The credit conundrum: a matter of opinion

What was one of the main criticisms arising from the post Icelandic fallout?  Over-reliance on credit ratings.  Looking a bit deeper into that, was it a criticism about ratings per se, or a criticism about reliance on the opinions of a third party to formulate the authority’s credit view?  Perhaps a bit of both.

I totally agree that relying on the slow-moving world of ratings alone is insufficient in order to make decisions worth millions of pounds of public money, and that this absolutely needs to be supplemented with other information available in the market which gives a more real-time indication of the latest position of these institutions.  In fact I would even go a little further and suggest that information available in the markets is the starting point for determining credit risk, supplemented by ratings.

But when it comes to the point about over-reliance on other’s opinions, I think this is much less clear cut.  The research and intelligence that goes into establishing a robust credit opinion that drives a rating is not something easily replicated within a LA treasury management function.  Let’s face it, resources are so stretched, there is barely the time to kick the tyres, let alone look under the bonnet.  So even though LAs remain accountable for investment decisions, it is very challenging to expect this to be done without taking account of the opinions of others who have specialist skills and expertise in the discipline of assessing credit risk.  These expert opinions need to then form a key role in the authority’s lending decisions – within an established framework.

So how are these expert opinions expressed, and how can each authority develop the framework to use them in such a way that the LA remains truly accountable for the decision?

In my view the first place to turn is the market. By definition, an effective market is reliant on all known information about a product being built into the price.  The credit markets are no different.  Provided there is liquidity, the Credit Default Swap (CDS) Spread can be used as a barometer as to the market’s perception of the risk of an entity defaulting.  These observable market indices move real time and are impacted by latest events, current fears and expected future directions of institutions.  They reflect the opinions of participants who actively trade in these instruments to make their livelihood.  These participants have the skills and expertise to form an opinion with conviction.  I would perhaps suggest that it is unlikely that treasury teams would have a more informed view of these entities risk characteristics than the view reflected in the CDS spreads.

But the spreads themselves don’t really give us a sufficient steer to help us form our credit opinion.  What is a good number?  How high is bad? Does reporting an average 5 year CDS of 189 on the portfolio really help Elected Member gauge the extent to which the Council is exposed to risk?   I’m not sure it does.  In absolute terms it doesn’t mean much – compare it to the historical values, and the values of other entities and it starts to mean more.  But can it help us decide on duration?  Not really.  Comparisons of the 5 year CDS of one institution vs another does not give us enough colour on whether lending for 9 months to one or 6 months to another is preferable from a credit risk perspective.

But there is a way to use this constantly moving data to form an active part of our decision making process, and to be able to provide more transparent information to those charged with governance.  This can be achieved by using the CDS spread as the basis for the calculation of a probability of default.  By calculating this probability for each deposit, this allows the credit risk of a 2 year deposit with one institution to be compared with a 1 year deposit of another on a completely level playing field.  Then, calculating a weighted average probability of default for the portfolio as a whole, gives elected members visibility on the risk of default in the portfolio.

In theory, this could be taken a step further.  If, when the investment strategy was set, a benchmark was included in respect of the weighted average probability of default – this would provide an expression of the authority’s credit risk appetite and allow the actual performance to be measured against this.  As well as the target for the weighted average, a further limit for individual deposits could be set.  If the portfolio was then monitored against these limits, it would allow true proactive management of the portfolio.  For example, if the weighted average was looking close to the benchmark set, subsequent investments could be placed in high credit quality organisations (e.g. DMO) in order to keep within the target set.

“But what about counterparties that don’t have an actively traded CDS?”, I hear you cry.    If we can’t tap into the expertise of the market, we need to cast our net a little further.  This is where we need to source another type of expert opinion.  The rating agencies don’t just express their credit opinions in formal ratings, but also in other products that can be of use as part of our credit tool armoury.  For example, FitchSolutions have a Bank Credit Model that analysts use in conjunction with available information on financial factors to provide an implied CDS.  This can then be used in the same way as traded CDS to imply a probability of default.  This is a really useful way that institutions without a formal rating and traded CDS (e.g. Building Societies) can be considered alongside other entities.

This chart shows the Probability of Default vs 12 month expected return for authorities participating in CIPFA’s June 2012 Risk Study, which each authority represented by a bubble, and the size of bubble reflecting the size of the portfolio.

A number of authorities are able to improve their profile by either taking less credit risk for the same returns (moving left) or getting a better return for the same amount of credit risk (moving up)

Figure 1 – CIPFA TM Risk Study June 2012

So the probability of default, I believe, is a measure of credit risk that should be reasonably accessible to a treasurer or elected member. But how do we come up with it? It’s reasonably straightforward to imply this from an institutions CDS spread. It’s also easy to look at their rating and see how frequently entities with similar ratings have defaulted historically. But to be clear, this is completely out of date. No institution, to my knowledge, uses this approach…other than perhaps in some marketing material (when the headline tells you how well something’s done, but the small print rightly reminds you that the past has no reflection on the future). I strongly suggest that you don’t calibrate your measurements of credit risk to history… even before you consider whether the last 20 years is a reasonably barometer for the next 20! Plus, what constitutes a default? Do historic defaults include the risk that an institution’s debt gets restructured and costs investors real money? Has Greece actually defaulted? CDS spreads are a forward looking indicator of credit risk, and can include the cost of quasi default through such things as debt restructuring. It would be foolish to calibrate your credit appetite to the past – indeed, this probably breaks the first rule of investment “the past is no indication of the future”. CDS levels are also used by banks in their capital calculations under Basel III, so we wouldn’t be the first sector to use this approach.

Without the ability to be confident about assessing the credit risk of different institutions, many LAs feel they are erring on the side of caution by keeping duration short.  Clearly this is one way to reduce credit risk, but at the obvious expense of increasing interest rate risk.  And whilst interest rate risk being realised may not make the same kind of headlines as credit risk, the ongoing impact is no less significant and in many cases significantly more so (more on this in my next blog).  As alluded to in my previous article, duration is primarily a tool for managing liquidity risk.

So, to conclude, it will be a long time before formal credit ratings become obsolete in the world of LA credit risk assessment, but in my view, their usefulness is falling behind more dynamic indicators of credit risk.  Local authority treasury teams are never going to be in a position to undertake the depth and breadth of credit analysis needed to be solely responsible for the credit view formed.  It is therefore necessary to use expert skills and opinions to help formulate the view.  If you are buying an opinion, make sure it is from organisations specialising in this art.  When those charged with governance set the credit risk appetite, this should be done in an informed and transparent way.  The risk appetite must be decided within the organisation, not from outsiders.  External service providers can assist with providing information for you to formulate your own view, so you can make truly informed investment decisions.  Also remember, duration is primarily a tool to manage liquidity risk, not a lazy way to manage credit risk!

Jackie Shute is the Co-founder of Public Sector Live