📁 Strategy Info¶
- Folder: /hummingbot/strategy/twap
- Configs: twap_config_map.py
- Maintainer: None
🏆 Strategy Tier¶
Community strategies have passed the Minimum Voting Power Threshold in the latest Poll and are included in each monthly release. They are not maintained by Hummingbot Foundation but may be maintained by a community member.
This strategy is a simple bot that places a series of limit orders on an exchange, while allowing users to control order size, price, and duration.
We recommend this strategy as a starting point for developers looking to build their own strategies, and it is used as reference for articles in Developer Reference: Strategies.
🏦 Exchanges supported¶
- SPOT CLOB CEX
🛠️ Strategy configs¶
||string||True||Enter the name of spot connector|
||string||True||Enter the token trading pair you would like to trade on
||string||buy||True||What operation will be executed? (buy/sell)|
||decimal||1||True||What is the total amount of [base_token] to be traded?|
||decimal||1||True||What is the amount of each individual order (denominated in the base asset, default is 1)|
||decimal||True||What is the price for the limit orders?|
||decimal||10||True||How many seconds do you want to wait between each individual order?|
||decimal||60||True||How long do you want to wait before cancelling your limit order (in seconds).|
||bool||False||False||Do you want to specify a start time and an end time for the execution?|
||decimal||False||Please enter the start date and time|
||decimal||False||Please enter the end date and time|
The description below is a general approximation of this strategy. Please inspect the strategy code in Trading Logic above to understand exactly how it works.
The TWAP strategy is a common algorithmic execution strategy used for splitting up large orders over time. Specifically, the TWAP strategy helps traders minimize slippage when buying or selling large orders. These features make the strategy more useful to traders and will help when creating future, more complex strategies:
- Incrementing / maintaining states over clock ticks
- Quantizing (rounding down to nearest tradable value) order size
- Dividing an order into segments
- Incorporating time delays between segmented orders
The TWAP strategy divides a large user order into chunks according to the following user configurations:
- Total order size
- number of individual orders
- time delay between orders
The orders are then split into tradable (quantized) amounts and executed sequentially with the indicated time delay in between orders. There is no time delay before the first order. Because only one order is placed in a clock tick, a state machine is needed to emit multiple orders over different clock ticks. To see the executed orders, type history into the command prompt.
Here are the additional user configurable parameters for the TWAP strategy (fields are added to
time_delay: Change the question to ask for the number of seconds to delay each individual order. (e.g. How many seconds do you want to wait between each individual order?)
num_individual_orders: a new field added to the config map. It should ask for the number of individual orders that an order should be split up into. (e.g.Into how many individual orders do you want to split this order?)
The TWAP strategy logic is trying to split a large order into smaller ones over time, and it does that by maintaining important information about the state when processing orders by adding state variables.
Custom state variables can be added to the strategy by setting variables in the
self._quantity_remaining: Indicates the quantity of order left to be placed as individual orders. This state variable is updated after each order is placed and persisted throughout until the order is done processing.
self._first_order: Indicates whether the current individual order is the first order.
TWAP processes orders when there is a remaining order quantity & the specified time_delay has passed. Specifically, some of the key elements in utilizing the remaining order quantity and time_delay are detailed below:
- If self._quantity_remaining is greater than 0 place an order
self._first_orderis true, we want to place order as soon as
self._current_timestamp > self._previous_timestampwe don't have a time delay before the first order
- If it isn't the first order, check that
self._current_timestamp > self._previous_timestamp + self._time_delay
- Once order is placed, update self._quantity_remaining by subtracting the amount of the order just placed
curr_order_amount: Either (total order amount)/(number of orders) or
self._quantity_remainingdepending on which is smaller
This demo is for instructional and educational purposes only. Any parameters used are purely for demo purposes only. We are not giving any legal, tax, financial, or investment advice. Every user is responsible for their use and configuration of Hummingbot.
ℹ️ More Resources¶
Strategy coding for dummies: This article is a blog post submission from our of our users. It is not directly related to TWAP strategy, but it demos how you can write a custom script for cross exchange market making strategy
Check out Hummingbot Academy for more resources related to this strategy and others!