Curriculum Part 6 Chapters 35–36

Structured Products

Everything here is decomposition: a structured product is a bundle of simpler claims, and the exam wants you to unbundle it, price the pieces, and say who is short which option.

Term structure modelling

Two families, and the distinction between them is the most reliably tested idea in this part.

Equilibrium models

Specify a process for the short rate from economic assumptions and derive the whole curve. Vasicek: dr = κ(θ − r)dt + σ dz — mean reversion to θ at speed κ with constant volatility, closed-form bond prices, but rates can go negative. Cox–Ingersoll–Ross: volatility scaled by √r, so volatility falls as rates approach zero and rates stay non-negative.

Strength: economic interpretation and parsimony. Weakness: with a handful of parameters the model will not reprice every observed bond, so it cannot be used for arbitrage-free relative-value pricing of derivatives against the current curve. Both are single-factor, so all rates move together — they cannot generate independent curve twists.

Arbitrage-free (no-arbitrage) models

Take today's observed curve as given and fit the model to it exactly, then price derivatives consistently with it. Ho–Lee adds a time-dependent drift to a normal short rate; Hull–White is essentially Vasicek with a time-dependent drift so the initial curve is matched; Black–Derman–Toy works in a binomial lattice on the log of the rate, keeping rates positive and calibrating to both the curve and volatilities.

Strength: exact consistency with market prices, which is a requirement for pricing and hedging interest-rate options. Weakness: calibrated parameters have no economic content, they must be re-fitted continually, and a model that fits today perfectly may imply implausible future dynamics.

How to choose in an answer: equilibrium models for economic scenario generation, long-horizon asset-liability projection and understanding drivers; arbitrage-free models for pricing and hedging traded derivatives against the current curve. Saying which purpose the model serves is what earns the mark.

Interest rate derivatives

Interest rate swap
Exchange fixed for floating on a notional. Economically a portfolio of forward rate agreements, or equivalently long one bond and short another; value at inception is zero and moves with the curve. The swap is the basic building block for converting fixed exposure to floating and for hedging liability duration.
Caps and floors
A cap is a series of options (caplets) on a reference rate, paying when the rate exceeds the strike — insurance for a floating-rate borrower. A floor is the mirror image, protecting a floating-rate lender. A collar combines a bought cap and a sold floor to cheapen the protection.
Swaption
An option to enter a swap. A payer swaption (right to pay fixed) gains when rates rise; a receiver swaption gains when rates fall. Embedded in callable and putable bonds — a callable bond holder has effectively sold a receiver swaption.
Forward rate agreement
Locks a single future interest rate; the elementary component of both swaps and caps.
Structured coupon notes
Inverse floaters, range accruals, callable step-ups: all are combinations of a bond, swaps and options. Unbundle, price each leg, and identify what the investor has sold.

The consistent exam move is to convert a structure into the bond-plus-options equivalent and then ask about convexity and negative convexity. A callable bond's upside is capped because the issuer calls when rates fall — negative convexity — which is why its price appreciation lags a comparable non-callable bond in a rally and why its yield is higher.

Asset-backed securities

The structure. Assets are sold to a bankruptcy-remote special purpose vehicle, which issues tranches of differing seniority. Credit enhancement comes from subordination, overcollateralisation, excess spread, and reserve accounts; the resulting waterfall pays interest and principal by seniority. The economic point is tranching: the same pool of assets is divided into claims with very different risk — a senior tranche is protected by everything below it, an equity or residual tranche absorbs first losses and is highly leveraged to the pool's loss rate.

Correlation is the hidden variable. For a given expected pool loss, higher default correlation makes extreme outcomes more likely, which harms senior tranches (they can now be reached) and can help equity tranches (a chance of very few defaults). Investors in a senior tranche are therefore implicitly short correlation — the lesson learned expensively in 2007–08, and a favourite examinable subtlety.

