In Pensions Age’s latest webinar, Peter Nakada, Managing Director of RiskMarkets at RMS, gives an overview of insurance linked securities and why they offer interesting opportunities for pension funds today

In our opinion, insurance linked securities (ILS) are some of the investment world’s best kept secrets. They have attractive yields, they have low or, you might even say, no correlation and they are relatively immune from liquidity contagion.

We often use the terms cat bonds and ILS relatively interchangeably. The overall reinsurance market, that is catastrophe risk transferred from primary insurers away to other providers of capital, is greater than $200 billion today. The collateralised reinsurance market, which is part of the ILS market, is in the order of $25 billion and then cat bonds themselves, which are the most liquid of the ILS, is $14 billion.

So why are insurance companies tapping the capital market? The starting point for all of this is that catastrophe risk itself has well outpaced inflation. Catastrophe risk is increasing so dramatically, but more because of the exposure to catastrophes, and less about the frequency and severity of catastrophes increasing that much. Essentially what’s happening there is more and more people want to build expensive homes and businesses in coastal areas that are exposed to hurricanes or areas that are on earthquake faults.

For example, the odds of a $100 billion hurricane hitting Florida is one in 30. If that would happen to the reinsurance capital base, you would have a 60 per cent hit to the cat reinsurance capital base, widespread insolvencies and a market that would struggle to function immediately following that event. Imagine on the other hand if that risk, instead of sitting on the balance sheet of a handful of re-insurers in Bermuda and Europe, had transferred into the capital markets, onto the balance sheets of pension funds. That $100bn would represent a 5 per cent reduction in pension funds’ assets and that’s something that the pension fund market could take in its stride much more so than the reinsurance market.

The second thing is that catastrophe risk is extremely concentrated. Hurricane is about 50 per cent of the total catastrophe risk and within hurricane Florida is more than 75 per cent of that risk. We’re talking about very concentrated risk, so roughly 15 per cent of a typical reinsurer’s entire balance sheet is invested in Florida hurricane.

Contrast that then with a typical pension fund allocation to hurricane, which would be 2 per cent. This would be in a portfolio of typically stocks, bonds and real estate, maybe some alternatives, which would completely be uncorrelated with that risk.

Prior to 2008 cat bonds were structured as follows: The issuer of a cat bond would go out and issue, say, $200 million worth of bonds. That money was then taken from investors and put into a collateral account where it was invested in ‘investment grade assets’.

So there were investment guidelines as to what you could put in there, but they were basically guidelines based on the rating of that collateral. As we all know there were many investment grade mortgage related securities that ended up being not so investment grade, as we saw the global financial crisis unwind. So what was supposed to be a bond that triggered based on catastrophes only, also had an element of credit risk in there.

At that time Lehman Brothers did a total return swap with the collateral account guaranteeing the principal amount and people were saying ‘what are the odds that you have an impairment in the collateral account and the bank you’re offering their total return swap goes bust?’. For those handful of bonds it was 100 per cent because that’s what happened. The collateral account took a significant discount and Lehman Brothers defaulted and investors said ‘wait a second, this is not what I signed up for’. Going forward, after that, the credit risk was removed from the collateral account and invested in treasury money markets securities.

So how does this represent opportunities for investors? To simplify things, let’s look at ILS with 1% expected loss. If you had 20 per cent of your portfolio invested in cat bonds at a 1 per cent expected loss, you would need to earn close to 8 per cent over Libor in order to produce a 1 per cent information ratio. That’s because you have a lot of concentration, the risk can’t get diversified away that much. But if you have a 2 per cent allocation in your portfolio, all you would need is a return of 2 per cent over Libor in order to earn a one time’s information ratio.

Why the big difference? What’s happening in the 2 per cent portfolio concentration case is the lion share of that risk is getting diversified away. The smaller the allocation of a non-correlated asset, the more of the risk is diversified away. So there’s where the attractive yield part comes from.
So given all this the sceptics among you might say ‘how do we know that those expected losses are reliable?’ Isn’t that just some fancy modelling, and hasn’t modelling been discredited after the global financial crisis? To which we would respond: not really. The models that we are using to analyse the expected loss are driven by science and not statistics. We’re not just looking backwards to see how things performed in the past. We’re actually modelling ground up the drivers of the cat risk and the behaviour of the exposures of cat risk. Let me give you an example. For hurricane risk we actually model the behaviour of hurricanes in terms of the wind speed and the size of the hurricane as it progresses over land. So we model 16,000 hypothetical hurricanes in a giant simulation and then we model what happens to a house with a given construction when the wind blows 100 m/hr on it, how much damage would be there.

Investors might also say: ‘Do I trust that the people who do this for a living aren’t transferring the worst risks to me?’ The answer to that is that you should take comfort in the fact that the models that we use to analyse cat risk for investors for ILS are exactly the same models that insurers and reinsurers are using themselves to measure and manage their own risk.
So there isn’t really an information asymmetry there. The investors have pretty much the same tools as the insurers and reinsurers. Another point that should give you comfort is that there are three independent cat modelling firms to keep the market honest.

So – what can go wrong? One risk is potentially badly structured bonds and concentrated risk. The badly structured bonds come in two flavours: one is poor collateral structure, for example, Lehman Brothers, and that’s been cleaned up. The second is poor trigger design. Poor trigger design relates to what measure of losses caused a bond to trigger. So be selective, don’t buy the market. If you had looked at the Swiss Re cat bond total return index from last year, it returned only 3.6 per cent, while 8 per cent return is the norm. At least 3 per cent of that was lost just due to badly structured bonds and these bonds had problems well known at the time. Another thing to do is to avoid concentration risk and diversify. If you buy the market, you only have a portion of this in any one peril and the perils are themselves uncorrelated with each other.

So to sum up, we believe that this is really an under-appreciated asset class and a lot of that is due to the fact that it has a non-familiarity premium rather than a risk premium. It means that early movers in this asset class would get excess return.

In association with RMS

Share Story: