What is counterparty credit risk? Basically the risk that
a counterparty who owes us money cannot pay us what is owed. For treasurers the
focus is mostly on financial counterparties, but many treasurers are also
responsible for commercial counterparties as well. In some companies there may
be overlap; for example if banks are customers then they may be both financial
and commercial counterparties. The focus of this article is on financial
counterparty credit risk.
Measuring Credit Risk
In risk management, the task is to ascertain the potential for loss when
determining how aggressively to manage the risk. As with other risks, to manage
credit risk it is first necessary to know the amount at risk, which is called
the exposure. When multiplying the amount at risk times the riskiness, the
product is a measure of risk. Risk equals exposure multiplied by riskiness.
Credit risk is normally measured in terms of:
- Default risk – the
risk of an event of default such as bankruptcy.
- Loss given default –
how much we are likely to lose in an event of default, which might be a haircut
rather than 100% loss.
So credit risk (the riskiness) can be
shown as: Credit risk% = Default risk% x Loss given default%.
further: Expected loss$ = Exposure$ x Credit risk%.
Each element has its
wrinkles. Let’s look at calculating exposure first.
The exposure is the amount that is at risk of a
credit loss. From a credit perspective it is not only the amount but also the
tenor that impacts the risk of loss. An overnight deposit is inherently less
risky than a 10-year bond because the likelihood of bad news overnight is much
lower than the likelihood of bad news over a period of 10 years.
complexity comes from the breadth of instruments we may have outstanding with
financial counterparties and from the spread of geographies in which we
interact with financial counterparties. For those of us lucky enough to have a
globally integrated treasury management system (TMS) that covers all
transactions with financial counterparties, this complexity is trivial.
Many treasurers find that not all financial transactions are captured in
their TMS, or are sometimes not captured fast enough. Even those who have all
regional treasury centres (RTCs) on a common TMS, may encounter issues such
- Subsidiaries in exchange control countries make deposits with
global relationship banks locally and these are only reflected in global
systems at month-end.
- Supply chain and logistics colleagues use
letters of credit (L/Cs) receivable from global relationship banks but are not
TMS users, so the L/C risk is either not captured, or only infrequently
captured, in TMS.
Another issue is to determine a consistent
exposure amount across different instruments. Should a US$10m deposit be
considered the same exposure as a US$10m forward foreign exchange deal? Most
would contend that they are not the same.
The best practice solution
for integrating financial counterparty credit exposures is to have them all in
one system – most likely a TMS or a corporate business intelligence (BI) system
or data warehouse. The problems include:
- The need to have all
geographies on the same or on an interfaced platform.
- The need to have
all instruments on the same or on an interfaced platform; for example L/Cs,
which generate bank credit risk, may be handled by sales or logistics who may
not use TMS.
- The need to have visibility into funds to track
underlying investments to manage limits holistically.
common to allocate limits at least to parts of the organisation that do not
access TMS but at least monthly consolidation is good practice. Such a limit
allocation will typically be both geographic and functional. For example,
limits may be allocated to regions and maybe to large subsidiaries, and also to
functions such as supply chain for L/C processing.
Limit allocation is,
of course, less efficient than online limit management. To mitigate the
inefficiency, the treasurer normally delegates to treasury centres (TCs) the
authority to reallocate limits when required.
Typically the counterparty
exposure limits will be different for different tenors. Generally longer tenors
will only be relevant to head office and treasury centres. Typical tenor
buckets might be less than three months; less than one year; less than three
years and three years and longer.
Most companies make their exposure
measures consistent across instruments, by using some kind of loan or cash
equivalence as per the typical example below:
If the company invests in money market funds (MMFs) and bond funds, the
counterparty exposures from such investments have to be integrated in the
overall credit exposure calculation. Best practice for this is to look through
the fund to the individual investments. This first requires that funds can
provide details of their investments in a timely and usable manner. While most
reputable funds provide weekly transparency into their investments, the
industry has no standard format for such data. So integrating fund data into
the counterparty exposure consolidation can be a non-trivial data processing
A practical alternative is to choose funds with strict
concentration limits and then to take a proportional haircut from the
counterparty exposure limits.
Measuring the Riskiness of
Before the global financial crisis most
corporates relied on agency ratings. It was known that they were statistically
unreliable but since the financial markets are built on ratings companies just
accepted what seemed an academic issue with rating reliability.
crisis showed us that the risk is real. Without government intervention
corporates would have sustained billions of losses on bank names. So treasury
departments looked for alternatives:
- CDS initially looked good, but
quickly were found to be too volatile for corporate risk management purposes,
raising too many false alarms.
