Counterparty and issuer risk: CVA, ETCs, certificates and swaps

Advanced12 min read · updated 8 September 2026 · 14 sources

Issuer risk means that a debt security is only as good as its issuer. If the issuer defaults, even a certificate on a rising index can become worthless. Counterparty risk is the counterpart for derivatives: the other party defaults while the trade is in profit for you. This article classifies ETFs, ETCs, ETNs, certificates, swap ETFs and securities lending, and shows how banks price the risk with the credit valuation adjustment (CVA), with a worked example.

Issuer risk: who pays in the end

Certificates, warrants, factor certificates, ETNs and ETCs are legally debt securities (Schuldverschreibungen): you have a claim against the issuer, usually a bank, but no share in segregated fund assets. If the issuer becomes insolvent, you are an insolvency creditor and receive only what is left after the proceedings, regardless of how the underlying has performed. The statutory deposit guarantee scheme does not help: debt securities issued by a credit institution are excluded from compensation (EinSiG 2015).

The best-known example in Germany are the certificates of Lehman Brothers Treasury Co. B.V., guaranteed by the parent company Lehman Brothers Holdings Inc. After the insolvency of the issuer and the guarantor in September 2008, they were largely worthless, including products with “Garant” (guarantee) in their name, whose capital protection was worth only as much as the bank's solvency. In the cases decided in 2011, a savings bank (Sparkasse) had sold the products; the Federal Court of Justice (Bundesgerichtshof, BGH) dismissed the investors' claims because they had been adequately informed about the risk of a total loss in the event of a Lehman insolvency (BGH 2011).

For funds, the law separates the investors' assets from the assets of the asset management company (Kapitalverwaltungsgesellschaft); they are not part of its insolvency estate (KAGB § 92 2013). An ETC, by contrast, is a debt security whose risk is reduced by collateralisation but not eliminated: legal and operational risks relating to the trustee, the custodian and the realisation of collateral remain. An ETN is usually unsecured, so the issuer risk is borne in full.

Counterparty risk in swap ETFs and securities lending

A swap ETF holds a basket of securities that need not match the index and exchanges the basket's return for the index return through a total return swap. If the index rises more than the basket, the swap counterparty owes the fund the difference. That is the exposure. If the counterparty defaults, the fund loses at most this positive swap value, not the basket. German funds may risk at most 5% of their value per counterparty from derivatives that are not centrally cleared, securities loans and repurchase agreements, or 10% for credit institutions from the EU or EEA or subject to equivalent supervision. Collateral counts after deductions (haircuts), and claims and liabilities under effective netting agreements are offset (DerivateV 2013). Directive (EU) 2024/2994 clarifies for the UCITS Directive that the limits apply to derivatives that are not cleared through an authorised central counterparty; this had to be transposed by 25 June 2026 (Directive (EU) 2024/2994 2024).

The risk of securities lending is similar: the fund lends securities against collateral. If the borrower defaults, the fund realises the collateral and buys the securities back; a loss arises only if the collateral is then worth less. Example: €10 million lent against 105% collateral, i.e. €10.5 million. By the time the collateral is realised, the stocks have risen by 8% to €10.8 million and the collateral has fallen by 3% to €10.185 million. The shortfall is €0.615 million, which for a fund of €100 million is 0.62%.

Collateralisation, netting and central clearing

Between banks and large investors, a master agreement with a collateral annex governs who posts collateral and when. Variation margin offsets the daily changes in value. Initial margin additionally covers what can move between the last exchange of collateral and the close-out after a default, the margin period of risk. Netting offsets all trades with one counterparty so that only the net balance counts. With central clearing, a central counterparty (CCP) steps in between the parties; in the EU, the EMIR regulation (European Market Infrastructure Regulation) requires this for certain standardised OTC derivatives and demands risk mitigation, such as the exchange of collateral, for non-cleared trades (Gregory 2020).

CVA: the price of default risk

The credit valuation adjustment (CVA) is the deduction from the risk-free value of the derivatives with a counterparty to account for that counterparty's default risk. Hull and White (2012) describe the usual procedure: the remaining term is divided into time steps; for each step you need the counterparty's probability of default, estimated from credit spreads, and the expected exposure in the middle of the step, calculated by Monte Carlo simulation of the market variables.

Potential future exposure (PFE), by contrast, is a high quantile of the exposure, such as 95%, and is used for limits. For an interest rate swap that is worth nothing at inception, the exposure has a hump shape: at the start, little time has passed for rates to change; towards the end, only a few payments remain outstanding. If the change in value is normally distributed with mean 0 and standard deviation σ_t, then EE = σ_t / √(2π) ≈ 0.40 × σ_t and PFE₉₅ = 1.645 × σ_t.

