Dakko Python SDK

Dakko Python SDK is a Python library for interacting with the Dakko Services. Supported exchanges:

• Coinbase (only "Exchange Domestic");
• Kraken;
• MEXC;
• Bitfinex;
• Bybit;
• OKX;
• Binance;
• Huobi.

Supported base assets:

• SOL;
• ETH;
• BTC;
• DOT;
• ADA;
• XRP;
• BCH;
• LINK;
• MATIC;
• LTC;
• AVAX.

Supported quote assets:

• USDT;
• USD;
• USDC.

Installation

Using pip:

pip install git+https://github.com/dakko-group/dakko-sdk.git

Using poetry:

poetry add git+https://github.com/dakko-group/dakko-sdk.git

Upgrading

Using pip:

pip uninstall dakko-sdk
pip install git+https://github.com/dakko-group/dakko-sdk.git

Using poetry:

poetry update dakko-sdk

Usage

Authentication

To authenticate with the Dakko Services, you need to provide an API token.

from dakko.mie import MarketImpactEstimation

mie = MarketImpactEstimation()
mie.auth(token="your_token")

Market Impact Estimation

The Market Impact Estimation (MIE) service provides tools for estimating the market impact of a trade with a given size, exchange(s), direction and instrument:

mie.get_optimal_trade_size(
    exchanges=["kraken", "binance"], # a list of exchanges or wildcard, i.e. "*" or "all"
    base_asset="eth",
    trade_size=1_000_000,
    is_sell=True,
    max_abs_slippage_bps=5,
)

In this case slippage threshold is a target value, so model will try to identify the best trade size for a given criteria. The MIE supports slippage estimation across different trade sizes. In this case using steps argument we build linear space with start = 1000 USD and end = trade_size USD points to build 2D distribution of slippage vs trade size:

mie.estimate_slippage_across_sizes(
    exchanges="*", # wildcard is used
    base_asset="sol",
    trade_size=100_000,
    steps=5,
    is_sell=False,
    max_abs_slippage_bps=5,
)

To get estimation of a given trade size allocation across exchanges you can use the following method:

mie.estimate_trade_allocation(
    exchanges="*", # wildcard
    base_asset="eth",
    quote_asset="usdt",
    trade_size=10_000_000,
    is_sell=True,
)

Backtesting Mode

The MIE supports baseline backtesting capabilities with 30 days back window. To run the model in backtesting mode you need to specify ts or isodate parameter (if isodate is provided, ts will be ignored). In this case for ts we require UNIX timestamp format in seconds (integer), for isodate we require ISO 8601 format (string). For example:

# example with ts
mie.get_optimal_trade_size(
    exchanges=["kraken", "mexc", "coinbase"],
    base_asset="sol",
    quote_asset="usdt",
    is_sell=True,
    trade_size=10_000_000,
    ts=1721440800 # corresponds to 2024-07-20T05:00:00Z
)

# example with isodate
mie.get_optimal_trade_size(
    exchanges=["mexc", "coinbase"],
    base_asset="btc",
    quote_asset="usdt",
    is_sell=True,
    trade_size=10_000_000,
    isodate="2024-07-20T05:00:00Z"
    ts=123 # will be ignored
)

The backtesting mode is also supported for slippage estimation across trade sizes. You need to use the same parameters to enable execution in backtesting for 30 days window:

mie.estimate_slippage_across_sizes(
    exchanges=["coinbase"],
    base_asset="eth",
    quote_asset="usd",
    is_sell=True,
    trade_size=10_000_000,
    steps=10,
    ts=1721440800 # corresponds to 2024-07-20T05:00:00Z
)

We gonna add support for wider versioning and model metrics access (based on benchmarks over historical data) to enhance results and user experience.

Help

To get method description you always can use help function. For example:

help(mie.estimate_slippage_across_sizes)