Understand Black-Scholes intuition — Stock-Market notes

WHY does Black-Scholes exist?

WHAT problem does it solve? Before 1973, nobody had a non-arbitrage price for an option. People guessed. Black, Scholes and Merton showed that an option can be perfectly replicated by continuously holding some stock plus some cash. If you can replicate it, the option's price is forced — otherwise there's free money (arbitrage).

WHY does this matter? It converts a gambling question ("will the stock go up?") into an engineering question ("what portfolio copies this payoff?"). The direction/drift of the stock drops out — a shocking result.

Building the formula from scratch

Step 1 — Model the stock as random walk (GBM)

Assume the stock follows Geometric Brownian Motion:

$dS = \mu S\,dt + \sigma S\,dW$

Why this step? Returns (not prices) are what compound multiplicatively, so we model $\frac{dS}{S}$ . The $\mu S\,dt$ term is steady growth; $\sigma S\,dW$ is random shocks scaled by price. This guarantees $S>0$ always.

Step 2 — Build a risk-free hedge (delta hedging)

Hold a portfolio: long 1 option $V(S,t)$ , short $\Delta$ shares:

$\Pi = V - \Delta S$

Change over $dt$ using Itô's lemma (the random-walk chain rule):

$dV = \left(\frac{\partial V}{\partial t} + \tfrac12\sigma^2 S^2\frac{\partial^2 V}{\partial S^2}\right)dt + \frac{\partial V}{\partial S}\,dS$

Why the extra $\tfrac12\sigma^2 S^2 V_{SS}$ term? Because for a random walk, $(dW)^2 = dt$ is not negligible — variance accumulates linearly in time. This is the whole soul of the formula.

Choose $\Delta = \frac{\partial V}{\partial S}$ . Then the random $dS$ terms cancel:

$d\Pi = \left(\frac{\partial V}{\partial t} + \tfrac12\sigma^2 S^2\frac{\partial^2 V}{\partial S^2}\right)dt$

Why this step? We killed randomness. A riskless portfolio must earn the risk-free rate $r$ , or arbitrage exists:

$d\Pi = r\Pi\,dt = r(V - \Delta S)\,dt$

Step 3 — The Black-Scholes PDE

Equate both expressions:

Why beautiful? Notice $\mu$ (real drift) is gone. Only $r$ and $\sigma$ remain.

Step 4 — Solve for a European call

With boundary condition $V(S,T)=\max(S-K,0)$ , solving the PDE gives:

Figure — Understand Black-Scholes intuition

Worked examples

Worked example Example 1 — At-the-money, short dated

$S=100$ , $K=100$ , $r=0$ , $\sigma=0.20$ , $T=0.25$ (3 months).

Step 1: $\sigma\sqrt{T} = 0.20\times0.5 = 0.10$ . Why? This is the total volatility over the option's life. Step 2: $d_1 = \frac{\ln 1 + (0 + 0.02)(0.25)}{0.10} = \frac{0.005}{0.10}=0.05$ . Why? $\ln(S/K)=0$ since ATM. Step 3: $d_2 = 0.05 - 0.10 = -0.05$ . Step 4: $N(0.05)\approx0.5199$ , $N(-0.05)\approx0.4801$ . Step 5: $C = 100(0.5199) - 100(0.4801) = 3.98$ .

Sanity approximation: ATM call $\approx 0.4\,S\,\sigma\sqrt{T} = 0.4(100)(0.10)=4.0$ . ✓ Matches!

Common mistakes

Forecast-then-Verify

Flashcards

What does Black-Scholes fundamentally assume the stock follows?

Geometric Brownian Motion,

dS=\mu S\,dt+\sigma S\,dW

(lognormal prices, normal log-returns).

Why does the real drift

\mu

disappear from the option price?

Delta hedging cancels stock exposure; a riskless portfolio must earn

r

, so pricing is under the risk-neutral measure where drift

=r

What is the delta hedge ratio

\Delta

\Delta = \partial V/\partial S = N(d_1)

for a call.

What does the term

\tfrac12\sigma^2 S^2 V_{SS}

come from?

Itô's lemma;

(dW)^2=dt

, so variance is not negligible over

dt

Interpret

N(d_2)

Risk-neutral probability the call finishes in-the-money (exercise happens).

Write the Black-Scholes call formula.

C=S N(d_1)-Ke^{-rT}N(d_2)

Give the quick ATM call approximation.

C\approx 0.4\,S\,\sigma\sqrt{T}

\sigma\to0

with

S>K

, what is a call worth?

\max(S-Ke^{-rT},0)

— just discounted intrinsic value.

What is the upper bound on a call price and why?

C\le S

; a call to buy the stock can't beat owning the stock outright.

State the Black-Scholes PDE.

V_t+\tfrac12\sigma^2 S^2 V_{SS}+rSV_S-rV=0

Recall Feynman: explain to a 12-year-old

Imagine a coupon that lets you buy a toy for ₹100 next month. If the toy will cost ₹150, your coupon is worth ₹50. If it'll cost only ₹80, you just throw the coupon away — it's worth ₹0, you never lose. So the coupon can only help you. Black-Scholes is a clever recipe that says: "Tell me how bouncy the toy's price is, how long till the coupon expires, and the bank's interest rate — and I'll tell you the fair price of the coupon." The trick: a shopkeeper can copy your coupon by keeping some toys and some cash and adjusting them daily. Because he can copy it, the price is fixed — no cheating allowed. And funnily, it doesn't matter whether you think the toy will get pricier; only how jumpy the price is.

Connections

Geometric Brownian Motion — the stock model underneath.
Itô's Lemma — where the $\tfrac12\sigma^2S^2V_{SS}$ term is born.
Delta Hedging — the replication trick that fixes the price.
Risk-Neutral Valuation — why $\mu\to r$ .
The Greeks — Delta $=N(d_1)$ , Gamma, Vega, Theta all come from differentiating $C$ .
Put-Call Parity — links the call formula to puts: $C-P=S-Ke^{-rT}$ .
Implied Volatility — invert Black-Scholes to back out $\sigma$ from market prices.

Concept Map

Hinglish (regional understanding)

Intuition Hinglish mein samjho

Dekho, option ek aisa contract hai jo tumhe right deta hai stock kharidne ka fixed price (strike) par — obligation nahi. Toh agar stock upar gaya toh profit, aur neeche gaya toh tum bas option chhod dete ho, loss zero. Yeh "upside milega, downside capped" wali property ko hi hum optionality kehte hain, aur Black-Scholes iski exact keemat batata hai.

Sabse magical baat: option ki price mein stock ka expected return ( $\mu$ ) matter hi nahi karta! Kyun? Kyunki ek trader delta-hedge karke apna directional risk hata sakta hai — thoda stock aur thoda cash rakh ke option ko copy kar leta hai. Jab copy ho sakta hai, toh arbitrage se bachne ke liye price fix ho jaati hai. Isliye sirf volatility ( $\sigma$ ), time ( $T$ ), interest rate ( $r$ ), aur strike ( $K$ ) matter karte hain. Yaad rakho: "drift dies, vol survives."

Formula ko English sentence ki tarah padho: $C = S\,N(d_1) - Ke^{-rT}N(d_2)$ . Yahan $N(d_2)$ matlab option ke ITM finish hone ki (risk-neutral) probability, $Ke^{-rT}$ matlab strike jo future mein doge usko aaj ki value mein discount kiya, aur $S N(d_1)$ matlab stock milne ki expected value. Simple: jo milta hai minus jo dete ho.

Ek quick trick regional exam ke liye — ATM option ke liye $C \approx 0.4 \times S \times \sigma\sqrt{T}$ . Isse tum seconds mein rough price bata sakte ho. Aur zyada volatility = zyada option value, kyunki randomness sirf upside badhata hai (downside toh already zero pe protected hai). Yehi Black-Scholes ka core intuition hai.

Test yourself — The Greeks

Connections