy and sell price for the mArgin maker bot, but rather you set a relative margin which adapts based on the history window - which continually adapts to what has happened. This means that it allows for prices to be dynamically chosen. There are other parameters you can set to ensure the bot behaves like you want it to such as: don’t sell below a certain price or ensure a min gain has been achieved.

The minimum effective gain allows you set the amount you want to make on each trade cycle. This is the amount after exchange fees have been subtracted. Setting a negative value allows the bot to make a loss in the hope that it can continue trading and make a gain. Otherwise, the bot would stop until the course came back to it.

The take profit mechanism has nothing to do with the mArgin maker bot, but rather it is a leonArdo margin trading feature. I discuss it in the margin trading video you can find here: http://www.marginsoftware.de/tutorial.html

Best wishes,
Jonathan


Ok. Thx for the prompt reply.

What I called of "take profit" is in fact the "limit sell" order the bot creates after a buy order is executed. I saw this "limit sell" order was not executed sometimes although the target price has been hit.

Like here: Why it didn't sell here > http://prntscr.com/ehfzo6 ?

You can see there's a "limit sell" order to 0.00009823, but price hit 0.00009994, but sell order was not executed

Another thing I noticed is for example selected timeframe is 3D, but candle displayed is for 1H timeframe > http://prntscr.com/ehglr6. Pls let me know if this is the correct behaviour.

Thanks.

Hi leomedina01,

The demo market tries to emulate what would happen on the real market, but seems like here it misfired? Without the logs I couldn't tell you why, but my guess would be that you had a network out issue at the point the course went through your sell order and that's why it didn't execute.

If you were using the real version and that was a real limit order on the exchange it would have been filled.

As for the candles, I am not sure I understand. The buttons at the top of the chart show the time frame and this then determines the time frame of the individual candles. For example in the 3 month view the candles are 1 day, in the 1D view the candles are 30 min and in the 30 minute view the candles are 1 minute candles.

All the best,
Jonathan



member
Activity: 64
Merit: 10
Hello,

can u pls calrify me somethings?

For the "Static Ping Pong" strategy, what's the stop loss policy for it (if there's one)? I mean if I'm having loss with a buy order how can I stop this loss and close my position (sell) automatically? Is there anywhere I can set this?

The same qustion applies to "Margin Maker". I'm using the demo version, and I was in loss, but no stop loss was applied to my buy order. How can I fix this? Also can u pls explain me better how this "Margin Maker" strategy works? what's buy/sell margin, and "min efferctive gain" option? What's the stop loss policy for it? what's take profit policy? I could see it adapts the "limit sell" order automatically sometimes. so how does it really work?

Thnx.

Hi leomedina01,

Thanks for checking out leonArdo.

There is currently no stop loss functionality in the bot - we are adding it soon.

We do support stop loss and trailing stop on the bitfinex exchange as they provide for it in their API. But we realise we cannot rely on all exchanges to add this into their API and there will have to be a leonArdo specific solution.

The main difference between the static ping pong and mArgin maker bots is that the ping pong bot is static and the mArgin maker in dynamic. What that means is that with the ping pong bot you explicitly set a buy and sell price and that is all the bot does. If the course bounces between those pre-determined prices the bot will cycle through buy and sell orders. In contrast to this you don't set a buy and sell price for the mArgin maker bot, but rather you set a relative margin which adapts based on the history window - which continually adapts to what has happened. This means that it allows for prices to be dynamically chosen. There are other parameters you can set to ensure the bot behaves like you want it to such as: don’t sell below a certain price or ensure a min gain has been achieved.

The minimum effective gain allows you set the amount you want to make on each trade cycle. This is the amount after exchange fees have been subtracted. Setting a negative value allows the bot to make a loss in the hope that it can continue trading and make a gain. Otherwise, the bot would stop until the course came back to it.

The take profit mechanism has nothing to do with the mArgin maker bot, but rather it is a leonArdo margin trading feature. I discuss it in the margin trading video you can find here: http://www.marginsoftware.de/tutorial.html

Best wishes,
Jonathan


Ok. Thx for the prompt reply.

What I called of "take profit" is in fact the "limit sell" order the bot creates after a buy order is executed. I saw this "limit sell" order was not executed sometimes although the target price has been hit.

Like here: Why it didn't sell here > http://prntscr.com/ehfzo6 ?

You can see there's a "limit sell" order to 0.00009823, but price hit 0.00009994, but sell order was not executed

Another thing I noticed is for example selected timeframe is 3D, but candle displayed is for 1H timeframe > http://prntscr.com/ehglr6. Pls let me know if this is the correct behaviour.

