BacktestSimplePricingModel#

tradeexecutor.backtest.backtest_pricing.BacktestSimplePricingModel Python class in Trading Strategy framework.

class BacktestSimplePricingModel[source]#

Bases: PricingModel

Look up the historical prices.

Open price of the price candle at the timestamp, or closest earlier timestamp is used
Different pricing model can be used for rebalances (more coarse) and stop losses (more granular)
This is a simple model and does not use liquidity data for the price impact estimation
We provide data_delay_tolerance to deal with potential gaps in price data

__init__(candle_universe, routing_model, data_delay_tolerance=Timedelta('2 days 00:00:00'), candle_timepoint_kind='open', very_small_amount=Decimal('0.10'), time_bucket=None)[source]#

Parameters:

candle_universe (GroupedCandleUniverse) – Candles where our backtesing date comes from
routing_model (RoutingModel) – How do we route trades between different pairs TODO: Now ignored
data_delay_tolerance –
How long time gaps we allow in the backtesting data before aborting the backtesting with an exception. This is an safety check for bad data.

Sometimes there cannot be trades for days if the blockchain has been halted, and thus no price data available.
candle_timepoint_kind – Do we use opening or closing price in backtesting
very_small_amount – What kind o a test amount we do use for a trade when we do not know the actual size of the trade.
time_bucket (Optional[TimeBucket]) –
The granularity of the price data.

Currently used for diagnostics and debug only.

Methods

`__init__`(candle_universe, routing_model[, ...])	param candle_universe:
`check_supported_quote_token`(pair)
`get_buy_price`(ts, pair, reserve)	Get the sell price for an asset.
`get_mid_price`(ts, pair)	Get the mid price by the candle.
`get_pair_fee`(ts, pair)	Figure out the fee from a pair or a routing.
`get_pair_for_id`(internal_id)	Look up a trading pair.
`get_sell_price`(ts, pair, quantity)	Get the sell price for an asset.
`quantize_base_quantity`(pair, quantity[, ...])	Convert any base token quantity to the native token units by its ERC-20 decimals.

__init__(candle_universe, routing_model, data_delay_tolerance=Timedelta('2 days 00:00:00'), candle_timepoint_kind='open', very_small_amount=Decimal('0.10'), time_bucket=None)[source]#

Parameters:

candle_universe (GroupedCandleUniverse) – Candles where our backtesing date comes from
routing_model (RoutingModel) – How do we route trades between different pairs TODO: Now ignored
data_delay_tolerance –
How long time gaps we allow in the backtesting data before aborting the backtesting with an exception. This is an safety check for bad data.

Sometimes there cannot be trades for days if the blockchain has been halted, and thus no price data available.
candle_timepoint_kind – Do we use opening or closing price in backtesting
very_small_amount – What kind o a test amount we do use for a trade when we do not know the actual size of the trade.
time_bucket (Optional[TimeBucket]) –
The granularity of the price data.

Currently used for diagnostics and debug only.

get_pair_for_id(internal_id)[source]#

Look up a trading pair.

Useful if a strategy is only dealing with pair integer ids.

Parameters:: internal_id (int) –
Return type:: Optional[TradingPairIdentifier]

get_sell_price(ts, pair, quantity)[source]#

Get the sell price for an asset.

Parameters:

ts (datetime) – When to get the price. Used in backtesting. Live models may ignore.
pair (TradingPairIdentifier) – Trading pair we are intereted in
quantity (Optional[Decimal]) – If the sel quantity is known, get the price with price impact.

Returns:

Price structure for the trade.

Return type:

TradePricing

get_buy_price(ts, pair, reserve)[source]#

Get the sell price for an asset.

Parameters:

ts (datetime) – When to get the price. Used in backtesting. Live models may ignore.
pair (TradingPairIdentifier) – Trading pair we are intereted in
reserve (Optional[Decimal]) – If the buy token quantity quantity is known, get the buy price with price impact.

Returns: