Northern Rock started of life in the mid-1800s as a UK building society, providing savings and mortgages. As with many building societies in the 1990s it became a bank. It continued to provide mortgages for its customers financing itself in the capital markets. In 2007 it shot to fame as becoming the first UK bank in living memory to have a run on it following it approach to the Bank of England for a loan facility.
Northern Rock had been financing itself cheaply on the short-term credit market whilst lending in the long-term market. With the credit crisis and the drying up of liquidity in the market Northern Rock’s ability to refinance itself on an ongoing basis evaporated, leading to its eventual demise.
It is plain to see that ILS funds ‘finance’ themselves in the short-term market. They are offering their investors a variety of liquidity propositions, anywhere from 1 week (even daily for certain structures) for some UCITs funds to 1 year for those who want to enter into annual reinsurance contracts as well as more liquid investments. Most funds sit somewhere between these two liquidity options. In reinsurance terms these periods are short term.
While investors invest in tradable instruments, such as cat bonds or even private cat bonds, there is less of an issue. If investors redeem then the instruments held must be sold, even though the price may not be attractive, and the proceeds paid to the redeeming investors.
The problem arises where funds use reinsurers to front for them, and is compounded when such funds receive leverage. The investments can be outstanding for years. For example: the Lyttelton earthquake event near Christchurch NZ, which occurred in February 2011, has still not yet paid all losses and the final ultimate loss quantum remains as yet unknown.
Long Term Investment
ILS funds are increasingly modelling themselves as ‘mini reinsurers’ in the way that they invest. Reinsurers such as Tokio Millenium Re, Hannover Re and Allianz Risk Transfer act as ‘fronters’, fronting reinsurance paper for ILS funds. This means that the ILS funds will set up a trust account whereby the reinsurer is the beneficiary and the ILS fund, or a cell reinsurer owned by the ILS fund, is the grantor. The funds, once they are in the trust account, must remain there until the reinsurer agrees that they may be repatriated. One should always bear in mind the wise words of Warren Buffet after have spent some years owning a reinsurance company: “Reinsurance is like hell: easy to enter and impossible to leave”.
Under the direction of the ILS underwriters, the fronter will enter into reinsurance in its own name, safe in the knowledge that it has assets under its control to pay losses should they occur. Unlike pure ‘collateralized reinsurance’, the reinsurance policies so issued are without collateral release provisions as the cedent does not have collateral, merely a promise to pay from the fronter.
As the portfolio grows and becomes more diversified the fronter will offer the fund leverage – with a diversified book, including reinstatements, the likelihood for all of the contracts to fail becomes close to an impossibility. At what level they are comfortable will vary, but when there is less than 1 basis point of risk they will likely assume the risk and charge up to 2% - good business for the fronter.
Whereas diversification is good, the downside with diversification is that the more you are diversified, the more you are likely to suffer one or more losses. The more leverage you have the more equity will be consumed by that single / few losses.
Further, the issue is not so much about losses – reinsurers and ILS funds are there to receive premiums and to pay losses – that is why cedents buy cover. The issue arises when losses that have occurred ‘may’ create a loss. In this scenario the fronter will not release the collateral but hold it until the final outcome is known (the 2011 Lyttelton Earthquake outcome is still unknown). Again diversification works against the fund – the more diversified the portfolio the more likely one or more contracts are extended, thereby blocking capital. The more leverage there is the greater impact this will have.
This blocking of capital is unlikely to be a problem while business continues, and even less so if it is growing. If there remains uncertainty about certain contracts the fronter is likely to allow the ILS fund to continue to write new business based on this ‘blocked’ capital. The probability for a loss on the affected contract(s) / collateral will be factored into the calculation of the new business to be underwritten, any new collateral required and any leverage offered. This is the way reinsurers manage their capital. However, whereas this method may work for a reinsurer which has permanent capital and a long-term view, for an ILS fund which has funded itself in the short-term it can be ruinous.
Preferred Liquidation to Support Redemptions
It is to be expected that ILS funds will rather sell their liquid cat bond investments to pay redemptions and continue to hold their illiquid fronted reinsurance investments. It is not possible for an ILS fund to exit from live reinsurance transactions that have been fronted for them.
In the short term, if there are meaningful redemptions across the market, this will translate directly into a supply of cat bonds, a fall in cat bond prices and corresponding increase in cat bond yields.
Correspondingly the reinsurance market, which does not reprice as frequently as the cat bond market, will not change.
