CurveSim simulates Curve finance pools with optimal arbitrageurs trading against them. It's primary use is to determine optimal amplitude (A) and fee parameters given historical price and volume feeds.
Python3, scipy, numpy, pandas, matplotlib, itertools, copy
To simulate a pool with a range of A and/or fee parameters, use the CurveSim.Asim function. Historical data for the coins in 3pool (2 months, 30 min intervals) are provided for testing. Data was acquired from nomics.com and averaged (volume-weighted) across exchanges. Coin pair data (e.g., DAI-USDC) are used to estimate optimal arbitrage trades at each time-point, limiting trade sizes according to historical volume. Coin prices in USD (e.g., DAI-USD) are used to estimate the pool's value over time.
To approximate realistic market conditions, we initiate the pool with the current pool size, and use the "vol_mult" argument to limit trade volume based on the expected proportion of market volume that goes through Curve (e.g., monthly pool volume / monthly price feed volume).
For example, to simulate 3pool with a range of A values, we use:
import CurveSim
coins = ['DAI','USDC','USDT']
D = int(856681803.23*10**18) #10**18 precision
quote = 'USD'
A_list = 'stable' #Default A range for stablecoins, 2**(np.array(range(11,25))/2)
vol_mult = .45
ar, bal, volatility, pool_value, slippage_cost, log_returns, log_returns_hold, err = CurveSim.Asim(D, coins, quote, A_list=A_list, vol_mult=vol_mult, plot=True)To simulate 3pool with a range of A values and fees (caution, may run for a long time), we use:
coins = ['DAI','USDC','USDT']
D = int(856681803.23*10**18)
quote = 'USD'
A_list = 'stable'
fee_list = np.linspace(.0001,.001,10)*10**10 #10**10 precision
vol_mult = .45
ar, bal, volatility, pool_value, slippage_cost, log_returns, log_returns_hold, err = CurveSim.Asim(D, coins, quote, A_list=A_list, fee_list=fee_list, vol_mult=vol_mult, plot=True)We provide our code for downloading data from nomics.com in the file "nomics.py". This code requires a paid API key. We will provide comparable code for CoinGecko (free) soon.