Worked example: CVA of a three-year interest rate swap

The adjustment looks small because a swap with no initial value builds up only a fraction of its notional as exposure. With long maturities, cross-currency swaps with an exchange of notional, or weak counterparties, it quickly becomes a multiple of that.

Wrong-way risk, DVA, FVA and the xVA family

According to Hull and White (2012), wrong-way risk arises when the probability of default and the exposure are positively correlated, and right-way risk when the correlation is negative. The authors link the counterparty's default intensity to variables in the simulation, for example the gold price in the case of a gold producer, and find that wrong-way and right-way risk change the CVA significantly. By how much depends on the collateralisation. The debit valuation adjustment (DVA) is the mirror image: the gain in value from the possibility that the bank itself defaults.

On top of that come further valuation adjustments, collectively known as xVA (Gregory 2020; Brigo et al. 2013): FVA for the cost of funding uncollateralised trades, ColVA for the terms of the collateral agreement, MVA for funding the initial margin and KVA for the cost of the capital that a trade ties up.

CVA in banking regulation: Basel III

Basel II covered the default and migration risks of counterparties, but not mark-to-market losses from CVA. According to Basel Committee (2011), however, around two thirds of the losses from counterparty risk in the financial crisis came from CVA losses and only about a third from actual defaults. Basel III therefore requires capital for CVA risk. In 2011, the Committee lowered the weight for CCC-rated counterparties in the standardised approach from 18% to 10% and expected capital for counterparty risk, including CVA, to roughly double compared with Basel II. In 2020, it completed the revision: adjusted risk weights, a separate treatment of certain cleared client transactions, and recalibrated standardised and basic approaches, aligned with the market risk framework (Basel Committee 2020).

Advanced: modelling exposure and default intensity

• Hump shape: with σ_t proportional to √t × (T − t), the maximum is at t = T/3, after one year in the example. For cross-currency swaps with an exchange of notional at the end, by contrast, the exposure grows until maturity.
• Default intensity: the approximation λ ≈ s / (1 − R) follows from the condition that the running CDS premium s equals the expected loss λ × (1 − R). Intensity models (reduced-form models) such as in Duffie and Singleton (1999) treat default as a jump with intensity λ; for several maturities, a curve of intensities is bootstrapped step by step from CDS spreads, like the spot rates in the article bond maths.
• Netting: CVA is not additive. With netting, the exposure of a portfolio is usually smaller than the sum of the individual exposures; a new trade that offsets existing positions can even reduce the CVA.
• Collateralisation: with variation margin, what remains is the change in value over the margin period of risk as well as thresholds and minimum transfer amounts; initial margin covers a further part of that.

What research shows

• Arora et al. (2012) analysed quotes from 14 CDS dealers selling default protection on the same companies: dealers' credit risk is priced, but the effect is vanishingly small. This is consistent with a market in which liabilities are collateralised.
• Duffie and Zhu (2011) show that a central counterparty for just one class of derivatives, such as credit default swaps, reduces netting efficiency in plausible cases and increases average exposure to defaults. Several separate CCPs for different classes always increase exposure compared with a single joint one.
• For portfolios of 25 instruments on five market variables, Hull and White (2012) found that wrong-way and right-way risk change the CVA significantly and that the percentage effect depends on the collateralisation.
• Basel Committee (2011) justified the CVA capital requirement with the losses of the financial crisis; textbooks such as Gregory (2020) and Brigo et al. (2013) cover exposure, collateralisation and the entire xVA family.

Limitations and common mistakes

• A “guarantee” is not security: a certificate's capital protection applies only as long as the issuer pays.
• Safe underlying, risky issuer: an unsecured certificate on gold is not gold.
• Treating swap ETFs and certificates as the same: with a swap ETF, only the positive swap value is at risk, which is limited and usually collateralised; with a certificate, it is the entire amount.
• CVA from a single point: a single CDS spread and a fixed recovery rate are strong simplifications. For many counterparties there is no liquid CDS; then only proxies based on rating, sector and region can help.
• Forgetting correlation: without wrong-way risk, CVA underestimates the risk precisely where it is greatest.

How Portfolio Lab calculates

Portfolio Lab has no tool for CVA or issuer risk. The option pricer values options with Black-Scholes-Merton, Black 76, Garman-Kohlhagen, binomial trees and Monte Carlo, without any default risk of the option writer. For warrants issued by a bank, the result therefore lacks the deduction for the bank's creditworthiness. For a swap ETF, the Look-through shows the fund's holdings list, i.e. the basket held and not the index; it treats derivatives such as swaps and futures like cash and, if the list gives no sector, assigns them to the “Cash/derivatives” sector. The withholding tax estimate in the Look-through flags swap ETFs and estimates them like physical funds; in the security search you can filter by replication method. How funds track their index is explained in the article index replication.

This article explains a method; it is not investment or tax advice. Worked examples are simplified and rounded. All terms in the glossary.