We are engaged in the development of a cryptocurrency exchange. It has already been developed. We need a market maker bot.
The main task of the bot is keeping a spread and ensuring the liquidity of transactions.
Additional tasks are:
1. internal arbitrage, i.e. the calculation of the value of coins in pairs and buying or selling the coins with low or high value.
2. external arbitrage, that is the same, but using the information from third-party exchanges.
Should be used the information from three exchanges - Binance, Bitfinex and HitBTC.
3. Filling the order book - adding the orders for buying and selling for the purpose of liquidity and holding the price in a certain price range.
The price range is set by several orders at the price of the maximum accumulation of orders on the other exchanges.
The main order is the largest, the order size is calculated randomly at the range specified in the configuration for the currency pair. It serves as an obstacle for leaving the price from the price range, that is calculated based on the accumulation of orders on other exchanges.
Preliminary, 5-10% of the basic order serves as an obstacle to price changes in the case when the trend is not defined and the signal. 1-0.5% of the price, at the close of the signal order. Analysts must be notified and use the bot settings to change the standard algorithm.
The standard algorithm provides:
in the case of breaking the order in the direction of the market (focus on external exchanges), the price ranges move in the direction of the price movement. Small orders are placed on the price movement to prevent a sharp jump in the price;
in case of a breaking the order against the market, the price ragers remain and additional liquidity adds to them.
A break is the execution of a signal order.
All settings of the bot should be changed through the control interface in real time without restarting the bot.
The software product must be submitted in the working version and source code.
The API of exchange provides the following data:
1. the depth of order book (returns all active orders for buying and selling), as well as the total sum according to the amount and the price (buying and selling) of order book,
the data is returned based on the transferred limit parameter, which determines how deep to dig for these active orders;
2. all orders for a certain user for a certain order book return not only the active orders, but also conditional orders stop loss,take profit and trading stop.
If you need any additional data, we are ready to implement it in the API.
I would like to see your suggestions about the implementation and algorithms, cost and terms.