What such a fund will do as its reinsurance contracts expire becomes the critical question. The ‘right’ thing to do would be to reduce underwriting fronted reinsurance and move some assets into more liquid investments (i.e. cat bonds). However there can be a strong negative in so doing. If the ILS fund wants to reduce the amount it underwrites through the fronter, the first thing that is unwound is the leverage. This reduces the premium income that the fund receives but does not reduce the collateral it holds with the fronter or return any collateral to the fund so that it may purchase cat bonds. After all the leverage has been unwound, which in itself can take time and therefore cost money, the ILS fund will be able to non-renew reinsurance contracts to withdraw collateral. At this time the reinsurer will ensure that it maintains sufficient collateral to pay all reasonable possible eventualities and allow the rest to be repatriated. This would result in a block of completely non-performing collateral with the reinsurer. If the redemption was a broad market redemption, the effect would be for reinsurance capacity to drop (from asset reduction and leverage) and therefore for reinsurance premium rates to correspondingly rise.
Much more palatable to the ILS fund, in respect of returns going forward, would however be to maintain the fronted portfolio and go forward with reduced liquid investments (cat bonds). Even though it may have a few contracts which are in extension, the probability that there will not be a payment under any such given contract would be considered by the reinsurer. Rather than block 100% collateral against a contract which say would have a 50% chance of a full loss and a 50% chance of no loss, the reinsurer would apply ‘reinsurance capital’ calculations and therefore allow for the benefit of the 50% chance of no loss and allow the fund to assume risk with this and therefore earn premium on it too. In essence it will only really be the leverage component that may reduce a little.
If the ILS fund chooses the latter path it will make it a more risky investment for the remaining investors as more of their capital / collateral is tied in ‘long-term’ investments. Further it may lead to a disjoint between cat bond pricing and reinsurance pricing whereby capital is artificially held within the reinsurance market depressing rates there and held out of the cat bond market enhancing yields by comparison to the reinsurance market.
Risk of Fund Gating
Going to the extreme, if the fronted reinsurance portfolio grows to a size where paying further redemptions from cat bond holdings becomes impossible, either due to the size of redemptions or a migration from hold liquid investments to illiquid investments, fund gating or widespread ‘side-pocketing’ becomes a reality.
Firstly it is impossible to rapidly unwind a fronted reinsurance portfolio and withdraw collateral (unlike private cat bonds which are ultimately tradable instruments) and secondly, as already explained, the unwind starts with the leverage component which does not release any collateral. Finally, once the risk period of the last contract in the portfolio expires, the repatriation of all collateral is unlikely due to the high likelihood of a portion of the contracts being blocked in their entirety due to the broader diversification of the portfolio, the whole reason why the ILS funds entered the fronted reinsurance space in the first place. (No diversification = no leverage).
Short-term financing of long-term investments is a dangerous strategy in any market and it is only a matter of time before the model breaks.
Over the last months we have seen increasing redemption in the ILS space and we have also seen increasing yields in the cat bond market, higher than those in the reinsurance market. The amount of the recent redemptions seen is probably unlikely to be large enough to be driving any market pricing today. However the question still remains, why are ILS funds not overweight cat bonds which are returning higher yields for the risk than reinsurance, especially when trading cat bonds is simple with no frictional costs and they are liquid, whereas reinsurance is complex with high frictional costs and totally illiquid? It will be more telling after 1 January 2016 once the main reinsurance contracts have been renewed and at what levels, and where the cat bond market is trading in relation. At this point one can maybe see the signs as to whether the funds are moving out of reinsurance and into cat bonds, or whether they remain in reinsurance to profit from leverage and not realising fully blocked collateral costs, but with it assuming the risks of potential gating should funds exit the market quickly as they did in Q4 2008.
Finally it must be stressed that it is important to understand that there is a definite place for reinsurance within an ILS portfolio. There will always be perils that one cannot find in the cat bond market, and contracts which do pay well, better than what is available in cat bonds. Ideally reinsurance contracts should transformed into tradeable products, such as private cat bonds, where each contract is separate from each other and can therefore be sold on the open market. Where fronters are used, there needs to be very well defined collateral release provisions between the fronter and the ILS fund and the fronter should not be in sole charge of releasing the collateral – release should be more mechanical and not based on judgement and whim. Finally leverage should be used carefully with a view to release and possible requirement to reduce underwriting.