Can do it numerically, or use the symbolic tool box (e.g. mathematica)

The hedging for sv model can be made by a stock (delta) and another option (vega).

如果用sv模型对欧式看涨期权进行对冲，那么其它期权用的是执行价格K不同的期权吗？（或者是T不同）
还有sv模型里没有隐含波动率为什么会有vega，sv模型的vega指的是对随机波动率的偏导吗？对随机项求偏导也是用数值方法？
谢谢~

1) yes, I prefer different Ks

2) vega is just a local sensitivity with repeat to an option parameter (not a concept only in BS model. Actually BS assumes constant volatility, but how can the model have vega? the answer by Hull is that empirically the result of taking volatility as a time varying process is consistent with that of vega in the BS model. So we can roughly regard it as reasonable).

3) Volatility is stochastic in SV models, but it is also an t adapted process (the same as the stock price), the only different thing is it is not observable (or tradable). Stock price is also stochastic, there is not much difference for the calculation of delta and vega.

best,

你好！
请问, 实务中做SV hedge时，构建组合使得对Stock和Variance的一阶导的系数等于0？还是根据SV model求出Implied Vol, 然后套用BS Model，对冲Greeks？

我谈谈我的想法，我觉得应该是对s和v求偏导，因为这就是随机波动率模型的风险源，没有必要在转换为隐含波动率。但是目前我看的文献有用最小方差方法得到的delta和gamma对冲效果比偏导的好。