Derivatives Course
CMS Options & Swaption-Based Hedging
Price CMS products with convexity adjustments, calibrate SABR, and build a swaption strip hedge.
Enrol now
- Lifetime access — all lessons & updates
- 10 lessons across 2 structured parts
- Fully worked Jupyter notebooks
- Market datasets + published benchmark prices
- Free lesson previews before you buy
- 14-day money-back guarantee
Secured by a no-questions-asked 14-day refund.
Overview
Introduction
Constant Maturity Swap products sit at the intersection of rates modelling and vol trading: the CMS rate is not a martingale under its natural payment measure, so a convexity adjustment is needed to price any CMS-linked coupon correctly. Getting that adjustment right requires both a model for the swaption smile and a static replication argument that connects CMS convexity to the full strip of swaption prices.
This course covers the two dominant industry approaches — Hagan's static replication formula and the linear Terminal Swap Rate approximation — and pairs them with the SABR stochastic volatility model to supply the swaption smile inputs. You calibrate SABR to a real swaption vol cube, compute CMS convexity adjustments, price a CMS spread option via Kirk's approximation, and build the swaption strip hedge that replicates the CMS payoff.
By the end you will be able to explain exactly why a CMS cap is more expensive than a co-terminal vanilla cap, compute the hedge notionals from first principles, and analyse the vega risk and hedge cost across tenors. These skills are directly applicable to any role involving structured rates products, exotic payoffs or vol-surface arbitrage.
Hands-on
The project you'll build
The project starts with SABR calibration. You are given a swaption vol cube — implied vols across expiries, tenors and strikes — and you calibrate the SABR parameters (α, β, ρ, ν) for each expiry-tenor pair by minimising the sum of squared errors against the Hagan et al. approximate vol formula. The calibrated surface becomes the input for all downstream calculations.
With the vol surface calibrated, you compute CMS convexity adjustments for a range of CMS tenors using both the Hagan static replication formula (numerical integration of g''(K) × swaption prices over the strike grid) and the linear TSR approximation (closed-form). You compare the two approaches and quantify the difference in adjustments across the curve.
You then price a CMS spread option on the 10Y−2Y swap rate differential using Kirk's approximation, construct the swaption strip hedge for the CMS cap by computing the replicating notionals N(K) = g''(K), cost the hedge portfolio, and compute delta and vega Greeks for the full CMS position — giving you a complete view of the product's risk profile.
Outcomes
What you'll learn
CMS convexity adjustment problem
Explain why the CMS rate is not a martingale under the payment measure and why a convexity adjustment is required for any CMS-linked coupon.
Hagan static replication
Implement the Hagan formula expressing the CMS convexity adjustment as a weighted integral over the full swaption smile.
Linear TSR approximation
Derive and implement the linear terminal swap rate approximation for a tractable closed-form convexity adjustment.
SABR calibration
Calibrate SABR parameters α, β, ρ, ν to a swaption vol cube using the Hagan et al. approximate implied vol formula.
CMS spread option pricing
Price a CMS spread option using Kirk's approximation and understand the model assumptions embedded in the formula.
Swaption strip hedge
Compute the replicating swaption notionals N(K) = g''(K) across the strike grid and cost the hedge portfolio.
Delta and vega Greeks
Compute numerical Greeks for CMS products with respect to the swap rate and the vol surface, and interpret the hedge implications.
Hedge cost analysis
Analyse the cost of the static replication hedge across tenors and discuss residual exposures and practical implementation constraints.
Stack
Tools & technology
The course teaches a workflow, not just a toolset — here is what we use and what you could swap in.
| Tool | Used for | Alternatives |
|---|---|---|
| Python 3 | SABR calibration, CMS convexity computation and hedge construction | Julia, MATLAB, QuantLib Python |
| NumPy & SciPy | Numerical integration of the Hagan formula, root-finding and optimisation for SABR fit | Gaussian quadrature from scratch, MATLAB fmincon |
| SABR Model | Parameterising the swaption smile for use in static replication and spread option pricing | SVI, local volatility, normal SABR |
| Kirk's Approximation | Closed-form pricing of the CMS spread option under log-normal dynamics | Margrabe's formula, Monte Carlo, copula methods |
| Jupyter | Interactive vol surface visualisation and convexity adjustment diagnostics | VS Code, plain Python scripts |
Syllabus
Course curriculum
10 lessons across 2 parts. Lessons marked Free preview can be read before you enrol.
Part 1 — CMS Convexity & SABRFrom the change-of-numeraire problem to SABR vol calibration.