Thanks.
legendary
Activity: 1988
Merit: 1008
Hello,

can u pls calrify me somethings?

For the "Static Ping Pong" strategy, what's the stop loss policy for it (if there's one)? I mean if I'm having loss with a buy order how can I stop this loss and close my position (sell) automatically? Is there anywhere I can set this?

The same qustion applies to "Margin Maker". I'm using the demo version, and I was in loss, but no stop loss was applied to my buy order. How can I fix this? Also can u pls explain me better how this "Margin Maker" strategy works? what's buy/sell margin, and "min efferctive gain" option? What's the stop loss policy for it? what's take profit policy? I could see it adapts the "limit sell" order automatically sometimes. so how does it really work?

Thnx.

Hi leomedina01,

Thanks for checking out leonArdo.

There is currently no stop loss functionality in the bot - we are adding it soon.

We do support stop loss and trailing stop on the bitfinex exchange as they provide for it in their API. But we realise we cannot rely on all exchanges to add this into their API and there will have to be a leonArdo specific solution.

The main difference between the static ping pong and mArgin maker bots is that the ping pong bot is static and the mArgin maker in dynamic. What that means is that with the ping pong bot you explicitly set a buy and sell price and that is all the bot does. If the course bounces between those pre-determined prices the bot will cycle through buy and sell orders. In contrast to this you don't set a buy and sell price for the mArgin maker bot, but rather you set a relative margin which adapts based on the history window - which continually adapts to what has happened. This means that it allows for prices to be dynamically chosen. There are other parameters you can set to ensure the bot behaves like you want it to such as: don’t sell below a certain price or ensure a min gain has been achieved.

The minimum effective gain allows you set the amount you want to make on each trade cycle. This is the amount after exchange fees have been subtracted. Setting a negative value allows the bot to make a loss in the hope that it can continue trading and make a gain. Otherwise, the bot would stop until the course came back to it.

The take profit mechanism has nothing to do with the mArgin maker bot, but rather it is a leonArdo margin trading feature. I discuss it in the margin trading video you can find here: http://www.marginsoftware.de/tutorial.html

Best wishes,
Jonathan




member
Activity: 64
Merit: 10
Hello,

can u pls calrify me somethings?

For the "Static Ping Pong" strategy, what's the stop loss policy for it (if there's one)? I mean if I'm having loss with a buy order how can I stop this loss and close my position (sell) automatically? Is there anywhere I can set this?

The same qustion applies to "Margin Maker". I'm using the demo version, and I was in loss, but no stop loss was applied to my buy order. How can I fix this? Also can u pls explain me better how this "Margin Maker" strategy works? what's buy/sell margin, and "min efferctive gain" option? What's the stop loss policy for it? what's take profit policy? I could see it adapts the "limit sell" order automatically sometimes. so how does it really work?

Thnx.
legendary
Activity: 1988
Merit: 1008
Id like to know if it works on a ubuntu 16.10 desktop version?

Yes, it works on Ubuntu!!
legendary
Activity: 1694
Merit: 1002
Go Big or Go Home.....
No ill intended, so no worries, just trying to help. I'm in IT/tech so have literally any and everything to play with and know the hardware in and out.

If your Skylake or Haswell i5's are showing extra high cpu usage, it is a system issue, or you are running too many bots on the exchange.
(I ran it briefly on my new laptop (Kaby Lake i5-7200u) and it ran nice and fast with 30% CPU usage under leonardo using 17 bots), so even with a
low-voltage dual core it runs fine.)

CPU scaling to new tech should not matter at all if it runs as well as it does on 10 year old tech. I tested it on some spare junk I had lying around (Ie Intel C2Duo and C2Quad systems) and found no need to dedicate a higher-end system for it. I might try it if I feel bored. I'll probably buy a second license for Leonardo with it as it has paid for itself after a few day's of trading (and it's been almost 4 now).  Wink
hero member
Activity: 868
Merit: 517
I am presently only paper trading on two machines.  One machine is a laptop and the other is a PC.  The load is pretty high on the laptop so I don't think I will keep running it there.  I have seen resource contention and some issues where the bots all seem to busy at the same time.  At that point CPU utilization has been high.  I setup a second machine where it does some gpu mining in the background and Leonardo with a bunch of bots, and again I do see a fairly high CPU use as it cycles the refresh on the bots.  The more bots the worse it gets.  I used to CPU mine on this machine also, but that caused Leonardo to crash if I tried to run Leonardo and the cpuminer at the same time.  The cpuminer is set to only use half the cores but apparently this is enough to interfere with Leonardo and cause some crashes.  It is more stable now that I stopped cpu mining that machine, but still has crashed a couple of times.   