Auto loan ABS
Amortising collateral, short weighted-average life, predictable prepayment (borrowers rarely refinance small car loans for rate reasons), granular pools, and a strong empirical loss record. Risks: obligor credit quality by tier (prime through subprime), used-vehicle recovery values, and servicer quality.
Credit card ABS
Non-amortising revolving collateral, so structures use a revolving period followed by an amortisation period. Performance is monitored by monthly payment rate, yield, charge-offs and excess spread; deterioration in excess spread triggers early amortisation, which repays investors ahead of schedule — protection for them, a funding shock for the sponsor.

Prepayment and extension risk generalise: faster prepayment shortens the life and returns cash when reinvestment rates are low; slower prepayment extends it. That is why average life, not stated maturity, is the relevant horizon for any amortising structure, and why an ABS investor's yield is model-dependent.

Insurance-linked securities: catastrophe bonds

The structure. A sponsor (an insurer or reinsurer) transfers catastrophe risk to capital markets through a special purpose vehicle that issues notes and holds the proceeds as collateral. Investors earn a spread over the collateral yield; if a defined event triggers the bond, principal is reduced to pay the sponsor's losses. The attachment point is the loss level at which investors begin to lose principal and the exhaustion point is where they lose all of it; the modelled probability of reaching attachment is the basis for pricing, expressed as a multiple of expected loss.

Four trigger types — the most examinable list in the chapter, because each shifts basis risk and moral hazard between the parties:

Indemnity
Based on the sponsor's actual claims paid. No basis risk for the sponsor, which is why sponsors prefer it — but investors must wait for claims to settle, and the sponsor has an incentive to underwrite loosely because its losses are hedged. That moral hazard means investors demand extra return for indemnity triggers.
Industry loss
Based on an independent third party's estimate of total industry losses, applied to the sponsor's market share. Settles faster, and moral hazard is reduced because the sponsor now bears basis risk between its own claims and the industry index.
Parametric
Based on measured physical parameters — wind speed at a location, earthquake magnitude. Fastest and most transparent settlement and lowest moral hazard, but the investor bears the largest basis risk: the parameter can be breached without a matching loss, or vice versa.
Modelled loss
The event's parameters are run through an agreed catastrophe model against the sponsor's exposure portfolio; the bond triggers if modelled losses exceed a threshold. Settles quickly without waiting for actual claims, at the cost of dependence on the model's integrity.

Investment characteristics. The risk driver is meteorological or seismic, so correlation with financial assets is genuinely low and remains low through financial crises — the core diversification case. Against that: negatively skewed returns with severe tail losses, catastrophe-model risk (and non-stationarity as climate changes), loss creep and delayed settlement on indemnity structures, trapped collateral, seasonality of peril exposure, and concentration of the market in a few perils and regions.

Life ILS: longevity & mortality risk

Two opposite exposures. Mortality risk is the risk that people die sooner than expected — the exposure of a life insurer, hedged with mortality catastrophe bonds that trigger on a spike in a population mortality index (pandemic, war, natural disaster). Longevity risk is the risk that people live longer than expected — the exposure of an annuity provider or a defined benefit pension plan, hedged with longevity swaps, longevity bonds and reinsurance.

Longevity swaps are the practically important instrument: the pension plan pays a fixed schedule of payments and receives payments matching the actual experience of its members, converting an uncertain lifetime liability into a fixed one. Structures are either indemnity/bespoke — matched to the plan's own members, no basis risk, expensive and slow to execute — or index-based, referencing a national population index, cheaper and faster but leaving the plan with basis risk between its members' longevity and the index. A pension question that mentions "hedging longevity" is asking you to make exactly this trade-off.

Life settlements complete the set: purchasing an in-force life insurance policy from the insured at more than surrender value but less than face value, then paying the premiums and collecting the death benefit. The return depends on life-expectancy estimation, so the risk is that the insured lives longer than underwritten (extension risk) plus premium-cost escalation, and the market carries genuine ethical and reputational considerations that an institutional investor must address explicitly.

