Covered bonds in LCR category 1?

One of the most pressing issues regarding banking regulation and covered bonds has been the treatment of covered bonds for the Liquidity Coverage Ratio.

The Basel Committee announced changes earlier this year that weren’t exactly picture perfect for covered bonds — while they reduced the minimum rating for corporate bonds down to BBB-, the AA- minimum for covered bonds remained.

In the more recent past, it has been the Europeans’ turn to publish further pieces on the matter:

• On 13 February the European Council published an addendum to the CRD IV text which included changes that had been provisionally agreed on by the Commission, Parliament and Council in the trilogue process

• On 21 February the European Banking Authority published a discussion paper on how to define liquid assets in the LCR

CRD IV and LCR treatment

Before going into the details, just a brief note on timing. Brussels wants to implement the LCR a year earlier than is proposed by the Basel Committee (full implementation already in 2018 thanks to a shorter transition phase). Also, the deadline for EBA to propose technical standards to the Commission has moved from June to the end of this year.

The CRD IV addendum reads fairly nicely for covered bonds and their potential treatment in the LCR. The focus continues to move from a very limited set of eligible assets to avoiding pushing banks into concentration risks. The scope of potentially eligible assets is thus set to widen:

• “A concentration of assets and overreliance on market liquidity creates systemic risk to the financial sector and should be avoided.”

• “A broad set of quality assets should therefore be taken into consideration during an initial observation period which will be used for the development of a definition of a Liquidity Coverage Ratio (LCR).”

In that context, the opening remarks of the EU document state that covered bonds could even move into category 1:

• “When making a uniform definition of liquid assets at least government bonds, and covered bonds traded on transparent markets with an ongoing turnover would be expected to be considered assets of extremely high liquidity and credit quality.”

We want to caution everyone against getting too excited about this news at this point. The category 1 statement in the CRD addendum sounds great but the EU stresses the fact that covered bonds eligible for this treatment have to have been traded on transparent markets with ongoing turnover.

As such, the ball is now in the EBA’s court because that is exactly what they have to determine: which products have had measurable liquidity over time and which haven’t.

EBA paper on technical standards

As mentioned above, EBA has to propose technical standards on the LCR to the European Commission. On 21 February they issued a discussion paper on the subject in which they stated their objectives and outlined the approach they plan to take. Their objectives in this exercise are:

• “(i) establish a ranking of asset classes based on their aggregate liquidity properties; and

•  “(ii) identify explanatory characteristics of individual securities that explain observed liquidity differences within asset classes.”

To achieve this, EBA plans to use a two step approach:

• “Assess a common set of liquidity metrics across all asset classes. These metrics will be computed first at the ISIN level, but the primary analysis will focus on aggregated results by asset class. An ordinal ranking of asset classes in terms of liquidity will be constructed.”

• “Test whether explanatory characteristics of individual securities within each asset class can be used to predict their liquidity in quantitative terms. EBA plans to attempt to construct definitions that should be fulfilled by individual assets within a particular eligible asset class, in order to be included in the liquid asset buffer as either transferrable assets of high or extremely high liquidity and credit quality.”

• “Finally, the analysis will identify the features that are of particular importance to market liquidity. Within individual asset classes that are found to contain assets of high liquidity and credit quality, appropriate haircuts will be proposed, based on the empirical evidence on historical price movements.”

The way we read the document at this point is that EBA wants to focus a lot on ex post turnover statistics and bid-ask spreads when determining the ultimate LCR treatment of individual assets. They then want to run extensive quantitative analysis on these figures. Qualitative criteria mentioned above are part of this exercise, but as things stand at the moment they will not have a very important role in the process. All EBA plans do with them is back test if they have any statistical relevance to refine the interpretation of the turnover and bid-ask spread data — they don’t seem to have any decisive meaning in their own right.

In our view, the proposed approach focuses too heavily on data that is still hard to come by and also hard to interpret. The LCR’s purpose is to define assets that banks can liquidate if they need to, which means we’re focussing on the bid only. Merely basing the LCR treatment on ex post trading turnover data is misleading in our view.

Let’s just take the first covered bond purchase programme of the ECB or look at late 2012/early 2013. Trading turnover was fairly low but the reason for this was not the absence of buyers but rather the absence of sellers. Looking at turnover numbers and bid-ask spreads for these periods would probably lead EBA to think that covered bonds were a rather bad asset to hold as a bank treasurer compared with, for example, corporate debt. However, the opposite would in fact have been true as there was plenty of demand for the bonds.

Additionally, starting at the asset class level to define liquidity ranking will put covered bonds at a disadvantage compared with other debt types because the share of peripheral bonds is higher than in most other asset classes. A strength of the market (that there is demand for these bonds while there is sometimes not for corporate or bank unsecured debt from these countries) would thus be turned into a disadvantage for the overall asset class.

Bottom line

While the CRD IV text talks about broadening the scope for LCR eligible assets, EBA’s very strong focus on trading statistics could in the end narrow the subset down substantially. The fact that EBA also wants to define the overall liquidity ranking at the asset class level but then define metrics that every bond has to fulfil to be eligible could lead to a very heterogeneous treatment of the asset class. This in itself would reduce transparency for bank treasurers (who would have to check each bond individually) and could ultimately have an adverse impact on liquidity.

So for now, realistically, the CRD statement and the EBA approach is something our friends in Denmark and Sweden should be really excited about as their domestic mortgage bond markets are probably the ones with the most extensive data on liquidity in the covered bond space. As such, they would be the most likely candidates for category 1 treatment.

Getting euro benchmark covered bonds in there will still require a lot of lobbying work by the European Covered Bond Council and even then it is nowhere near certain.

Florian Eichert

Senior Covered Bond Analyst

Crédit Agricole CIB