Black-Scholes assumes asset returns are lognormal → constant σ, no jumps, symmetric bell curve. If that were literally true, IV would be a flat horizontal line across all strikes.
But real markets violate this in three ways:
Fat tails — extreme moves happen more often than the normal distribution predicts. Options that pay off only in the tails (deep OTM) are therefore worth more than BS says → to match that higher price you need a higherσ → IV rises in the wings → smile.
Negative skew of stock returns — stocks crash down faster than they rally up. So the left tail is fatter than the right. Traders pay up for downside protection (OTM puts) → those low strikes get inflated IV → skew/smirk.
Crash fear / leverage effect — after 1987, dealers permanently price in disaster insurance on the downside. Falling prices raise leverage → raise volatility, creating an inbuilt correlation.
Derivation sketch of vega: Start from
c=SN(d1)−Ke−rTN(d2),d1=σTln(S/K)+(r+21σ2)T,d2=d1−σT.
Differentiate w.r.t. σ. The tricky cross-terms cancel because of the identity
Sϕ(d1)=Ke−rTϕ(d2),
leaving only the clean term Sϕ(d1)T. (The identity is what makes Greeks so tidy.)
Imagine a weather app that has to guess how wild tomorrow will be. Black-Scholes is a lazy app that says "every kind of day is equally wild." But people know storms (crashes) hit harder than sunny days, and they'll pay extra to be insured against a storm. So insurance for a bad crash-day costs more than the lazy app expects. When we plot how much each "day" costs, the storm side sticks up — that's the skew. If both a super-hot and super-cold day scare people, both sides stick up like a smile. The bumpy curve is just people telling us which disasters they're most afraid of.
The volatility σ that makes the Black-Scholes price equal the observed market price of the option.
Why is IV plotted against strike/moneyness not flat, as BS assumes?
Because real returns have fat tails and skew, so options at different strikes are worth more/less than lognormal predicts, needing different σ to match prices.
Why does equity skew slant so OTM puts have higher IV?
Crash fear / leverage effect / demand for downside insurance makes the left tail fatter, so puts are bid up → higher IV.
What guarantees a unique IV for each option price?
Positive vega (∂c/∂σ=Sϕ(d1)T>0), so price is strictly increasing in σ.
Formula for log-moneyness?
m=ln(K/F), with m=0 ATM, m<0 OTM puts, m>0 OTM calls.
Is a high-IV OTM put automatically overpriced?
No. High IV reflects the market's probability/crash view; richness is judged vs your own forecast of realized vol, not vs ATM.
What does the shape of the smile physically represent?
The market's implied probability distribution of the underlying — bumps show which price outcomes the market deems more likely/valuable than lognormal.
Why is Black-Scholes still used despite the smile?
It's a quoting/translation language (IV↔price per strike), not a claim of the true distribution.
What is the vega identity that simplifies its derivation?
Sϕ(d1)=Ke−rTϕ(d2), causing cross-terms to cancel and leaving ν=Sϕ(d1)T.
Dekho, Black-Scholes model ek assumption pe khada hai: har strike ke liye ek hi single volatility. Matlab agar tum IV ko strike ke against plot karo, toh ek flat straight line milni chahiye. Lekin real market mein aisa hota hi nahi. Jab tum actual option prices se IV nikaalte ho (backwards solve karke), toh curve bend ho jaata hai — kabhi ek slant (skew) ban jaata hai, kabhi ek U-shape (smile). Yeh curve hi bata deta hai ki market crash se kitna darta hai.
Kyun hota hai yeh? Kyunki stock returns lognormal nahi hote — unka left tail fat hota hai, matlab crash zyada tez aur zyada often aate hain rally ke muqable. Isliye log downside protection (OTM puts) ke liye extra paisa dete hain. Zyada price = zyada IV us strike pe. Equity index mein isliye left side upar uthta hai — yeh hai skew ya smirk. FX mein up aur down dono symmetric hote hain, toh dono taraf IV upar — yeh hai smile.
Important cheez: high IV ka matlab yeh nahi ki option overpriced hai. IV toh sirf market ki probability ki soch hai. Woh crash ka insurance fairly bhi priced ho sakta hai. Aur ek aur baat — vega positive hota hai (∂c/∂σ>0), iska matlab price hamesha σ ke saath badhta hai, isliye har price ke liye ek unique IV milta hai. Yehi reason hai ki hum smile curve draw kar bhi paate hain.
Yaad rakhne ka trick: "Stocks SKEW down in fear, FX SMILES both ways." Curve ki shape hi market ka fingerprint hai — jahan IV upar, wahan market zyada dara hua hai. Exam mein aur trading mein dono jagah yeh distribution ka intuition kaam aata hai.
