Applying the Chain Rule to the Black–Scholes Spot Derivative
Summary
The document explains how to transform the Black–Scholes partial derivatives after changing variables from stock price and calendar time to log moneyness and scaled time. It focuses on the second derivative with respect to the stock price, building from the first derivative and applying the chain rule and product rule again.
The key result expresses the spot gamma as a scaled difference between the second and first derivatives of the transformed function. The derivation tracks the dependence of the exponential factor on log moneyness and uses the relation between log moneyness and spot. This is a symbolic clarification of a coordinate transformation used when reducing Black–Scholes to the heat equation; it provides no numerical example or independent validation, and the discussion is limited to the stated transformation.
Key ideas
- The transformed first derivative with respect to spot includes a factor that depends on log moneyness.
- Differentiating that first derivative requires both the product rule and the chain rule.
- The spot second derivative depends on the difference between the transformed second and first derivatives.
- The result is useful when converting the Black–Scholes equation into a heat equation.
Tags
Full text
# How to apply the chain rule for partial derivatives to transformations?
# How to apply the chain rule for partial derivatives to transformations?
I'm currently working to solve the Black-Scholes model partial differential equation (it's a model for a.o. stock option prices). The Black-Scholes equation for a calloption C(S,t) is given by
$ \frac{∂C}{∂t}+\frac{1}{2} σ^2 S^2 \frac{∂^2 C}{∂S^2}+rS \frac{∂C}{∂S}-rC=0$
S is the stock price and t is time to expiration.
I am solving the equation by transforming it to the heat equation. I am having a little trouble with the first few transformation. The transformations are:
- $x=\ln(\frac{S}{K})$ which gives $S=Ke^x$
- $τ=\frac{σ^2}{2} (T-t)$ which gives $t=T-\frac{2τ}{σ^2}$
The function becomes:
- $U(x,τ)= \frac{1}{K} C(S,t)=\frac{1}{K} C(Ke^x,T- \frac{2τ}{σ^2})$
These transformation applied to the partial differential equation above gives the following outcomes for the different terms. I found the following solution on the internet, my problem is that I don't really understand what they do here:
$\frac{∂C}{∂t}=K \frac{∂U}{∂τ} \frac{∂τ}{∂t}=\frac{-Kσ^2}{2} \frac{∂U}{∂τ},$
$\frac{∂C}{∂S}=K \frac{∂U}{∂x} \frac{∂x}{∂S}=\frac{K}{S} \frac{∂U}{∂x}=e^{-x} \frac{∂U}{∂x},$
$\frac{∂^2 C}{∂S^2}=\frac{-K}{S^2} \frac{∂U}{\partial x}+\frac{K}{S} \frac{∂}{∂S} (\frac{∂U}{∂x})$ $=\frac{-K}{S^2} \frac{∂U}{∂x}+\frac{K}{S} \frac{∂}{∂x} (\frac{∂U}{∂x}) \frac{∂x}{∂S}$ $=\frac{-K}{S^2} \frac{∂U}{∂x}+\frac{K}{S^2} \frac{∂^2 U}{∂x^2}$ $=\frac{e^{-2x}}{K} (\frac{∂^2 U}{∂x^2} \frac{-∂U}{∂x}$)
Can someone help me out? I would really appreciate it!!!!
thanks in advance
## Answer by LocalVolatility (score 2)
https://quant.stackexchange.com/a/31487
As you state in your comment, you only have trouble with the second partial derivative w.r.t. the spot. So you understand how the first partial derivative is obtained
\begin{equation} \frac{\partial C}{\partial S} = e^{-x} \frac{\partial U}{\partial x}. \end{equation}
Then you just carefully apply the chain rule again. In ALL details:
\begin{eqnarray} \frac{\partial^2 C}{\partial S^2} & = & \frac{\partial}{\partial S} \left\{ e^{-x} \frac{\partial U}{\partial x} \right\}\\ & = & \frac{\partial}{\partial x} \left\{ e^{-x} \right\} \frac{\partial U}{\partial S} + e^{-x} \frac{\partial}{\partial x} \left\{ \frac{\partial U}{\partial S} \right\}\\ & = & \frac{\partial}{\partial x} \left\{ e^{-x} \right\} \frac{\partial x}{\partial S} \frac{\partial U}{\partial x} + e^{-x} \frac{\partial^2 U}{\partial x^2} \frac{\partial x}{\partial S}\\ & = & e^{-x} \frac{\partial x}{\partial S} \left( \frac{\partial^2 U}{\partial x^2} - \frac{\partial U}{\partial x} \right)\\ & = & e^{-x} \frac{1}{S} \left( \frac{\partial^2 U}{\partial x^2} - \frac{\partial U}{\partial x} \right)\\ & = & e^{-2 x} \frac{1}{K} \left( \frac{\partial^2 U}{\partial x^2} - \frac{\partial U}{\partial x} \right) \end{eqnarray}Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.