Convertible Bond Pricing: Black-Scholes, Binomial Trees, and the Greeks

Introduction

Convertible bond pricing sits at the intersection of fixed income and equity derivatives, and the structurers who build the pricing models are typically the most quantitatively sophisticated members of the equity-linked desk inside an investment bank. The economic intuition is straightforward: a convertible decomposes into a bond component (the present value of coupons and principal discounted at the issuer's credit spread) plus an embedded equity call option (struck at the conversion price and exercisable through the bond's life). The implementation is harder because the bond's call and put provisions create early-exercise features that standalone Black-Scholes cannot handle cleanly, which is why practitioners use Tsiveriotis-Fernandes binomial-tree models in production. ECM bankers rarely build the pricing models themselves, but they negotiate against them on every live deal: the structurer's vol input drives the conversion premium they can market, the Greeks dictate what the convertible arbitrage funds (roughly 70 percent of typical demand) will pay, and the cheapness print on the screen is what investors anchor to in the bookbuild call.

The economic intuition behind every convertible pricing model is the same two-component decomposition.

The Bond Component

The bond component is the present value of the convertible's fixed-income cash flows (coupons and principal at maturity) discounted at the issuer's straight-debt yield. If the issuer's 5-year senior unsecured non-convertible bond would yield 7 percent in the open market, the convertible's bond component values the $1,000 par bond's coupon stream and principal at that 7 percent rate. The bond component is the convertible's "bond floor": the theoretical price below which the convertible should not trade because investors would always prefer to hold for the coupon and principal payment. The bond floor is what gives the convertible its downside protection and is the principal reason convertibles outperform common equity in stock-down scenarios.

The Option Component

Layered on top of the bond component is an embedded call option on the issuer's stock, struck at the conversion price and exercisable through the bond's life. The option component captures the value of the conversion right: if the stock trades above the conversion price during the bond's life, the bondholder can exchange the bond for shares worth more than the bond's redemption value, and the option's value rises with the stock. The option's value at issuance depends primarily on the stock's implied volatility, the conversion premium (effectively the option's moneyness), the bond's maturity (time to expiration), and the dividend yield (which reduces the option's value because dividends paid before conversion belong to the existing shareholders). Black-Scholes is the standard closed-form benchmark for the embedded call:

where $S$ is the stock price, $K$ the conversion price, $T$ the time to maturity, $r$ the risk-free rate, $q$ the dividend yield, and $\sigma$ the implied volatility.

The Convertible's Total Value

The convertible's total fair value is the sum of the bond component and the option component, less an "option dilution" adjustment that accounts for the fact that conversion creates new shares:

At issuance, the convertible should trade at par; if the structurer's pricing produces a fair value materially above or below par, the term-sheet inputs (coupon, conversion premium, maturity) need to be adjusted to bring the model output back to par.

The intuitive bond-plus-option decomposition feels like it should let practitioners use Black-Scholes for the option component and a standard fixed-income model for the bond component. In practice, the embedded provisions break that simple decomposition.

The Issuer Call Cap

The issuer's soft-call provision lets the issuer force conversion when the stock has rallied past the soft-call trigger (typically 130 to 150 percent of the conversion price). The call cap means the option's upside is capped: even if the stock trades to $500 with a $130 conversion price, the issuer will call the bond after the soft-call trigger is met and investors will convert at the $130 strike rather than continuing to capture upside. Standalone Black-Scholes assumes infinite upside, which overstates the convertible's option value.

The Investor Put Floor

Investor put options at year 3 or 5 raise the bond's effective floor in scenarios where the issuer's credit deteriorates: investors can put back at par before maturity, capturing the full $1,000 rather than holding through credit deterioration. Standalone bond pricing models do not handle the put feature cleanly.

Path Dependence and Coupon Capitalization

Convertible coupons are paid through the bond's life, and conversion before a coupon date forfeits accrued interest. The decision to convert at any point is path-dependent: the optimal strategy depends on the entire stock-price history, not just the terminal stock price. Black-Scholes is closed-form for European options (exercise only at maturity) and does not handle path-dependent American-style early exercise without modification.

