Skip to content
All library documents

Replicating Reciprocal-Price and Price-Ratio Derivative Payoffs

Article Quant Q&A · Author: benh

Summary

The discussion considers hedging a payoff that depends on the reciprocal of a stock price, then extends the problem to a payoff proportional to the ratio of a future stock price to its current price. One response uses a static option-replication identity: a smooth payoff can be represented using a linear position in the underlying plus weighted calls and puts. Applied to the reciprocal payoff, the weighting depends on strike, and the example centers the linear approximation at the expected stock price.

For the price-ratio payoff, the responses suggest holding shares in proportion to the reciprocal of the initial price, with dividend cash flows accounted for separately. Another proposed route uses the conditional expected stock price and replicates with forwards and bonds. The material does not provide a complete derivation or a general model-free hedge: one response invokes Black–Scholes without showing its formulas, while the option-based construction depends on available strikes and pricing assumptions. Dividend timing and discounting also affect the stated adjustment.

Key ideas

  • A reciprocal-price payoff can be represented using a linear underlying position and a continuum of calls and puts.
  • The option weights in the replication depend on the curvature of the reciprocal payoff across strikes.
  • A payoff proportional to the ratio of terminal and initial stock prices can be constructed by holding shares based on the initial price.
  • Dividends received during the holding period must be included when pricing the price-ratio payoff.
  • The responses leave model assumptions and some replication details unspecified.

Tags

Full text
# How to hedge a derivative that pays the reciprocal of the stock price?


# How to hedge a derivative that pays the reciprocal of the stock price?












1) Suppose S is the stock price, how to hedge a derivative that pays $1/S_t$ at time $t$?

2) Suppose there will be a dividend of amount $d$ between $t$ and $T$, how to hedge a derivative that pays $100 $*$ S_T/S_t$ at time $T$?

The person who asked me the question said we don't need to assume the distribution of S here.

Thanks!

## Answer by Gordon (score 3)

https://quant.stackexchange.com/a/14224

Note that, for a smooth function and constant a $$f(S_t) = f(a) + f'(a) (S_t-a) + \int_a^{\infty}(S_t-x)^+f^{''}(x)dx + \int_{0}^a(x - S_t)^+f^{''}(x)dx.$$ Then, the payoff $1/S_t$ can be approximately hedged by call and put options: $$\frac{1}{S_t} = \frac{1}{a} -\frac{1}{a^2}(S_t-a)+ 2\bigg[\int_a^{\infty}\frac{(S_t-x)^+}{x^3}dx + \int_{0}^a\frac{(x - S_t)^+}{x^3}dx \bigg], $$ where $a = E(S_t)$.

As for $S_T/S_t$, let $d$ be the dividend paid at $t_1$, where $t<t_1<T$. Note that $$E(S_T \mid \mathcal{F}_t) =S_t \exp\Big(\int_t^T r_s ds \Big) - d\exp\Big(\int_{t_1}^T r_s ds \Big). $$ We replicate the payoff $1/S_t$ at time $t$. Then we replicate by forwards and bonds.

## Answer by emcor (score 1)

https://quant.stackexchange.com/a/14231

We can explicitly value the Inverted Option under Black-Scholes Model as follows:

Then the delta-hedging ratio is given as:

## Answer by Fab (score 0)

https://quant.stackexchange.com/a/14362

For question 2): At time $T$, we need to pay $100\cdot\frac{S_T}{S_t}$ (in domestic currency, say \$). To do this, we need to buy 100\$ worth of shares at time $t$: that gives us $N=100\cdot \frac{1}{S_t}$ shares, with the desired final value of $$N\cdot S_T = 100\cdot\frac{S_T}{S_t}$$ at expiry. Needless to say, today's PV of 100\$ at time $t$ is $100\,B(0,t)$.

However, then at time $t_1$, we hold $N$ shares, so we get a dividend of $d$ per share, so we receive $d\,N = 100\cdot d/S_t$. Being the nice investment bankers that we are, we charge the client correspondingly less, namely the PV of that, which is $100\,d$ times the answer to question 1.

Thus, final answer: $100\,(B(0,t) - d\cdot Q_1)$

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.