Topic: margin terminal - over 25 spot and futures exchanges! Bots and more - page 156. (Read 268864 times)

No,
Take a look at this screenshot.

https://imgur.com/Fk5YIwI

The eff margin gain will only kick in when the top horizontal red dotted line (the bottom of the buy margin) is lower than the last buy amount.

Another question on negative, when would the bot cancel positive sell order and switch to negative one if I set min.eff to be negative? After price drop below my buying price?

Hey leonArdo team, is it possible to add the option to allow us to include USD from BTC/USDT chart into the Y Axis notation beside the current BTC/ETH pricing? For example instead of just listing 0.0013, we can set it to show 0.0013($5).

There is no guaranteed way to detect trend.  As far as the bot goes I would like a feature added such as I described earlier. Essentially the bot would record x number of the last completed orders, lets say 300 for example, the number is set by the user. Then the user defines a trend number. Lets say 50 for example. So in this example the bot would compare the average price of the last 50 orders versus the average price of the last 300. If the aversge price of the last 50 is higher than the 300 then the price is trending up, and vice versa. Think of it as comparing two short moving averages to determine immediate trend. Its a simple, straightforward way to help the bot buy/sell at more profitable times.

As far as manual trading goes, look for confluence. Such as a MACD cross combined with a RSI rising above 30. You want a buy signal confirmed at least two indicators combined with a review of price action. Tons of great vids on YouTube about trading strategies and confluence.

Defvader do you know of a simple method to detect the trend of a coin?
I am trying to learn to read MACD but everytime I think I got it right it doesnt go the way I thought it would.

I agree that a trend watcher is a must have feature in the bot. I am seeing alot of good profit oportunities being lost because the bot did not adapt to the trend of the coin.
As a couch trader I really hope those features come soon enough to be able to take advantage of the market right now.
I bought Leonardo because I saw potential in it and the talk about new features really has me excited about it.

I'm not sure if he's going to add it just yet. But basically its a simple way for us to set parameters so that the bot won't buy a crashing coin or sell a rising coin too early. He has been receptive to the idea and my hope is that it gets included at some point. A trend watcher is critical to making automated trading work.

Having said that, Jonathan has told me about some of the new features in the pipeline and I think the community is going to be REALLY excited. If the dev team succeeds with what they're talking about then leonArdo is going to a major player in the crypto world.

Can you let us know what the feature would do?
I am anxiously waiting to see what new versions will bring.
Leonardo is a great piece of software and I am sure it will improve even more.
Its always nice to have access to the developers and that they actually reply back.

For me theres two things that I want to see Leonardo do, first improve its data collections so that it reduces its bandwidth.
Because I have a monthly limit I can only use a few pairs at a time.
And second, I hope to see the bot being able to ride the trend swings to maximize profits from the buy/sell.

I've spoken to Jonathan a few times over the last week about implementing a feature that could help with situations such as this. He has been gracious with his time and we should be thankful for such a responsive dev.  Let's hope the new leonArdo lives up to its name.

heres a perfect example of what I am saying:


Buy at 3:20pm, coin dropps and it sells at 4:32 for a loss of 1.66%

If it made a buy at that price right away it would have recoved the loss because look how much the coin went up.
now I am stuck with a low buy value and its going to be an hour before the statistics window buy margin moves up.

Jonathan, have you equationed changing how the bot places a buy order after a sell is made because of the negative eff gain%?

Right now the bot will calculate a new buy order based on the buy margin % as it normally does but from my relative short experience playing with it, perhaps it would be more profitable to have the bot place a buy order at whatever the price is right away.
What I am seeing is that when the sell order is executed the price is swinging up and it misses on a possible profitable situation.
What I am suggesting is when a sell order is completed because of the eff min gain % to place a buy order immediately and after the buy order is filled to set up a sell order based on the first buy order that generated the loss plus the eff min gain % to try to recover the amount of the loss.
example buy at 1btc, the price drops and when the sell takes place at 0.9btc (-1%) make a buy at whatever the price is (hopefully at 0.9btc) and then create a sell at (1btc + 0.1%+fees) 1.1btc and then the bot resumes from it.

I have to say that I have really thought too much if this is a good strategy or not but I have been using startegies with -1 and -2% eff gains % and everytime the sell happens the price is going up. This causes the bot to place a buy order that is way too low and it might take alot of time for it to catch up with the coin if the price stays stable or keeps going up.