The Tsiveriotis-Fernandes Solution

The Tsiveriotis-Fernandes (TF) binomial-tree model addresses these issues by building a tree of stock-price paths, propagating the convertible's value backward from maturity, and at each node deciding whether the optimal action is to hold, convert, put, or accept an issuer call. The TF model splits the convertible's value at each node into a cash component (subject to credit risk) and an equity component (not subject to credit risk because conversion settles in stock), discounting each component at the appropriate rate:

where $V_{\text{cash}}$ is the portion subject to credit risk (discounted at the issuer's risky rate $r + s$) and $V_{\text{equity}}$ is the portion settled in stock (discounted at the risk-free rate $r$). The TF approach is the standard practitioner model for vanilla convertibles and the foundation of most production pricing systems on the bank's equity-linked desk.

Tsiveriotis-Fernandes (TF) Model

The standard practitioner binomial-tree model for pricing convertible bonds with credit risk, developed in 1998 by Kostas Tsiveriotis and Chris Fernandes. The model splits the convertible's value at each tree node into a cash component (subject to the issuer's credit spread) and an equity component (not subject to credit risk because conversion settles in stock), discounting each component at the appropriate rate. The TF approach is widely used because it handles path-dependent early-exercise features (issuer calls, investor puts, voluntary conversion) cleanly while incorporating credit risk in a tractable way.

The two inputs that dominate convertible pricing on the bank's structuring desk are implied volatility and credit spread.

Implied Volatility

The volatility input is typically taken from the listed-options market for the same issuer at a maturity matching (or close to) the convertible's maturity. The most common approach is to use the at-the-money (ATM) implied Black-Scholes volatility from listed options, adjusted for the convertible's longer maturity if necessary (listed options rarely trade past 2 to 3 years, while convertibles run 5 to 7 years). Issuers without liquid listed options force structurers to use historical realized volatility as a proxy, which introduces meaningful pricing uncertainty. The implied-volatility input is the most contested element of any convertible-pricing conversation because small changes in vol produce large changes in option value, and the marketing of the bond involves anchoring investors on a particular vol level.

Credit Spread

The credit-spread input is the issuer's straight-debt yield over the risk-free rate, used to discount the bond's cash flows. The spread is determined by the issuer's credit profile (cash flow, leverage, business profile) and is typically observed from the issuer's outstanding non-convertible bonds or estimated from comparable issuers. The credit spread input drives the bond floor, and for issuers without observable straight-debt prices, the structurer estimates the spread using rating-agency credit comparisons or CDS spreads if available.

The model's outputs, the Greeks, describe how the convertible's price changes as inputs move. Convertible arbitrage funds use the Greeks to construct delta-hedged positions and to manage risk through the bond's life.

Delta

Delta is the change in convertible fair value per dollar change in the underlying stock price. At issuance with a 30 percent conversion premium, delta is typically 0.4 to 0.7 depending on volatility and the stock's distance to the conversion price. As the stock rallies and approaches the conversion price, delta rises toward 1.0 (the convertible behaves more like equity); as the stock falls below the bond floor, delta approaches 0 (the convertible behaves like a straight bond).

Gamma

Gamma is the change in delta per dollar change in the stock. Gamma is highest when the stock is near the conversion price (the option is at-the-money), where small stock moves produce large changes in delta. The convertible's gamma is what convertible arbitrage funds capture through dynamic delta-hedging: as the stock moves, the arbitrageur re-hedges the short equity position at the new delta, and over time the rebalancing captures realized volatility above implied volatility.

Vega

Vega is the change in convertible fair value per 1 percentage point change in implied volatility. Convertibles have positive vega because higher vol increases the option's value. Vega is highest when the option is at-the-money and the bond has substantial time to maturity. Convertible arbitrage funds care about vega because their position is implicitly long volatility: they bought the bond's embedded option and short-sold the stock, leaving them with net long vol exposure.

