我正在寻找一种可以在Python中更快地计算隐含波动率的库。我有大约1百万行期权数据需要计算隐含波动率。有什么最快的方法可以计算隐含波动率呢?我尝试过使用 py_vollib,但它不支持向量化。计算需要大约5分钟左右。是否有其他库可以帮助加速计算?人们在实时波动率计算中使用什么,当每秒钟会出现数百万行数据时?
我正在寻找一种可以在Python中更快地计算隐含波动率的库。我有大约1百万行期权数据需要计算隐含波动率。有什么最快的方法可以计算隐含波动率呢?我尝试过使用 py_vollib,但它不支持向量化。计算需要大约5分钟左右。是否有其他库可以帮助加速计算?人们在实时波动率计算中使用什么,当每秒钟会出现数百万行数据时?
您需要意识到,隐含波动率计算是计算密集型的,如果您想要实时数据,也许Python不是最佳解决方案。
以下是您需要的函数示例:
import numpy as np
from scipy.stats import norm
N = norm.cdf
def bs_call(S, K, T, r, vol):
d1 = (np.log(S/K) + (r + 0.5*vol**2)*T) / (vol*np.sqrt(T))
d2 = d1 - vol * np.sqrt(T)
return S * norm.cdf(d1) - np.exp(-r * T) * K * norm.cdf(d2)
def bs_vega(S, K, T, r, sigma):
d1 = (np.log(S / K) + (r + 0.5 * sigma ** 2) * T) / (sigma * np.sqrt(T))
return S * norm.pdf(d1) * np.sqrt(T)
def find_vol(target_value, S, K, T, r, *args):
MAX_ITERATIONS = 200
PRECISION = 1.0e-5
sigma = 0.5
for i in range(0, MAX_ITERATIONS):
price = bs_call(S, K, T, r, sigma)
vega = bs_vega(S, K, T, r, sigma)
diff = target_value - price # our root
if (abs(diff) < PRECISION):
return sigma
sigma = sigma + diff/vega # f(x) / f'(x)
return sigma # value wasn't found, return best guess so far
计算单个值足够快
S = 100
K = 100
T = 11
r = 0.01
vol = 0.25
V_market = bs_call(S, K, T, r, vol)
implied_vol = find_vol(V_market, S, K, T, r)
print ('Implied vol: %.2f%%' % (implied_vol * 100))
print ('Market price = %.2f' % V_market)
print ('Model price = %.2f' % bs_call(S, K, T, r, implied_vol))
隐含波动率:25.00%
市场价格= 35.94
模型价格= 35.94
但是,如果您尝试计算很多次,您会意识到这需要一些时间...
%%time
size = 10000
S = np.random.randint(100, 200, size)
K = S * 1.25
T = np.ones(size)
R = np.random.randint(0, 3, size) / 100
vols = np.random.randint(15, 50, size) / 100
prices = bs_call(S, K, T, R, vols)
params = np.vstack((prices, S, K, T, R, vols))
vols = list(map(find_vol, *params))
耗时:10.5秒
def bs_put(S,K,r,vol,T)
,它返回bs_call(S,K,r,vol,T) - S + np.exp(-r * (T)) * K
,但我不确定这将如何影响“def find_vol”方法...你能帮忙吗? - JC23find_vol
函数基本上是使用函数及其导数的牛顿-拉弗森方法来寻找根。用于定价看涨期权和看跌期权的BS公式的导数相对于波动率的变化(vega)是相同的,因此您只需更改函数以相应地确定价格(将看涨期权更改为看跌期权)。您可以通过看涨期权和看跌期权的平价关系或将定价公式更改为K * np.exp(-r * T) * N(-d2) - S * N(-d1)
来定价看跌期权。 - David Duartenorm.cdf()
方法的调用更改为ndtr()
,你将获得2.4倍的性能提升。如果将norm.pdf()
方法更改为norm._pdf()
,您将获得另外一个(巨大的)增长。实施这两个更改后,上面的示例从17.7秒降至0.99 s。虽然您可能不需要所有这些,但您将失去错误检查等。参见:https://github.com/scipy/scipy/issues/1914。在scipy.special
中有ndtr()
方法。最近,py_vollib
的向量化版本已经发布在 py_vollib_vectorized 上。该版本基于 py_vollib
构建,可以更快地定价数千个期权合约并计算希腊值。
!pip install py_vollib
import py_vollib
from py_vollib.black_scholes import black_scholes as bs
from py_vollib.black_scholes.implied_volatility import implied_volatility as iv
from py_vollib.black_scholes.greeks.analytical import delta
from py_vollib.black_scholes.greeks.analytical import gamma
from py_vollib.black_scholes.greeks.analytical import rho
from py_vollib.black_scholes.greeks.analytical import theta
from py_vollib.black_scholes.greeks.analytical import vega
import numpy as np
#py_vollib.black_scholes.implied_volatility(price, S, K, t, r, flag)
"""
price (float) – the Black-Scholes option price
S (float) – underlying asset price
sigma (float) – annualized standard deviation, or volatility
K (float) – strike price
t (float) – time to expiration in years
r (float) – risk-free interest rate
flag (str) – ‘c’ or ‘p’ for call or put.
"""
def greek_val(flag, S, K, t, r, sigma):
price = bs(flag, S, K, t, r, sigma)
imp_v = iv(price, S, K, t, r, flag)
delta_calc = delta(flag, S, K, t, r, sigma)
gamma_calc = gamma(flag, S, K, t, r, sigma)
rho_calc = rho(flag, S, K, t, r, sigma)
theta_calc = theta(flag, S, K, t, r, sigma)
vega_calc = vega(flag, S, K, t, r, sigma)
return np.array([ price, imp_v ,theta_calc, delta_calc ,rho_calc ,vega_calc ,gamma_calc])
S = 8400
K = 8600
sigma = 16
r = 0.07
t = 1
call=greek_val('c', S, K, t, r, sigma)
put=greek_val('p', S, K, t, r, sigma)
py_vollib
和其底层的 lets_be_rational
库仅适用于欧式期权(https://github.com/vollib/py_vollib/issues/7)。还有,GitHub 中的评论有些误导,我发现 py_vollib
基本上无法处理深度 ITM 的美式期权。在您深入研究之前,请牢记这一点。 - Oleg Medvedyevdef goalseek(spot_price: float,
strike_price: float,
time_to_maturity: float,
option_type: str,
option_price: float):
volatility = 2.5
upper_range = 5.0
lower_range = 0
MOE = 0.0001 # Minimum margin of error
max_iters = 100
iter = 0
while iter < max_iters: # Don't iterate too much
price = proposedPrice(spot_price=spot_price,
strike_price=strike_price,
time_to_maturity=time_to_maturity,
volatility=volatility,
option_type=option_type) # BS Model Pricing
if abs((price - option_price)/option_price) < MOE:
return volatility
if price > option_price:
tmp = volatility
volatility = (volatility + lower_range)/2
upper_range = tmp
elif price < option_price:
tmp = volatility
volatility = (volatility + upper_range)/2
lower_range = tmp
iter += 1
return volatility