Asymptotic Behavior of the Black-Scholes Formula¶

The Black-Scholes formula exhibits well-defined limiting behavior as parameters approach extreme values. These limits give intuition for option behavior across market regimes and serve as sanity checks for numerical implementations.

Limits in Stock Price¶

Section goal: what happens to call and put prices as the spot drifts to \(0\) and to \(\infty\).

1. As S → ∞ (Deep In-the-Money Call)¶

Behavior of \(d_1\) and \(d_2\):

Cumulative probabilities:

Call price:

Interpretation: Deep ITM call behaves like forward contract with guaranteed exercise. The option value equals the stock price minus the present value of the strike.

Put price (via put-call parity): \(P = C - S + Ke^{-rT} \to 0\) — the deep OTM put is worthless.

2. As S → 0 (Deep Out-of-the-Money Call)¶

Now \(d_1, d_2 \to -\infty\), so \(\mathcal{N}(d_1), \mathcal{N}(d_2) \to 0\) and \(C \to 0\). For the put, \(\mathcal{N}(-d_2) \to 1\) gives \(P \to Ke^{-rT}\): the deep ITM put approaches the present value of the strike.

Limits in Time to Maturity¶

Section goal: option value at the expiry boundary (\(T \to 0\)) and in the long-dated limit (\(T \to \infty\)).

1. As T → 0 (Approaching Expiration)¶

Case 1: \(S > K\) (ITM call)

As \(T \to 0\):

Case 2: \(S < K\) (OTM call)

As \(T \to 0\):

Case 3: \(S = K\) (ATM call)

Both \(\mathcal{N}(d_1)\) and \(\mathcal{N}(d_2)\) tend to \(\tfrac{1}{2}\), and \(S\,\mathcal{N}(d_1) - Ke^{-rT}\,\mathcal{N}(d_2)\) becomes a \(\tfrac{K}{2} - \tfrac{K}{2}\) near-cancellation. To see the limit cleanly, substitute the Taylor expansion \(\mathcal{N}(\epsilon) = \tfrac{1}{2} + \frac{\epsilon}{\sqrt{2\pi}} + O(\epsilon^3)\) with \(\epsilon_i = d_i\). With \(S = K\):

The call price becomes

Each bracketed term is \(O(T^{1/2})\) as \(T \to 0\) — the first via \(1 - e^{-rT} = rT + O(T^2)\), the second via \(d_i = O(\sqrt{T})\) — so

The leading \(K\sigma\sqrt{T}/\sqrt{2\pi}\) term is exactly the ATM approximation derived later in this section.

Summary: As expiration approaches:

This recovers the terminal payoff condition.

2. As T → ∞ (Very Long Maturity, r > 0)¶

Assumption: \(r > 0\) throughout this subsection. The case \(r = 0\) is qualitatively different (see Exercise 5).

Decompose

The first term decays as \(T^{-1/2}\), the second grows as \(T^{1/2}\) with positive coefficient (since \(r + \tfrac{1}{2}\sigma^2 > 0\)); the second term dominates strictly. So \(d_1, d_2 \to +\infty\) at rate \(\sqrt{T}\), and \(\mathcal{N}(d_1), \mathcal{N}(d_2) \to 1\).

Call price: Both terms in \(C = S\mathcal{N}(d_1) - Ke^{-rT}\mathcal{N}(d_2)\) converge: \(S\mathcal{N}(d_1) \to S\) and the strike term vanishes by exponential decay — \(Ke^{-rT} \to 0\) exponentially while \(\mathcal{N}(d_2)\) is bounded. Therefore

The discounted strike's exponential decay always dominates any sub-exponential factor in \(\mathcal{N}(d_2)\), which is why this limit is unconditional in \(\sigma\) (given \(r > 0\)).

Put price: \(P = C - S + Ke^{-rT} \to S - S + 0 = 0\).

Limits in Volatility¶

Section goal: behaviour as \(\sigma \to 0\) (deterministic stock) and \(\sigma \to \infty\) (extreme uncertainty).