Theta

Theta is the change in convertible fair value per day of time decay. Convertibles have negative theta (the option loses time value each day), partially offset by the bond's coupon accrual. Theta is highest when the option is near-the-money and time to maturity is shortest. The dynamic creates the convertible arb's principal cost: holding the bond means paying option time decay each day, which has to be earned back through realized vol or coupon and short rebate.

Rho

Rho is the change in convertible fair value per 1 basis point change in interest rates. Convertibles have negative rho because higher rates lower the bond floor (PV of fixed cash flows declines), partially offset by the option's positive rate sensitivity. The total rate sensitivity is typically modest and usually managed at the desk level rather than through individual position hedges.

The Greeks are formally the partial derivatives of the convertible's value $V$ with respect to each input:

The Greeks translate directly into the convertible arbitrage hedging strategy that dominates the convertible buyer base.

Constructing the Delta Hedge

The arb fund buys the convertible and shorts the underlying stock equal to the bond's delta times the conversion ratio:

For a $1 billion convertible with delta 0.5, the hedge involves shorting roughly $500 million of stock at issuance, neutralizing directional equity exposure between the long convertible and the short stock leg.

Capturing Gamma and Carry

The position is not perfectly hedged because the convertible has positive gamma: as the stock moves, the convertible's delta changes faster than the short stock leg adjusts. Arbitrage funds re-hedge dynamically, increasing the short as the stock rallies and reducing it as the stock falls. The dynamic re-hedging captures realized volatility, with each round-trip stock move generating a small profit:

The arb position also generates carry from the bond's coupon plus the rebate on the short stock position. If realized vol exceeds implied vol, the gamma capture plus carry exceeds the time decay (theta), and the position generates positive returns.

Cheapness Analysis

ECM bankers and convertible arbitrage funds evaluate new convertibles by comparing the bond's market price (offered price at issuance) to the model fair value at the same vol input. A convertible offered at par with model fair value above par at the proposed vol is "cheap" (attractive to investors); a convertible offered at par with model fair value below par at the proposed vol is "rich" (less attractive). The cheapness analysis informs both investor decisions to participate and the bank's negotiation with the issuer about the term sheet.

Convertible Cheapness

The difference between a convertible bond's market price and its theoretical fair value at the prevailing implied volatility input, expressed as a percentage of par or in basis points. A "cheap" convertible trades below its theoretical fair value (attractive to investors); a "rich" convertible trades above fair value (less attractive). Convertible arbitrage funds and outright fund managers use cheapness analysis to evaluate new issues and secondary trades, and the cheapness measure is one of the principal inputs to their participation decisions. ECM bankers also track cheapness across the market to advise issuers on whether the current environment supports favorable convertible terms.

To make the gamma capture concrete, walk through a realistic delta-hedged convertible position over a single week.

Setup

Arbitrage fund holds $10 million par of a convertible with delta 0.5, gamma 0.005 (per dollar stock move), and 50 percent implied volatility. Underlying stock is at $100. Initial hedge: short approximately 50,000 shares of stock (delta times conversion ratio across the position).

Day 1: Stock Rallies to $103

Convertible value rises by approximately delta times the conversion ratio times the stock move (plus a small half-gamma-times-stock-move-squared term), equaling roughly $15 per $1,000 par on the $3 stock move. Across $10 million par this is approximately $150,000 of convertible gain. The 50,000-share short loses $3 $\times 50{,}000 =$ $150,000. Net P&L before re-hedge: approximately zero (mostly delta-offset), with a small positive gamma contribution. Delta has now risen from 0.50 to approximately 0.515 (gamma effect); the fund increases short by 1,500 shares to maintain neutrality.

Day 2: Stock Falls to $98

Convertible value falls by less than the stock move (because delta is now lower); short stock gain exceeds convertible loss. The combined position generates a small profit equal to roughly half the gamma times the squared stock move (the asymmetric outcome from positive gamma). Delta falls back toward 0.50; fund reduces short by approximately 2,500 shares.