Why these appeal to institutions: the driver is demographic rather than financial, giving low correlation to markets. Why they are hard: pricing depends on actuarial models with long horizons and structural trend uncertainty (medical advances), liquidity is minimal, valuation is model-based, and counterparty exposure on a decades-long swap is material — which is why collateralisation terms are a central negotiating point.

Hybrid products: mezzanine debt

Where it sits. Mezzanine is subordinated debt ranking between senior secured debt and equity, used chiefly in buyouts and mid-market growth financing when senior lenders will not stretch far enough and the sponsor does not want to issue more equity. It is a hybrid because its return is engineered from several components: a cash coupon, often a PIK (payment-in-kind) accrual, arrangement and exit fees, and an equity kicker in the form of warrants or a conversion right.

Why it exists. For the borrower it is cheaper than equity and less dilutive, typically with looser covenants, bullet maturity and no amortisation; for the sponsor it increases the leverage available and therefore equity returns. For the lender it offers an equity-like return with a contractual claim ahead of equity — the classic risk-return middle ground.

Structural risks
Subordination means low recovery in default — mezzanine is close to equity in a downside scenario despite being debt. Intercreditor agreements govern payment blockages, standstill periods and enforcement rights: read them, because they determine whether the mezzanine lender has any leverage in a restructuring.
PIK risk
Accrued rather than cash interest compounds the exposure and masks stress — the position grows precisely as the borrower's ability to pay deteriorates.
Equity kicker
Warrants supply the upside that justifies the risk; their value depends on the exit, so mezzanine returns are correlated with sponsor exit conditions, not just with credit.
Valuation
Unquoted, model-marked, and hard to compare across deals — the ODD questions about valuation policy apply directly.

In an allocation answer: mezzanine and private credit generally are hired for contractual income with equity-like return, funded from either the credit or the return-seeking sleeve depending on the institution's framing — and the honest caveat is that in a severe downturn its correlation with private equity is high, because it is exposure to the same companies further down the same capital structure.

Confusion pairs

Equilibrium vs arbitrage-free
Economically derived, will not reprice every bond, good for scenarios vs calibrated to today's curve exactly, required for derivative pricing.
Vasicek vs CIR
Constant volatility, rates can go negative vs √r volatility, rates bounded at zero.
Cap vs swaption
A strip of options on a floating rate, period by period vs one option on an entire swap.
Attachment vs exhaustion point
Where investor principal starts being lost vs where it is fully lost.
Indemnity vs parametric trigger
Sponsor's actual claims — no sponsor basis risk, moral hazard, slow settlement vs physical measurement — fast, transparent, investor bears basis risk.
Mortality vs longevity risk
Dying sooner than expected (the insurer's exposure) vs living longer than expected (the annuity and pension exposure).
Bespoke vs index longevity swap
Matched to the plan's own members, no basis risk, costly vs population index, cheaper, basis risk retained.
Revolving vs amortising ABS
Credit card structures with a revolving period and early-amortisation triggers vs auto loans that amortise from the start.
Senior tranche vs correlation
Senior investors are implicitly short correlation; equity tranche holders can benefit from it.

Practice

Six multiple-choice questions in exam style, with the reasoning — not just the letter.

1. An analyst must price a portfolio of swaptions consistently with today's observed yield curve. The most appropriate model family is:

A. Equilibrium single-factor models such as Vasicek or CIR
B. Arbitrage-free models such as Hull–White or BDT
C. A historical simulation of rate changes
D. A regression of swap rates on macro variables
B. Only a no-arbitrage model calibrated to the current curve prices derivatives consistently with observable bond prices. Equilibrium models are for scenario generation and economic interpretation, not for relative-value pricing.

2. Holding expected pool loss constant, an increase in default correlation across an ABS collateral pool most likely:

A. Benefits the senior tranche and harms the equity tranche
B. Harms the senior tranche and may benefit the equity tranche
C. Affects neither
D. Harms both equally
B. Higher correlation fattens both tails: the chance of losses large enough to reach the senior tranche rises, while the chance of very few defaults — which is all the equity tranche needs — also rises. Senior investors are implicitly short correlation.

