An issue has been bubbling away in the background of the catastrophe bond and ILS fund market since late 2015 which has the potential to impact the way managers of UCITS catastrophe bond funds deal with diversification.
It came to light late last year that, at the instigation of the European Securities & Markets Authority (ESMA), the financial regulator in Luxembourg, a popular domicile for UCITS catastrophe bond funds, was seeking to apply the so-called 20/35 diversification (or perhaps concentration risk) rule to cat bond funds.
This rule, 20/35, is designed to ensure that a UCITS investment fund’s maximum concentration of a single risk is 35%, with none of the other risks exceeding a maximum of 20%. The rule is designed to ensure that funds are not set up with large exposures to a single equity or commodity position, so reducing the risk by increasing the diversification of UCITS funds.
A very sensible idea and something that you would imagine the ILS fund market could easily support, given the multitude of ways you can segment an ILS or catastrophe bond fund strategy, so as to protect the strategies end-investors.
However, the way the rule appears to be implemented in this case is not ideal, in fact it could be deemed extremely restrictive for the catastrophe bond fund market.
The regulator in Luxembourg is applying the rule to newly registered UCITS cat bond funds on a notional exposure basis. That means that, in a market where U.S. hurricane risk currently makes up 60%+ of the outstanding notional available to fund managers and investors, achieving the sort of diversification implied by a 20/35 becomes very difficult.
But what about regional perils, I hear you ask? Or expected loss? Or return periods? Or attachment points? Or triggers? Or the multitude of other ways that ILS funds segment their assets into pillars in order to diversify their portfolios and ensure that no single loss event can have a major impact on the fund?
As it stands, it appears that the regulator is leaving it up to managers of UCITS cat bond funds to establish their own peril buckets, such as going down the lines of establishing buckets of regional perils. But, given the proliferation of U.S. wide and multiple peril cat bond deals, splitting up a portfolio to get to a 35% maximum concentration is hard and getting to the 20% extremely difficult.
Artemis has spoken with a number of managers of UCITS cat bond funds in recent months, from those running recently launched strategies which are already being mandated to follow this rule, to more mature Luxembourg based UCITS cat bond funds with hundreds of millions of asset under management already.
Opinions of the ILS managers we’ve spoken with on this topic over recent months range from the rule being “severely restrictive” through to “unworkable” and all said that there are other ways to ensure that a UCITS cat bond fund is not overly exposed to a single major loss event than by applying the 35/20 rule on a notional exposure basis.
We understand that regulators from other UCITS domiciles have not yet decided whether to apply the rule to cat bond funds registered there, but the conversation has begun, we understand.
It is to be hoped that the European domiciles of UCITS funds could find a common way to approach this rule, with a pragmatic and workable concentration risk and diversification solution, which takes into account the natural diversity within cat bonds and catastrophe risks as assets.
While regional perils is one way of approaching this, although difficult to work given the way cat bonds are structured, other ways would perhaps be preferable, such as a diversification based on peril and expected loss or return period.
So that would imply a maximum of 35% of, say, Florida hurricane risk that has an expected loss of 0% to 2% and then 20% maximum of Florida hurricane risk in the expected loss brackets of 2.01% to 4%, 4.01% to 6%, etc.
The 20/35 rule was originally designed for index funds, to ensure sufficient diversification within UCITS funds and to prevent over-exposure to single counterparties or assets. It now appears to be being applied to an asset class where the diversification is much more nuanced than simply a peril exposure concentration rule.
In fact, the way this rule appears to have been implemented for ILS funds would suggest that offering a UCITS fund investing in equities of companies of a single nation should be impossible, as, under this interpretation of the rule, the entire fund is technically exposed to one risk (or peril in cat bond terms).
The newly launched UCITS catastrophe bond funds in Luxembourg, which have, according to our sources, been directed to follow this rule, will have a tough time building scale with such restriction in place on their diversification. And it may well mean that the rule forces them to hold a significant cash allocation in order to comply with it, hardly in the interest of investors.
For the mature Luxembourg domiciled UCITS cat bond funds which are already large (some in the many hundreds of millions of dollars size range) this will be very troubling, as to shift to a portfolio that meets this diversification restriction, as the Luxembourg regulator has defined it, will be very difficult.
As other European regulators become aware of this, which they are certain to as it has been pushed by the Europe-wide regulator ESMA, and look at how to apply this rule to any UCITS cat bond funds they play host to, it is to be hoped that there will be a chance for a conversation about how best to apply this to an asset class which is very different in terms of correlation between risks, and for which diversification is not as simple as pure exposure-type based.
We’ll keep you updated as any further information emerges on this topic.