Cumulative Effect

Across many such daily moves, the position systematically captures gamma profit on each round-trip stock move. Annualized, the gamma capture roughly equals one-half times gamma times realized variance times the bond's notional. If realized vol exceeds implied vol, this gamma capture exceeds the theta cost; if realized vol falls short, theta dominates and the position loses money. Convertible arbitrage funds returned approximately +4 percent through May 2025, against the broader hedge fund average of +2.6 percent, with the gamma capture component of returns driving most of the outperformance.

The arbitrage P&L on a delta-hedged convertible position decomposes into specific components, each driven by a different Greek.

The Daily P&L Equation

Daily P&L approximately equals: gamma P&L (from realized stock moves squared) plus theta P&L (negative; daily time decay) plus vega P&L (from changes in implied vol) plus rho P&L (from rate moves) plus carry (coupon plus short-stock rebate, both positive). The components are tracked separately on the arbitrage desk's risk system, and the principal driver of returns over a typical horizon is the spread between realized and implied volatility.

Gamma vs. Theta Trade-Off

The position's structural trade-off is gamma capture versus theta decay. High-gamma convertibles (at-the-money, longer maturity) earn more gamma per unit of stock move but also cost more in theta. Arbitrage funds calibrate their portfolio to maximize the gamma-to-theta ratio across positions, favoring "sweet spot" converts (typically 30 to 60 percent moneyness with 3 to 5 years to maturity).

Vega Trades

Some arbitrage strategies are designed specifically to be long vega: the fund buys converts where implied vol appears low relative to expected future vol, holding the position until implied vol rebounds. Vega trades pay gamma capture along the way but are primarily betting on the implied-vol move rather than the realized-vs-implied spread.

Carry and Borrow Income

The arb position generates positive carry from the convertible's coupon (paid by the issuer) and the rebate the fund earns on short-stock proceeds (typically the prevailing short-rebate rate, currently 4 to 5 percent on most US listed equities). The carry income partially offsets the theta cost and is one of the structural reasons convertible arbitrage works in elevated-rate environments.

The model framework above produces a theoretical fair value, but the actual market price reflects supply, demand, and the specific buyer mix on a given deal. Equity-linked structurers calibrate their pricing to reflect three real-world factors that the underlying TF model does not capture directly.

The New-Issue Concession

Convertibles typically price at a small concession to model fair value at issuance to compensate investors for supply-absorption risk and incentivize participation. Concessions of 1 to 3 percent of par are routine; 4 to 6 percent for less liquid issuers or stressed windows. The concession produces the first-day trading profit arbitrage funds typically capture.

Buyer-Mix Sensitivity

Large arbitrage demand at the bookbuild lets the bank price at tighter coupons and higher premiums than outright-dominated deals would support. Structurers track indicative orders from each buyer category and adjust term-sheet inputs (coupon, premium, soft-call trigger) to match available demand.

Volatility Surface and Skew

The simple Black-Scholes input uses a single volatility number, but listed-options markets show volatility surfaces that vary by strike and maturity. Convertible structurers adjust the model input to reflect the specific portion of the volatility surface that the convertible's strike and maturity correspond to, and the adjustment can move the theoretical fair value by 100 to 200 basis points relative to a flat vol input. The volatility surface adjustment is one of the harder elements of convertible pricing and one of the main reasons real-deal pricing involves judgment beyond the model output.

A bond floor, an embedded call option, a Tsiveriotis-Fernandes tree handling the early-exercise features, and a Greek profile that drives the arbitrage hedge: those are the moving parts every convertible pricing model carries, and the framework every equity-linked structurer translates into a live term sheet. The structures up to this point have been optional convertibles where the holder chooses whether to exchange; a different structure forces conversion at maturity and reshapes the accounting and dilution profile materially. Mandatory convertibles and convertible preferred stock are next.