Pricing and Hedging a European Straddle with Calls and Bonds
Summary
The document derives the risk-neutral value of a European straddle with payoff equal to the absolute difference between the terminal asset price and strike. Under the Black–Scholes assumptions, it expresses the value as a call and put, then rewrites it as two calls plus a bond position and a short position in the underlying. This decomposition gives the proposed constant holdings: short one share, hold a bond paying the strike at maturity, and own two calls at that strike.
The replies support the replication by decomposing the terminal payoff into a call and put and adding their hedges. This is an algebraic replication in the standard frictionless Black–Scholes setting; it does not establish that the same fixed holdings apply under transaction costs, dividends, or changing model assumptions. One reply’s notation for the call price is imprecise, but the intended decomposition is clear.
Key ideas
- A straddle payoff can be written as the sum of a call and a put with the same strike and expiry.
- The put-call payoff identity also expresses the straddle as two calls, short one share, and a bond paying the strike.
- Under Black–Scholes assumptions, this decomposition yields the straddle’s value process and a constant replication portfolio.
- The hedge follows by adding the hedges of the component options and underlying positions.
Tags
Full text
# Price and constant hedging portfolio for straddle: $X=|S(T)-K|$
# Price and constant hedging portfolio for straddle: $X=|S(T)-K|$
wondering if somebody could check my answer for a homework question!
Given a straddle, characterized by its pay-off at maturity $X=|S(T)-K|$, I am asked to find the price of the (simple) claim at any time $t\in [0,T]$. Starting at $$V(t)=e^{-r(T-t)}\mathbb{E_Q}[|S(T)-K||\mathcal{F}_t]$$ with $S(t)=S_0\exp{((r-\frac{1}{2}\sigma^2)t+\sigma W^\mathbb{Q}(t))}$, I finally end up with the value process: $$V(t)=2S(t)\Phi(d_1)-2Ke^{-r(T-t)}\Phi(d_2)-S(t)+Ke^{-r(T-t)}$$ where $$d_1=\frac{\ln(S(t)/K)+(r+\frac{1}{2}\sigma^2)(T-t)}{\sigma\sqrt{T-t}},\quad d_2=d_1-\sigma\sqrt{T-t}$$
Now, it is suggested that there exists a constant hedging portfolio for this claim. The portfolio consists not only of stocks and bonds, but also contains European call options. That is, I wish to find a portfolio with value process $\hat{V}(t)=x_tS(t)+y_tB(t)+z_tC(t)$, where $(x_t,y_t,z_t)$ remains constant for $t\in[0,T]$. My gut tells me that $(x_t,y_t,z_t)=(-1,Ke^{-rT},2)$ as the value process of the claim can be written as: $$V(t)=Ke^{-r(T-t)}-S(t)+2C(t)$$ with $C(t)$ the price for a call option (by the Black-Scholes formula). I can "semi-justify" this by defining $V(t)=v(t,x)+2C(t)$ with $v(t,x)=Ke^{-r(T-t)}-x$, so that $v$ satisfies the Black-Scholes partial differential equation and thus the portfolio with $$x_t=v_x(t,S(t))=-1, \qquad y_t=e^{rt}(v(t,S(t))-x_tS(t))=Ke^{-rT}$$ hedges the claim $K-S(t)$. Then I just add $2C(t)$ to both the value process of the portfolio and $v(t,S(t))$.
So! My two questions here are:
- Have I calculated the value process $V(t)$ for the claim correctly?
- Have I correctly justified the hedging portfolio for the claim?
Thanks in advance!
## Answer by Mark Joshi (score 4)
https://quant.stackexchange.com/a/27441
just take a call and a put struck at $K$ and add them together. For the hedge just add the hedges together as well.
## Answer by user16651 (score 1)
https://quant.stackexchange.com/a/27442
Another approach as follow. The $T$-Straddle option $X$, i.e. $$X=\left\{ \begin{align} & K-S(T)\quad ,\quad 0<S(T)\le K \\ & S(T)-K\quad ,\quad S(T)>K \\ \end{align} \right. $$ has then following contract function $$\Phi (x)=\left\{ \begin{align} & K-x\quad ,\quad 0<x\le K \\ & x-K\quad ,\quad x>K \\ \end{align} \right. $$ which can be decomposed into the following basic contract functions written $$\Phi (x)=K\times {{\Phi }_{B}}(x)-1\times {{\Phi }_{S}}(x)+2\times {{\Phi }_{C,K}}(x)$$ such that \begin{align} & {{\Phi }_{B}}(x)=1 \\ & {{\Phi }_{S}}(x)=x \\ & {{\Phi }_{C,K}}(x)=\max \{x-K\,,0\} \\ \end{align} then $$\Pi (t,X)=K{{e}^{-r(T-t)}}-{{S}_{t}}+2{{C}_{BC}}({{S}_{T}},K)$$ Note \begin{align} & E_{t}^{Q}\left[ {{\Phi }_{B}}(x) \right]=E_{t}^{Q}\left[ 1 \right]={{e}^{-r(T-t)}} \\ & E_{t}^{Q}\left[ {{\Phi }_{S}}(x) \right]=E_{t}^{Q}\left[ {{S}_{t}} \right]={{S}_{t}} \\ & E_{t}^{Q}\left[ {{\Phi }_{C,K}}(x) \right]=E_{t}^{Q}\left[ {{({{S}_{T}}-K)}^{+}}]=C_{BC}(S_T,K \right) \\ \end{align}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.