1. As σ → 0 (Zero Volatility)¶

Deterministic evolution: Stock grows at risk-free rate:

Case 1: \(S_0 > Ke^{-rT}\) (Forward price above strike)

As \(\sigma \to 0\), \(d_1, d_2 \to +\infty\) so \(\mathcal{N}(d_1), \mathcal{N}(d_2) \to 1\). The stock finishes above \(K\) with probability approaching \(1\):

Case 2: \(S_0 < Ke^{-rT}\) (Forward price below strike)

Now \(d_1, d_2 \to -\infty\), so \(\mathcal{N}(d_1), \mathcal{N}(d_2) \to 0\) and

Case 3: \(S_0 = Ke^{-rT}\) (Exactly ATM forward)

Summary:

This is the forward value with no volatility premium.

Put limit:

2. As σ → ∞ (Infinite Volatility)¶

Behavior of \(d_1\) and \(d_2\): Decompose

The first piece decays as \(\sigma^{-1}\), the \(\pm\sigma\sqrt{T}/2\) pieces grow linearly in \(\sigma\). So linear growth dominates inverse decay, with opposite signs:

Therefore \(\mathcal{N}(d_1) \to 1\) and \(\mathcal{N}(d_2) \to 0\). (The "indeterminate \(\infty - \infty\)" form for \(d_2 = d_1 - \sigma\sqrt{T}\) resolves as soon as one writes both \(d_1\) and \(d_2\) in terms of the same dominant scale \(\sigma\sqrt{T}\) — they have opposite limits, not a cancellation.)

Call price:

Interpretation: \(\mathcal{N}(d_1) \to 1\) and \(\mathcal{N}(d_2) \to 0\), so as in the deep-ITM case the call approaches the full stock price.

Put price: Put-call parity gives \(P = C - S + Ke^{-rT} \to Ke^{-rT}\).

Summary:

Limits in Interest Rate¶

Section goal: how prices respond as \(r \to 0\) and as \(r \to \infty\).

1. As r → 0 (Zero Interest Rate)¶

Call formula:

where:

Put-call parity becomes:

(no discounting since \(e^{-r \cdot 0} = 1\))

Interpretation: Options behave as if cash flows are not discounted.

2. As r → ∞ (Infinite Interest Rate)¶

As \(r \to \infty\), \(d_1, d_2 \to +\infty\) (driven by the \(r\sqrt{T}/\sigma\) term) and \(e^{-rT} \to 0\). The discounted strike vanishes (cf. \(T \to \infty\)), giving \(C \to S\) and \(P \to 0\).

Limits in Strike Price¶

Section goal: limits as \(K \to 0\) (call \(\to S\)) and \(K \to \infty\) (call \(\to 0\)).

1. As K → 0 (Zero Strike)¶

As \(K \to 0\), \(d_1, d_2 \to +\infty\), so \(C \to S\) and \(P \to 0\) — a zero-strike call is equivalent to owning the stock.

2. As K → ∞ (Infinite Strike)¶

As \(K \to \infty\), \(d_1, d_2 \to -\infty\), so \(\mathcal{N}(d_1), \mathcal{N}(d_2) \to 0\) and \(C \to 0\) (no chance of exceeding the strike). The put price obeys \(P = Ke^{-rT}\mathcal{N}(-d_2) - S\mathcal{N}(-d_1)\) with \(\mathcal{N}(-d_1), \mathcal{N}(-d_2) \to 1\), so \(P \sim Ke^{-rT} - S \to +\infty\). The general upper bound \(P \leq Ke^{-rT}\) remains valid — both sides diverge at the same rate, so \(P/(Ke^{-rT}) \to 1\), i.e., the put approaches its theoretical maximum value relative to the discounted strike.

ATM Approximation for Small Time¶

Section goal: the clean short-time expansion \(C_{\text{ATM}} \approx 0.4\, S\sigma\sqrt{T}\) and its derivation.

