VWAP Execution by Slicing Orders Against Order Book Depth
Summary
This execution strategy divides a target purchase or sale into repeated market orders sized from visible order book depth. It converts a total quote-currency budget into a fixed base-asset target using the initial price, then checks the book out to a configured price spread. At each interval, it proposes an order equal to a configured fraction of the cumulative available volume, capped by the remaining target. A balance checker adjusts each proposed order before submission, and fill events update the remaining quantity and progress toward completion.
The example exposes settings for exchange, pair, side, total quote volume, spread, book-volume fraction, and delay between orders. The source also records realized quote volume and trades, though it does not use those records to optimize later slices. This is an execution template, not a tested performance study: no benchmarks or market-impact results are given. Its depth-based sizing relies on current order book snapshots, and market conditions may change before orders fill; the code also does not describe a schedule tied to market volume over time.
Key ideas
- The target base quantity is calculated from the quote budget and the initial price.
- Each slice uses a fraction of order book volume available within the configured spread.
- The proposed order is capped by the remaining target and adjusted for account balances.
- Fill events reduce the outstanding target and determine when the execution is complete.
- The example gives no evidence comparing execution quality with alternative schedules.
Tags
Full text
# VWAPExample
# VWAPExample
Configuration parameters for the VWAP strategy.
## Source (Apache-2.0)
```python
import logging
import math
import os
from decimal import Decimal
from typing import Dict, List
from pydantic import Field
from hummingbot.connector.connector_base import ConnectorBase
from hummingbot.connector.utils import split_hb_trading_pair
from hummingbot.core.data_type.common import MarketDict
from hummingbot.core.data_type.order_candidate import OrderCandidate
from hummingbot.core.event.events import OrderFilledEvent, OrderType, TradeType
from hummingbot.strategy.strategy_v2_base import StrategyV2Base, StrategyV2ConfigBase
class VWAPConfig(StrategyV2ConfigBase):
"""
Configuration parameters for the VWAP strategy.
"""
script_file_name: str = os.path.basename(__file__)
controllers_config: List[str] = []
connector_name: str = Field("binance_paper_trade", json_schema_extra={
"prompt": lambda mi: "Exchange where the bot will place orders",
"prompt_on_new": True})
trading_pair: str = Field("ETH-USDT", json_schema_extra={
"prompt": lambda mi: "Trading pair where the bot will place orders",
"prompt_on_new": True})
is_buy: bool = Field(True, json_schema_extra={
"prompt": lambda mi: "Buying or selling the base asset? (True for buy, False for sell)",
"prompt_on_new": True})
total_volume_quote: Decimal = Field(1000, json_schema_extra={
"prompt": lambda mi: "Total volume to buy/sell (in quote asset)",
"prompt_on_new": True})
price_spread: float = Field(0.001, json_schema_extra={
"prompt": lambda mi: "Maximum price spread to use when placing orders (0.001 = 0.1%)",
"prompt_on_new": True})
volume_perc: float = Field(0.001, json_schema_extra={
"prompt": lambda mi: "Percentage of the order book volume to buy/sell (0.001 = 0.1%)",
"prompt_on_new": True})
order_delay_time: int = Field(10, json_schema_extra={
"prompt": lambda mi: "Delay time between orders (in seconds)",
"prompt_on_new": True})
def update_markets(self, markets: MarketDict) -> MarketDict:
markets[self.connector_name] = markets.get(self.connector_name, set()) | {self.trading_pair}
return markets
class VWAPExample(StrategyV2Base):
"""
BotCamp Cohort: 7 (Apr 2024)
Description:
This is an updated version of simple_vwap_example.py. Changes include:
- Users can define script configuration parameters
- Total volume is expressed in quote asset rather than USD
- Use of the rate oracle has been removed
"""
def __init__(self, connectors: Dict[str, ConnectorBase], config: VWAPConfig):
super().__init__(connectors, config)
self.config = config
self.initialized = False
self.vwap: Dict = {"connector_name": self.config.connector_name,
"trading_pair": self.config.trading_pair,
"is_buy": self.config.is_buy,
"total_volume_quote": self.config.total_volume_quote,
"price_spread": self.config.price_spread,
"volume_perc": self.config.volume_perc,
"order_delay_time": self.config.order_delay_time}
last_ordered_ts = 0
def on_tick(self):
"""
Every order delay time the strategy will buy or sell the base asset. It will compute the cumulative order book
volume until the spread and buy a percentage of that.
The input of the strategy is in quote, and we will convert at initial price to get a target base that will be static.