So I do think this could benefit if it has the ability to scale to more cores and threads.  Since we now hear rumblings from Intel about more cores and threads coming it can't hurt to start looking at leveraging this new hardware.  And one observation is that the coins still show the open buy or sell order when bringing Leonardo back up after a restart/crash, but the bot it no longer running for any of those positions.  I don't know if that is just a quirk of paper trading with this miner or if that is what is expected.  YMMV

 



Sounds like you are running it on a PC with a low-end CPU. It should be a quad core or at least an i5 with HT.
Like I said, on an old Quad-COre Q6600 it runs with 10-20 bots under 25% CPU utilization, and it only has 4GB of ram.

If you're concerned with CPU usage, why are you even running it on a Rig that you mine on??!

Haswell and Skylake i5's in both machines.  Everything around here has multiple jobs.  The Haswell is the laptop, but that really is only a 2 core cpu.  So that isn't cutting it.  So 4 cores really is the minimum. 

Please don't mistake what I am saying for complaining.  I simply asked if the new CPU's are going to be supported. This is not a critique of Leonardo.  I can see that if I were to use this and begin trading with my own money I will use a dedicated machine that does nothing but trading.   

As it stands right now I am still in the stage of trying to figure out if I can make money with it.   So I was not going to dedicate an entire machine at this point until I can figure out how to make this work for me.  I am building some new machines and I will dedicate one of those for trading.  Hence the question about supporting the new CPU's.   
legendary
Activity: 1988
Merit: 1008
Just a general question,
I am having a few issues on Bitfinex, I can only assume the bot is trying to be too precise, is there a way to buy and sell just down to 1 decimal?  Eg. $1250.5
bot keeps stopping and this is the error that comes up in real time on the exchange " Invalid order price precision (5 sig figures limit)."


Price Precision
The precision level of all trading prices is calculated based on significant figures.

For all pairs on Bitfinex we use 5 significant digits. Examples of five significant digits would be 12.123, 1.1234, 123.43, and 1234.5.

This mimics how traditional global markets dynamically handle precision of small, medium, and larger values. The rationale behind this is that the higher the amount the less relevant the number of decimals becomes. The corollary is true for very small amounts, where more precision is more useful.

Note: API will truncate price with precision > 5

Hi Dan,

We saw that there was an issue there in their API as far as I remember. I'll look into it and get back to you soon.

Best wishes,
Jonathan

sr. member
Activity: 432
Merit: 250
Just a general question,
I am having a few issues on Bitfinex, I can only assume the bot is trying to be too precise, is there a way to buy and sell just down to 1 decimal?  Eg. $1250.5
bot keeps stopping and this is the error that comes up in real time on the exchange " Invalid order price precision (5 sig figures limit)."


Price Precision
The precision level of all trading prices is calculated based on significant figures.

For all pairs on Bitfinex we use 5 significant digits. Examples of five significant digits would be 12.123, 1.1234, 123.43, and 1234.5.

This mimics how traditional global markets dynamically handle precision of small, medium, and larger values. The rationale behind this is that the higher the amount the less relevant the number of decimals becomes. The corollary is true for very small amounts, where more precision is more useful.

Note: API will truncate price with precision > 5
legendary
Activity: 1694
Merit: 1002
Go Big or Go Home.....
I am presently only paper trading on two machines.  One machine is a laptop and the other is a PC.  The load is pretty high on the laptop so I don't think I will keep running it there.  I have seen resource contention and some issues where the bots all seem to busy at the same time.  At that point CPU utilization has been high.  I setup a second machine where it does some gpu mining in the background and Leonardo with a bunch of bots, and again I do see a fairly high CPU use as it cycles the refresh on the bots.  The more bots the worse it gets.  I used to CPU mine on this machine also, but that caused Leonardo to crash if I tried to run Leonardo and the cpuminer at the same time.  The cpuminer is set to only use half the cores but apparently this is enough to interfere with Leonardo and cause some crashes.  It is more stable now that I stopped cpu mining that machine, but still has crashed a couple of times.   

So I do think this could benefit if it has the ability to scale to more cores and threads.  Since we now hear rumblings from Intel about more cores and threads coming it can't hurt to start looking at leveraging this new hardware.  And one observation is that the coins still show the open buy or sell order when bringing Leonardo back up after a restart/crash, but the bot it no longer running for any of those positions.  I don't know if that is just a quirk of paper trading with this miner or if that is what is expected.  YMMV

 



Sounds like you are running it on a PC with a low-end CPU. It should be a quad core or at least an i5 with HT.
Like I said, on an old Quad-COre Q6600 it runs with 10-20 bots under 25% CPU utilization, and it only has 4GB of ram.

