Replicating a Claim on the Time Average of Squared Brownian Motion
Summary
The document considers a claim paying the time average of squared Brownian motion when the underlying follows standard Brownian motion and the interest rate is zero. It gives the initial fair value as the expected payoff, then uses Itô calculus to rewrite the claim as that initial amount plus a stochastic integral against the traded asset.
The integral identifies a replicating dynamic position: hold an amount proportional to the current Brownian level and decreasing with time remaining. Starting with the stated initial value and adjusting this holding continuously reproduces the terminal claim in the model. The result is a specific example of riskless replication, not a general strategy for arbitrary asset dynamics or nonzero rates. The document assumes the Brownian process is tradable and supplies no discussion of transaction costs, discrete rebalancing, or market frictions.
Key ideas
- The claim pays the time average of the squared Brownian price process.
- Under the stated zero-rate setup, the fair initial value is the expected payoff.
- Itô calculus expresses the payoff as initial capital plus a stochastic integral.
- The integrand gives the dynamic holding needed to replicate the claim in the idealized model.
- The replication result assumes continuous trading in the specified Brownian asset.
Tags
Full text
# Dynamic hedging strategy example
# Dynamic hedging strategy example
I am faced with the following problem. Let the standard Brownian motion $W_t$ be the price process of a traded asset in an economy with zero interest rate. Define $$A_T=\frac{1}{T}\int_0^T W_t^2 dt$$
I have two questions:
- What is the fair price at time $t=0$ of a contract that offers $A_T$?
- How do we form a dynamic hedging strategy that eliminates all risk in having to deliver this claim?
I answered part 1 by simply taking the expectation. The fair price is $E(A_T\mid \mathcal{F}_0)=\frac{T}{2}$. How could the dynamic hedging be strategised?
## Answer by Bravo (score 5, accepted)
https://quant.stackexchange.com/a/8190
Consider a dynamic hedging strategy where you invest $H_t$ in the stock at time $t$. To eliminate all risk, the value of the investment must be equal to the claim at time $T$. Using Ito's calculus, we could express $A_T$ as follows:
$$A_T=\frac{T}{2}+\int_0^T 2W_t \left(1-\frac{t}{T}\right) dW_t=\frac{T}{2}+\int_0^T H_tdW_t$$
Thus the strategy would be to start with an amount $T/2$ (fair price at $t=0$) and invest $H_t=2W_t(1-t/T)$ dynamically in the stock.
PS: This is a special case of the Black-Scholes setup where the interest rate $r=0$. If $X_t$ is the value of the holding and $S_t$ is the stock price, the value of $dX_t$ is $dX_t=H_tdS_t+(X_t-H_tS_t)rdt$. $H_t$ is the amount to be investment dynamically in the stock, and is also known as the delta of the option.
## Answer by IMK (score 0)
https://quant.stackexchange.com/a/8171
Your fair price formula is not general enough. You need to make W(0) appear, then differentiate wrt it. This will be your delta. You assumed W(0) = 0 and got rid of it, and now you're stuck with nothing to differentiate wrt.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.