- Create proposal (a list of order candidates)
- Check the account balance and adjust the proposal accordingly (lower order amount if needed)
- Lastly, execute the proposal on the exchange
"""
if self.last_ordered_ts < (self.current_timestamp - self.vwap["order_delay_time"]):
if self.vwap.get("status") is None:
self.init_vwap_stats()
elif self.vwap.get("status") == "ACTIVE":
vwap_order: OrderCandidate = self.create_order()
vwap_order_adjusted = self.vwap["connector"].budget_checker.adjust_candidate(vwap_order,
all_or_none=False)
if math.isclose(vwap_order_adjusted.amount, Decimal("0"), rel_tol=1E-5):
self.logger().info(f"Order adjusted: {vwap_order_adjusted.amount}, too low to place an order")
else:
self.place_order(
connector_name=self.vwap["connector_name"],
trading_pair=self.vwap["trading_pair"],
is_buy=self.vwap["is_buy"],
amount=vwap_order_adjusted.amount,
order_type=vwap_order_adjusted.order_type,
price=vwap_order_adjusted.price)
self.last_ordered_ts = self.current_timestamp
def init_vwap_stats(self):
# General parameters
vwap = self.vwap.copy()
vwap["connector"] = self.connectors[vwap["connector_name"]]
vwap["delta"] = 0
vwap["trades"] = []
vwap["status"] = "ACTIVE"
vwap["trade_type"] = TradeType.BUY if self.vwap["is_buy"] else TradeType.SELL
vwap["start_price"] = vwap["connector"].get_price(vwap["trading_pair"], vwap["is_buy"])
vwap["target_base_volume"] = vwap["total_volume_quote"] / vwap["start_price"]
# Compute market order scenario
orderbook_query = vwap["connector"].get_quote_volume_for_base_amount(vwap["trading_pair"], vwap["is_buy"],
vwap["target_base_volume"])
vwap["market_order_base_volume"] = orderbook_query.query_volume
vwap["market_order_quote_volume"] = orderbook_query.result_volume
vwap["volume_remaining"] = vwap["target_base_volume"]
vwap["real_quote_volume"] = Decimal(0)
self.vwap = vwap
def create_order(self) -> OrderCandidate:
"""
Retrieves the cumulative volume of the order book until the price spread is reached, then takes a percentage
of that to use as order amount.
"""
# Compute the new price using the max spread allowed
mid_price = float(self.vwap["connector"].get_mid_price(self.vwap["trading_pair"]))
price_multiplier = 1 + self.vwap["price_spread"] if self.vwap["is_buy"] else 1 - self.vwap["price_spread"]
price_affected_by_spread = mid_price * price_multiplier
# Query the cumulative volume until the price affected by spread
orderbook_query = self.vwap["connector"].get_volume_for_price(
trading_pair=self.vwap["trading_pair"],
is_buy=self.vwap["is_buy"],
price=price_affected_by_spread)
volume_for_price = orderbook_query.result_volume
# Check if the volume available is higher than the remaining
amount = min(volume_for_price * Decimal(self.vwap["volume_perc"]), Decimal(self.vwap["volume_remaining"]))
# Quantize the order amount and price
amount = self.vwap["connector"].quantize_order_amount(self.vwap["trading_pair"], amount)
price = self.vwap["connector"].quantize_order_price(self.vwap["trading_pair"],
Decimal(price_affected_by_spread))
# Create the Order Candidate
vwap_order = OrderCandidate(
trading_pair=self.vwap["trading_pair"],
is_maker=False,
order_type=OrderType.MARKET,
order_side=self.vwap["trade_type"],
amount=amount,
price=price)
return vwap_order
def place_order(self,
connector_name: str,
trading_pair: str,
is_buy: bool,
amount: Decimal,
order_type: OrderType,
price=Decimal("NaN"),
):
if is_buy:
self.buy(connector_name, trading_pair, amount, order_type, price)
else:
self.sell(connector_name, trading_pair, amount, order_type, price)
def did_fill_order(self, event: OrderFilledEvent):
"""
Listens to fill order event to log it and notify the Hummingbot application.
"""
if event.trading_pair == self.vwap["trading_pair"] and event.trade_type == self.vwap["trade_type"]:
self.vwap["volume_remaining"] -= event.amount
self.vwap["delta"] = (self.vwap["target_base_volume"] - self.vwap["volume_remaining"]) / self.vwap[
"target_base_volume"]
self.vwap["real_quote_volume"] += event.price * event.amount
self.vwap["trades"].append(event)
if math.isclose(self.vwap["delta"], 1, rel_tol=1e-5):
self.vwap["status"] = "COMPLETE"
msg = (f"({event.trading_pair}) {event.trade_type.name} order (price: {round(event.price, 2)}) of "
f"{round(event.amount, 2)} "
f"{split_hb_trading_pair(event.trading_pair)[0]} is filled.")
self.log_with_clock(logging.INFO, msg)
self.notify_hb_app_with_timestamp(msg)
def format_status(self) -> str:
"""
Returns status of the current strategy on user balances and current active orders. This function is called
when status command is issued. Override this function to create custom status display output.
"""
if not self.ready_to_trade:
return "Market connectors are not ready."
lines = []
warning_lines = []
warning_lines.extend(self.network_warning(self.get_market_trading_pair_tuples()))
balance_df = self.get_balance_df()
lines.extend(["", " Balances:"] + [" " + line for line in balance_df.to_string(index=False).split("\n")])
try:
df = self.active_orders_df()
lines.extend(["", " Orders:"] + [" " + line for line in df.to_string(index=False).split("\n")])
except ValueError:
lines.extend(["", " No active maker orders."])
lines.extend(["", "VWAP Info:"] + [" " + key + ": " + value
for key, value in self.vwap.items()
if isinstance(value, str)])
lines.extend(["", "VWAP Stats:"] + [" " + key + ": " + str(round(value, 4))
for key, value in self.vwap.items()
if type(value) in [int, float, Decimal]])
return "\n".join(lines)
```Shown in full with attribution under the source's licence. Licence: Apache-2.0
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.