- Some larger corporates tried to carry
out their own credit analysis, but reading bank balance sheets is a non-trivial
exercise fraught with timing and complexity issues.
- Composite credit
metrics, or implied ratings, such as Thomson Reuters StarMine, Bloomberg CRAT,
and Moody’s KMV provide a useful middle road, each with fewer false alarms and
less prone to market manipulation than CDS but much better at giving early
warning than agency ratings.
Implied ratings provide a reasonably
stable but near real-time assessment of credit risk. They are not subject to
the kind of trading volatility that afflicts CDS spreads. Yet implied ratings
will normally give warning signals well before agency ratings changes and watch
Much academic literature has been written about
implied ratings and market data algorithms to determine credit risk. A
mathematically-inclined treasurer could probably construct their own model
using data feeds from Thomson Reuters or Bloomberg or even Google or Yahoo
Finance. But why reinvent the wheel?
This is why I have suggested taking
the lower of the agency rating and the implied rating as a conservative
measure. StarMine comes with the Thomson Reuters package as a special set of
Reuters Instrument Codes (RIC) so it is quite easy to integrate for anyone who
already uses Thomson Reuters.
For simplicity’s sake I will use StarMine
for this article. StarMine credit scores are available as RICs and so easy to
integrate where RICs are already used in TMS and Excel. StarMine credit scores
are also part of the Thomson Reuters service, available without extra charge,
which means it is free if you already have a Reuters terminal or feed. StarMine
covers approximately 2000 banks. Be aware that there are implied rating
products available from Bloomberg and also from the credit rating agencies
Implied ratings have a
long academic and market history, including metrics such as KMV and z-scores.
Modern technology and market bandwidth have enabled increasing richness in raw
data and sophistication in the models themselves.
Credit Risk model (SCR) builds on these antecedents to provide forward looking
estimates of credit risk that generally anticipate agency rating changes.
Experience with the collapse of Lehman Brothers shows that SCR indicated
problems two months before the rating agencies. (In fact, Lehman Brothers was
still rated “A” by the agencies when it collapsed.)
Figure 1: Comparison of Agency Rating and SCR Rating for Lehman
The SmartRatios Model uses a third generation
model to calculate risk using a wide array of accounting ratios that are
predictive of credit risk. It also employes several sets of metrics, some
generic and some industry specific, clustered around the following components:
profitability; leverage; coverage; liquidity; growth and stability; sector and
region specific metrics.
The model uses option pricing
mathematics which evolved from the Merton structural credit risk model. This
produces default probability as a percentage which can be mapped to agency
ratings as a comparison. So treasurers who already have a counterparty credit
risk management process in place that relies on agency ratings can simply
modify the process to use (conservatively) the lower of the agency rating or
the StarMine rating equivalent which is available as a RIC.
The StarMine default probability can also be used directly to
quantify credit risk, using the formula given above. (Loss Given Default (LGD)
is calculated as 1-Recovery Rate * Exposure. Recovery rate is available on
Thomson Reuters’ CDS RICs (covering 2800 issuers). Exposure is dependent on the
Thomson Reuters have done much research to
determine how accurate StarMine’s SmartRatios Model is compared to earlier
models that calculate default probabilities:
Figure 2: Accuracy Ratio Comparison.
Post-crisis counterparty credit risk has become a major concern
for treasurers. Spooked boards are rightly asking difficult questions and
treasurers must ensure that they have a comprehensive view on counterparty
credit exposure globally and group wide. Wide press coverage of rating agency
deficiencies has prompted boards to ask for more rigorous credit metrics.
Implied ratings provide a solution that meets board members’ concerns at
reasonable cost in terms of time and money.
Many banks around the world, large and small, continue to experience major security failures. Biometric systems such as pay-by-selfie, iris scanners and vein pattern authentication can help.
The implementation date of Europe's revised Markets in Financial Instruments Directive, aka MiFID II, is fast approaching. Yet evidence suggests that awareness about the impact of Brexit on MiFID II is, at best, only patchy and there are some alarming misconceptions.
Banks might feel justified in victim blaming when fraud occurs, but it does little for customer confidence.
Politicians have united in urging the Reserve Bank of Australia to lend its backing to the digital currency by officially recognising it.