3. A sponsor wants a catastrophe bond with no basis risk to its own losses. The appropriate trigger, and its consequence for investors, is:

A. Parametric; investors settle quickly
B. Industry loss; investors bear moral hazard
C. Indemnity; investors face slower settlement and moral hazard, so demand extra spread
D. Modelled loss; investors bear no model risk
C. Indemnity triggers eliminate the sponsor's basis risk by paying on actual claims — which means waiting for claims to settle and accepting the sponsor's incentive to underwrite loosely. Investors price that.

4. A defined benefit plan enters an index-based longevity swap rather than a bespoke one. The principal consequence is:

A. It eliminates all longevity risk
B. It retains basis risk between its members and the reference population
C. It converts longevity risk into interest rate risk
D. It removes counterparty exposure
B. Index hedges are cheaper and faster to execute but pay on national population experience, which need not match the plan's own membership. Counterparty exposure remains and is managed with collateral.

5. Which feature of mezzanine debt most increases risk as a borrower deteriorates?

A. Arrangement fees
B. Bullet maturity
C. PIK interest that accrues rather than being paid in cash
D. Warrant coverage
C. PIK compounds the exposure precisely while cash generation is weakening, and masks distress because no payment default occurs. Warrants are the upside, not the risk; bullet maturity matters but is known in advance.

6. Credit card ABS differ structurally from auto loan ABS principally because:

A. They carry no credit risk
B. Collateral is revolving, requiring a revolving period and early-amortisation triggers
C. They are not tranched
D. They have longer stated maturities than auto loans
B. Non-amortising receivables require a structure that reinvests principal during a revolving period, with excess-spread and performance triggers that force early amortisation if the pool deteriorates.

Constructed-response practice

Write these under time. Each outline is the shape the rubric rewards, not a model answer to memorise.

Prompt A (14 minutes). A pension plan is considering a 3% allocation to catastrophe bonds. Explain the investment case, the four principal risks, and how you would implement and monitor the allocation.

Outline:
  • Case: a non-financial risk driver, so genuinely low correlation that persists in financial crises; spread over collateral yield; contractual, defined triggers.
  • Risk 1 — negative skew and tail severity: many small gains, rare large principal losses.
  • Risk 2 — model risk in the catastrophe model, aggravated by climate non-stationarity in the historical hazard record.
  • Risk 3 — basis, settlement and structural risks by trigger type; loss creep and trapped collateral on indemnity structures.
  • Risk 4 — concentration by peril, region and season, plus limited liquidity in the secondary market.
  • Implementation: diversified fund or mandate across perils and regions rather than single bonds; sizing on stress loss, not volatility; monitoring modelled expected loss, attachment probabilities, exposure by peril and season, and independent review of the model vendor.

Prompt B (12 minutes). A mid-market buyout sponsor offers a mezzanine tranche at a 10% cash coupon plus 3% PIK plus warrants. Assess the risk-return proposition for an insurer, and identify the three documentation terms you would examine most closely.

Outline:
  • Decompose the return: cash coupon, PIK accrual, fees, and warrant value dependent on the exit — note that the last makes the return partly equity-like.
  • Risk: subordination implies low recovery in default, so downside is closer to equity than to senior debt; PIK compounds exposure as stress builds.
  • Insurer-specific lens: capital treatment and rating-agency view, valuation of an unquoted model-marked asset, and the correlation of the position with the plan's other private-equity exposure.
  • Terms: the intercreditor agreement (payment blockage, standstill, enforcement rights), covenant package and equity-cure provisions, and the warrant/conversion mechanics including anti-dilution.
  • Conclude with the condition for proceeding — for example only within a diversified private-credit mandate, sized on stress-case recovery rather than on the headline coupon.
← 13 · Hedge Funds Study desk →