>> For positive what if min.eff gain is larger than eff.margin? would it wait until min.eff gain meet or sell at eff.margin

The mArgin maker bot sets the price of the next buy or sell dependent on the current statistics - which change over time as new trades enter the trade window. That's the first thing. However, before the bot places an order it checks the Min effective gain parameter. Does placing the order violate the Min effective gain? In your case above the answer is yes. If you set a Min effective gain of 10% the bot has to place the order at a price that delivers a 10% gain, after trading fees, from the previous trade that went through.

>> For negative I still don't get it. If I buy at 1btc eff.margin is 5% and I set min.eff to -1 when will this take effect? when price drop to 0.95btc? How thats different from stop-loss %

Allowing the bot to make a loss, by setting a negative Min effective gain, means that if the bot buys but after that the price falls the bot can potentially continue to trade. To keep things simple, you set a Min effective gain of -1% and the bot buys at a price of 1 BTC. If the price drops below 1 BTC and never recovers, then without a negative Min effective gain it would never place another trade. With it turned on the bot has a chance to place a sell order as low as 0.99 BTC, which might mean it can make a number of trades thereafter which make gains.

Of course there are no guarantees.

It's different from stop-loss because stop-loss is an event that also stops the bot. That's your get out of the market because you are not prepared to make a single cent more of a loss than where you place your stop loss.

Thanks for your compliments.

30 bots is quite a lot. I'd try reducing the number to 25 and see how you go. Your system is fine.

The next release will include stability fixes. The one after the next one will be focused on new bots.

Best wishes,
Jonathan

For positive what if min.eff gain is larger than eff.margin? would it wait until min.eff gain meet or sell at eff.margin

For negative I still don't get it. If I buy at 1btc eff.margin is 5% and I set min.eff to -1 when will this take effect? when price drop to 0.95btc? How thats different from stop-loss %

If you set it a 0 it does nothing.
If you set it at a posive value it will cause a sell as soon the coin price increase by that amount in relation to the buy. Ie, buy at 1.00 and eff gain 5%, then if price reaches 5.00 it sells regardless of the eff margin being 10%.

If you set with a negative value it behaves completely differently from the positive values.

First it the sell margin % has to fall below the buy price and then the bot starts creating sells orders at a loss up to the eff gain %. ie, you buy a coin at 1.00 and the sell margin was a nice 5% (so eff margin is 5%).
The coin value would have to drop those 5% or more for the sell margin to be below 1.00 for the eff gain % to kick in.

Can someone explain about min.effective gain? I don't understand what it does.

When I create a bot I set buying and selling windows and Eff.margin is calculated based on that window let say 1%
So if I put 1btc worth on that bot it will buy when price reach bottom blue bar and sell when it reach top and I will gain 1% of 1 BTC which is 0.01 BTC

What does min. eff. gain do in this case if I set it to -1, 0, and 1 % ?

Hello and thank you for this great software.
I'm really looking forward for the next update and its developed trend riding skills and the stop buttom after the last buy.
However what i experienced in the past days and weeks is that Leonardo would stop working at a certain point (usually after 2-3 days) and crashes.

My System is:
Intel Core M-5Y10c, 8gb ddr 3, 64 bit Windows 10, Internet: 25 Mbit

Leonardo:
Bittrex, 30 pairs, 1 bot per pair, (MB:=450, 1/s= 10.4, kb/s= 320)

My question is if it would be better to reduce the pairs for the sake of stability? 
Also is this next update focusing on new features alone or is overall stability included?

Thank you for your hard work and awsome support!

Thank you for your response!

as far as ı experienced, bittrex is flawless, poloniex is better now compared to the ddosed days.
ı am using both of them without problem

There are no perfect settings for the bot. However, if you set the Min effective gain to be a positive, there is no way the bot can a loss. There are a number of important options to consider like how long your order stays active for (max order age), whether you want the protection of stop loss, whether you want to allow the bot make a loss in the hope that future gains will be made (Min effective gain), the initial history trade window size, the initial effective gain amount, the relative (within the trade history window) buy and sell margin percentages.

Do you want a fast acting bot making lots of trades and smaller gains or a slower mover making big gains? You can set mulitple bots, some high frequency and others low frequency, on a single pair.

And if there is a pump and dump in action it might be wise to switch to manual trading.

It definitely takes practice, no doubt about that. Using a demo market is a great way to try out ideas without getting burnt.

Finally, we are working hard on a host of new indicator based bots (Bollinger band, MACD, EMA cross-over etc.) so it might be that you prefer those. First up is another release and then those new bots will be our main focus.

Thanks for your kind words regarding the interface and good luck!

Best wishes,
Jonathan