Balancer Pool's spot prices are the price of one token bound to the pool in relation with the rest (for example, in a WETH / DAI pool, the spot price of WETH will be in DAI. This is, for every WETH i'll get out XXX DAI from the pool (swap fees apply)
This logic is in a state update method, that will apply calc_spot_pricefrom BMath
Every pool update in the policies (i.e. balance changes in the pool's bounded tokens) will signal the update of spot prices
def calculate_spot_prices(pool: dict, ref_token: str):
swap_fee = pool['swap_fee']
balance_in = pool['tokens'][ref_token].balance
weight_in = pool['tokens'][ref_token].weight
spot_prices = {}
for token in pool['tokens']:
if token == ref_token:
continue
balance_out = pool['tokens'][token].balance
weight_out = pool['tokens'][token].weight
price = BalancerMath.calc_spot_price(token_balance_in=Decimal(balance_in),
token_weight_in=Decimal(weight_in),
token_balance_out=Decimal(balance_out),
token_weight_out=Decimal(weight_out),
swap_fee=Decimal(swap_fee))
spot_prices[token] = price
return spot_prices
def s_update_spot_prices(params, substep, state_history, previous_state, policy_input):
pool = policy_input.get('pool_update')
if pool is None:
return 'spot_prices', previous_state['spot_prices']
ref_token = params[0]['spot_price_reference']
spot_prices = calculate_spot_prices(pool, ref_token)
return 'spot_prices', spot_prices
The spot price reference can be set as a cadCAD parameter in the Notebooks, allowing for parameter sweeps to run the simulation several times with different spot price result for each token.
The code below will generate 2 subsets of results in the dataframe output after the simulation is run:
import pprint
pp = pprint.PrettyPrinter(indent=4)
from decimal import Decimal
# Spot price reference must be a symbol of a token in the pool in ALL_CAPS, you can ignore the spot price parameter for the simulations in this notebook.
parameters = {
'spot_price_reference': ['DAI', 'WETH'],
}
