Deriving Exponential Stock Growth from a Deterministic Return Equation
Summary
The document derives the deterministic stock-price formula from the differential equation that results when uncertainty is removed from a proportional price process. Starting with the rate of change in price proportional to the current price, it writes the relation in differential form as the price change divided by price equaling the drift times time.
Integrating both sides from the initial time to the horizon gives the change in the natural logarithm of price as the drift multiplied by elapsed time. Exponentiating then yields the familiar exponential growth expression. The explanation identifies this as solving a first-order ordinary differential equation and connects infinitesimal percentage returns with log returns. Its scope is deliberately limited to deterministic growth; it does not address stochastic price dynamics or uncertainty.
Key ideas
- With deterministic proportional growth, the differential equation is dS divided by S equals drift times dt.
- Integrating the reciprocal-price differential produces a change in log price.
- Exponentiating the log-price equation gives the exponential stock-price path.
- The derivation is an ordinary differential equation solution for a deterministic process.
- Infinitesimal percentage returns equal infinitesimal log returns.
Tags
Full text
# Derivation of stock price formula John C. Hull 9th Ed p309
# Derivation of stock price formula John C. Hull 9th Ed p309
It says assuming a no-uncertainty Weiner process that models stock price: $$ \Delta S = \mu S\Delta t $$ Can be rearranged to (after taking the limit of $\Delta t \to 0$... $$ \frac{dS}{S}=\mu dt $$ Then integrating between time 0 and T to get: $$ S_T=S_0 e^{\mu T} $$
I don't understand the last step. Are they integrating with respect to t? How does the exponential come about when there was no exponential in the prior equation? Is this step a condensation of a complex calculation that they didn't show?
## Answer by Kevin (score 3, accepted)
https://quant.stackexchange.com/a/53254
It's simpler than you think. Hull is just solving an ODE.
You can naively put integral signs on both sides of the equation: \begin{align*} \frac{\mathrm{d}S_t}{S_t} &=\mu \mathrm{d}t \\ \implies \int_0^T\frac{\mathrm{d}S_t}{S_t} &=\int_0^T\mu \mathrm{d}t\\ \implies \ln(S_T)-\ln(S_0) &=\mu T \\ \implies S_T&=S_0e^{\mu T}. \end{align*}
Perhaps this makes it easier: Since $S_t$ is deterministic, it is a ``normal'' function. Thus, you may want to write $y(x)=S_t$. The above equation then turns into $\frac{\mathrm{d}y}{y} =\mu \mathrm{d}x\Leftrightarrow y'=\mu y$. So, it's simply about solving a first-order ODE.
Also note that $\frac{\mathrm{d}S_t}{S_t}=\mathrm{d}\ln(S_t)$, i.e. percentage returns correspond to log-returns if time is infinitesimal. So, you shouldn't be surprised to find an exponential here.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.