For at-the-money options (\(S \approx K\)) with small \(T\), a useful closed-form approximation emerges.

1. ATM Call Formula¶

When \(S = K\) and \(T\) is small:

Using Taylor expansion of \(\mathcal{N}(x)\) near \(x = 0\):

After substitution and simplification:

This is the ATM volatility rule of thumb: ATM option value is roughly 40% of \(S\sigma\sqrt{T}\).

Example: \(S = 100\), \(\sigma = 20\%\), \(T = 0.25\) (3 months):

Summary Table¶

Practical Implications¶

Recall (see § Properties and Bounds and § Common Numerical Issues): the bounds \(0 \leq C \leq S\), \(0 \leq P \leq Ke^{-rT}\) provide sanity checks, and the limiting formulas above motivate the special-case branches used near \(T \approx 0\), \(\sigma \to 0\), or \(|d_i|\) large in production implementations. For the Greek-level limits (\(\Delta \to 1\) deep ITM, \(\Gamma \to \infty\) as \(T \to 0\) near the strike — pin risk — and \(\nu \to 0\) as \(\sigma \to 0\)), see § Greeks in Black-Scholes. Real-market violations of these asymptotics signal model misspecification, transaction-cost frictions, or American features.

Summary¶

The five regimes — deep ITM (forward-like), near expiration (intrinsic), long maturity (call \(\to S\)), zero/infinite \(\sigma\) — are summarised in the table above and serve as bounds, sanity checks, and stability triggers in implementations.

Exercises¶

Exercise 1. Let \(C(S, K, T, r, \sigma)\) denote the Black-Scholes call price. Prove rigorously that

by analyzing the rate at which \(\mathcal{N}(d_1) \to 1\) as \(S \to \infty\). What does this tell you about the delta of a deep in-the-money call?

Exercise 2. For an at-the-money call (\(S = K\)) with \(r = 0\), use the Taylor expansion of \(\mathcal{N}(x)\) around \(x = 0\) to show that the call price satisfies

for small \(T\). Compute the percentage error of this approximation relative to the exact Black-Scholes price when \(S = K = 100\), \(\sigma = 0.25\), \(r = 0\), and \(T = 0.01\).

Exercise 3. Consider the put price \(P\) as \(\sigma \to \infty\). Show from the Black-Scholes formula (not from put-call parity) that \(P \to Ke^{-rT}\). Explain the financial intuition: why should the put attain its maximum theoretical value when volatility is infinite?

Exercise 4. Examine the behavior of the Black-Scholes call price as \(r \to -\infty\) (negative interest rates). Determine \(\lim_{r \to -\infty} d_1\), \(\lim_{r \to -\infty} d_2\), and \(\lim_{r \to -\infty} C\). Is the result consistent with the economic interpretation of a very negative interest rate?

Exercise 5. The summary table states that as \(T \to \infty\), the put price satisfies \(P \to 0\). However, this assumes \(r > 0\). What happens to \(\lim_{T \to \infty} P\) when \(r = 0\)? Prove your answer by carefully analyzing \(d_1\) and \(d_2\) in this special case.

Exercise 6. Using the ATM approximation \(C_{\text{ATM}} \approx 0.4 \, S \sigma \sqrt{T}\), derive an approximate formula for the ATM implied volatility \(\sigma_{\text{impl}}\) given a market price \(C_{\text{mkt}}\) of an at-the-money call. Apply your formula to find the approximate implied volatility when \(S = 50\), \(T = 0.25\), and \(C_{\text{mkt}} = 2.00\).

Exercise 7. Consider the ratio of a European put to a European call, \(P/C\), for a fixed strike \(K\) and maturity \(T\). Determine

for the three cases \(S > Ke^{-rT}\), \(S < Ke^{-rT}\), and \(S = Ke^{-rT}\). Discuss why the \(S = Ke^{-rT}\) case requires careful treatment.