- 1.CMS products and the convexity adjustmentWhy the CMS rate is not a martingale under the payment measure and what to do about it.Free preview
- 2.Hagan static replication formulaExpress the convexity adjustment as an integral over swaption prices.Free preview
- 3.Linear TSR approximationApproximate the change-of-numeraire density linearly for a tractable CA formula.
- 4.SABR stochastic volatility modelImplement the Hagan et al. approximate implied vol formula.
- 5.Calibrating SABR to swaption smilesFit alpha, beta, rho and nu to market swaption implied vols.
Part 2 — Spread Options & HedgingCMS spread options, strip hedging and delta/vega.
- 1.CMS spread option via Kirk's approximationPrice a CMS spread caplet using the approximate spread vol formula.
- 2.Building the swaption strip hedgeCompute hedge notionals N(K) = g''(K) across the strike grid.
- 3.Hedge cost and residual vegaCompute the cost of the replicating swaption portfolio and residual exposure.
- 4.Delta and vega Greeks for CMS productsNumerical Greeks with respect to swap rate and vol surface.
- 5.Conclusion: from CMS to exoticsHow convexity adjustments appear in TARNs, range accruals and other exotics.
Before you start
Prerequisites
- Solid understanding of swap and swaption mechanics — par swap rate, swaption payoff, PVBP.
- Familiarity with the Black-76 model for swaption pricing and implied volatility conventions.
- Working Python and NumPy; numerical integration with scipy.integrate and optimisation with scipy.optimize.
- The Short Rate Models or HJM course is useful but not required — the change-of-numeraire intuition is explained from scratch.
Fit
Who this course is for
Rates derivatives quants
You price vanilla swaptions and caps and want to move into CMS products, structured notes and exotic payoffs that involve the full smile.
Vol traders and structurers
You trade swaption vol and want to understand how CMS convexity ties together the entire vol surface into a hedgeable exposure.
Fixed income researchers
You study interest rate modelling and want a rigorous, code-first treatment of SABR and CMS convexity that goes beyond the original papers.
Risk managers for rates exotics
You manage Greeks on a structured rates book and want to understand the model risk embedded in SABR-calibrated convexity adjustments.
Faculty
Your instructor
QuantIndex Faculty
QuantIndex Faculty are practising derivatives quants with direct experience pricing and hedging CMS products, calibrating SABR surfaces and managing vega risk on structured rates books. This course is written by quants who implement these models in production and is reviewed for both theoretical rigour and practical relevance before each release.
Benchmarked & verified
Every price you compute is checked against published results from leading textbooks and market data. You do not finish the course hoping your model is right — you finish it knowing.
Questions
Frequently asked questions
What exactly is a convexity adjustment and why can I not just ignore it?
The convexity adjustment accounts for the fact that the CMS rate, when evaluated under the payment date's measure rather than the natural swap measure, has a non-zero drift. For short tenors and short observation lags the adjustment is small, but for long-tenor CMS products — 10Y or 30Y swap rate as an index — it can be 20–50 basis points, which is too large to ignore in any pricing context.
Do I need to understand measure theory to follow the course?
No. The change-of-numeraire argument is explained intuitively — you are changing which asset you use as the unit of account, and this changes the drift of the swap rate. The key result is the static replication formula, which is derived from first principles without requiring Girsanov's theorem or Radon-Nikodym derivatives.
Is SABR still the industry standard for swaption smiles?
SABR remains the dominant quoting and interpolation model for swaption smiles in most rate currencies, particularly because the Hagan et al. formula provides fast, differentiable implied vols that are easy to calibrate. More sophisticated models (free-boundary SABR, stochastic normal vol) are used for low-rate environments, and the course discusses these limitations.
How does Kirk's approximation compare to a full Monte Carlo for the spread option?
Kirk's approximation is accurate for spread options where the two underlying rates are highly correlated and the tenors are not too far apart, which covers most practical CMS spread cases (10Y−2Y, 30Y−10Y). For exotic structures where the correlation is uncertain or the distribution is far from log-normal, a Monte Carlo with correlated SABR dynamics is more appropriate.
Is the swaption strip hedge practical to implement in a real book?
The static replication hedge is the theoretical ideal; in practice, the hedge is truncated to a finite number of liquid strikes and rebalanced as the vol surface moves. The course discusses the residual vega from truncation and the practical minimum-strike constraint imposed by negative rates and market liquidity.
Is there a refund policy?
Yes — a 14-day, no-questions-asked money-back guarantee. If the course is not right for you, email us and we will refund you in full.
Ready to start?
- Lifetime access — all lessons & updates
- 10 lessons across 2 structured parts
- Fully worked Jupyter notebooks
- Market datasets + published benchmark prices
- Free lesson previews before you buy
- 14-day money-back guarantee
Secured by a no-questions-asked 14-day refund.