If you're concerned with CPU usage, why are you even running it on a Rig that you mine on??!
hero member
Activity: 868
Merit: 517
I am presently only paper trading on two machines.  One machine is a laptop and the other is a PC.  The load is pretty high on the laptop so I don't think I will keep running it there.  I have seen resource contention and some issues where the bots all seem to busy at the same time.  At that point CPU utilization has been high.  I setup a second machine where it does some gpu mining in the background and Leonardo with a bunch of bots, and again I do see a fairly high CPU use as it cycles the refresh on the bots.  The more bots the worse it gets.  I used to CPU mine on this machine also, but that caused Leonardo to crash if I tried to run Leonardo and the cpuminer at the same time.  The cpuminer is set to only use half the cores but apparently this is enough to interfere with Leonardo and cause some crashes.  It is more stable now that I stopped cpu mining that machine, but still has crashed a couple of times.   

So I do think this could benefit if it has the ability to scale to more cores and threads.  Since we now hear rumblings from Intel about more cores and threads coming it can't hurt to start looking at leveraging this new hardware.  And one observation is that the coins still show the open buy or sell order when bringing Leonardo back up after a restart/crash, but the bot it no longer running for any of those positions.  I don't know if that is just a quirk of paper trading with this miner or if that is what is expected.  YMMV

 

legendary
Activity: 1694
Merit: 1002
Go Big or Go Home.....
With the new release of AMD's Ryzen CPU's with more cores and threads is there any plan to compile and support the enhanced capabilities of these new chips?  Seems like it would be great fit for Leonardo.  It looks like Leonardo could benefit from the additional horsepower.   Grin

I'm using Leonardo on an older Intel Q6600 Quad Core system with only 4GB of ram and an SSD running Windows 8.1 Pro.
I don't do much else with it other than remote access into it and have it monitor some of my mining rigs as well. Overall it has been very stable and CPU / Resource usage is minimal. I use between 10-20 bots on it with no slow downs and CPU usage under 25%.

Leonardo has crashed 3 times over the past month or so of running it but at least 2x it was my fault/cause of the crash.

I don't think there is a need for high - spec CPU's unless you are planning to run 100 bots under one Leonardo.
In that case I'd suggest splitting up over multiple instances and PC's of Leonardo and staying under 20bots each. Less chance of down-time in case of a crash.
hero member
Activity: 868
Merit: 517
With the new release of AMD's Ryzen CPU's with more cores and threads is there any plan to compile and support the enhanced capabilities of these new chips?  Seems like it would be great fit for Leonardo.  It looks like Leonardo could benefit from the additional horsepower.   Grin
lfo
member
Activity: 85
Merit: 10
BNM founder
Id like to know if it works on a ubuntu 16.10 desktop version?
legendary
Activity: 1694
Merit: 1002
Go Big or Go Home.....
Question. When I see lots of little volume trades under my trade history (larger orders are set as ping pong or market maker though).
Do those count as individual sales/buys? If they are very low amounts, would they count as a loss after fees?

No need to worry!

The bots always try to sell the given amount. The execution of these orders can, however, be split over several trades into so called partial order fills. leonArdo visualises the remaining amount left of the order. As the percentages are just calculated on the individual amounts of each trade, it does not matter whether they are split or executed as a single order.

Thank you as always.
legendary
Activity: 1988
Merit: 1008
Due to the recent Cloudflare bug, which caused a memory leak, some bitcoin exchanges decided to be cautious and delete all API keys.

We have already been contacted by many of our customers who connect to BTC-e and have been affected.  We have no problem providing anyone affected with a new copy of leonArdo with an updated API key.

Just contact us Smiley

All the best,
The leonArdo guys
legendary
Activity: 1988
Merit: 1008
Question. When I see lots of little volume trades under my trade history (larger orders are set as ping pong or market maker though).
Do those count as individual sales/buys? If they are very low amounts, would they count as a loss after fees?

No need to worry!

The bots always try to sell the given amount. The execution of these orders can, however, be split over several trades into so called partial order fills. leonArdo visualises the remaining amount left of the order. As the percentages are just calculated on the individual amounts of each trade, it does not matter whether they are split or executed as a single order.
legendary
Activity: 1694
Merit: 1002
Go Big or Go Home.....
Question. When I see lots of little volume trades under my trade history (larger orders are set as ping pong or market maker though).
Do those count as individual sales/buys? If they are very low amounts, would they count as a loss after fees?
Jump to: