What are realistic long-term return expectations for ILS?


Insurance linked securities (ILS) are generally, and correctly, acknowledged to be a diversifying  asset class. ILS correlation with traditional financial assets is generally low. However, the novelty and structured nature of ILS means they often get put into the “alternatives” bucket by asset allocators.

This association brings certain return expectations, which we think may mislead some investors. What are realistic return expectations? And can they be sustainably delivered?

What drives ILS returns?

Catastrophe bonds and other collateralized reinsurance contracts provide protection against tail risk. They are essentially fixed income securities that have a “short put” return profile. That is to say, the maximum return one can achieve (unlevered) is the premium received for selling the protection. The downside can be up to the full notional of the trade for which the premium was received.

This raises the question of what can realistically be expected of a strategy of selling tail risk protection.

Insurance operates on the idea of building broad diversified portfolios of risk. It relies upon the “law of large numbers” to provide risk transfer services to the individuals buying protection at a reasonable price. This works so long as the risks insured are independent of one another.

Insurers must assess the degree to which new exposure correlates with the loss behavior of the existing insurance risks in the portfolio. Once an insurer is faced with a concentration of exposures, each additional unit of exposure assumed increases the marginal additional capital required.

For example, if an insurance writer covers many properties in the same region, a large portion of that portfolio could be affected by the same event. For this reason insurers and reinsurers (henceforth collectively referred to as (re-)insurers) may find it more efficient to transfer some of their concentration risk to another counterparty.

There are a few parts of the world where the concentration of insured exposures is so great that the amounts of protection sought are beyond the reinsurance industry’s ability to fully absorb the demand. These regions are commonly referred to as “Peak Risk” zones. They include:

  • The US Atlantic Coast (most specifically Florida and Gulf Coast hurricane)
  • The US West Coast (earthquake)
  • Japan (earthquake and typhoon)
  • Northern Europe (winter storm)

For these areas there comes a point where each additional dollar of exposure assumed would require an additional dollar of capital if the (re-)insurer is to maintain a capital adequacy that is compatible with the regulatory and rating agency requirements.

It is these marginal additional capital requirements that determine, at the end of the day, what sort of price is charged for the transfer of catastrophe risk. This price is commonly referred to in the (re-)insurance industry as the premium.

The components of risk transfer premium

The premium for risk transfer comprises three main components:

  1. Expected loss
  2. The (re-)insurer’s expenses
  3. The loading for profit.

If an additional unit of exposure adds significantly to the diversification within a (re-)insurer’s portfolio, the (re-)insurer will require only a marginal amount of additional capital. This results in a lower risk transfer premium required to meet the reinsurer’s target return on capital.

Conversely, adding additional peak risk exposure requiring a dollar of capital for each dollar of protection provided would require a risk transfer premium that generates an expected return or profit margin that is at least equivalent to the reinsurer’s desired return on capital. Otherwise such additional business would dilute the reinsurer’s overall return.

This explains why non-peak risk regions attract relatively lower reinsurance prices than protections purchased covering peak risk exposure. The table below illustrates the above.

ILS_returns_calculation.JPG

For illustrative purposes only. Source: Schroders Capital

So what is a sustainable level of return?

A (re-)insurance company’s balance sheet contains a number of risks that consume capital. These are essentially broken down into three areas of risk.

  • Insurance risk
  • Investment risk
  • Operational risk.

The capital requirements for a (re-)insurance company are a function of the riskiness of the individual sub-components within the balance sheet.

ILS instruments are exclusively focused on the insurance risk component, so the benchmark for potential returns follows the target returns that (re-)insurers use in managing that business. Reinsurers typically target returns on equity of between 700 basis points (bps) and 900 bps over a given “risk free” benchmark[1].  Therefore, for a peak risk exposure the required return would have to be equal to or greater than the above benchmark. Note also that such returns are aimed for on a so-called “through the cycle” basis which assumes a certain degree of volatility around the target from year to year. Volatility will depend on loss activity and the underlying pricing cycle characteristic of the non-life (re-) insurance market.

The ILS market emerged in the 1990s. ILS provided a supplemental pool of risk-bearing capital amid a recognition that demand for peak risk protection exceeded the appetite of (re-)insurance companies’ combined balance sheets. As a diversifying asset class, ILS could theoretically be priced at a level that was lower than the marginal cost of capital required by professional reinsurers.

At the same time, and from the point of view of ILS investors, the target returns described above represent the theoretical cap on the sustainable level of return for the pure insurance risk in fully collateralized ILS because higher returns would result in additional capital entering the reinsurance industry. An additional component of the return ILS investors receive is whatever return can be generated on the collateral securing the ILS instrument’s obligation. In recent years this has been effectively zero, but in the past was as high as 5.5% in 2007[2] and may again increase as central banks increase their reference rates.

Has the asset class has lived up to expectations?

Analysis[3] by Lane Financial LLC found the compound insurance return (i.e. excluding any return on collateral) for the 20 year period from January 1, 2002 to December 31, 2021 was 5.9%.

This makes sense. The universe of perils insured is indeed dominated by US hurricane (peak risk) but also contains some diversifying perils at lower yields. Factoring those in, one would assume the resulting performance might be close to the lower end of the 700-900 bps range cited above, and it has been.

When looking at historic returns, one must also bear in mind that certain instruments were impacted by perils that were either not originally modeled or poorly modeled. They were therefore not considered when setting the price of the trade. Examples include wildfire, winter freeze or severe convective storm.  

We would argue that on a risk-adjusted basis, and compensating for previously unmodeled perils, the ILS market is offering returns close to reinsurers’ own cost of capital. As discussed in an earlier paper, the reinsurance market currently is experiencing a so-called “hard market” with a significant mismatch between supply and demand. This has led spreads to rise to their highest levels since 2009. Over time, we would expect capital to be drawn back into the sector allowing returns to revert to their longer term mean.

The short track record of data is also a challenge here. The sector - in the form of catastrophe bonds for which there is publicly available data -  only has a track record going back around 20 years. Not a bad start, but in a market selling protection against events that may have event return periods as high as 50 to 100 years, in some cases even longer, it’s perhaps not yet enough.

Can ILS managers achieve better returns?

In our view, there are only two potential avenues to boost target returns.

Either one can:

  1. Position the portfolio more aggressively and hope that no major losses occur, or
  2. Use financial or structural leverage to boost returns[4].

Option 1 is a strategy that was offered by certain managers and did generate high returns in the years 2006 to 2016, but which proved not to be sustainable as loss activity reemerged in the years since 2017. Option 2 is a strategy that is available and used by many managers. Here one needs to keep in mind that the use of leverage can increase the volatility of results and the potential for a total loss in adverse scenarios. The ultimate question of whether leverage can improve performance sustainably depends upon the costs of such a facility and the risk appetite of investors. 

Pure diversifier

ILS are a structured product that enable investors to gain exposure to pure insurance risk as a new asset class. The structured nature of the instruments results in many investors assigning the instruments into their “alternatives” bucket. Another reason for this allocation may be the reliance on specialist managers with the necessary domain knowledge.

However, the performance of ILS does not exhibit a high degree of correlation to the other “alternatives” or, for that matter, to the traditional asset classes. This has been demonstrated through both the Global Financial Crisis in 2008, the Covid-19 shock in March 2020 and again in February 2022. It is a pure diversifier that has generated consistent positive absolute returns since its inception some 25 years ago. As we have outlined above, there is a natural cap to the maximum expected return for a fully collateralised strategy which is dictated by the (re-)insurance industry’s cost of capital.

Achieving higher returns is possible, but at the cost of increased risk. The ultimate return derived is a function of the investor’s risk appetite and tolerance for large drawdowns.      

 

[1] Swiss Re Target ROE 700 bps + Risk Free (10 year Government Bonds); Hannover Re Target ROE 900 bps + Risk Free (rolling five year average of 10 year Government Bonds); Munich Re Ambition 2025 ROE 12-14% (no explicit spread over RF benchmark). Source: Company Presentations (Swiss Re Group Company presentation as at April 2021; Hannover Re company presentation June 2022; The Munich Re Group Equity Story June 2022) 

[2] Trade Notes_Quarterly Market Performance Report Q4 2021, December 31, 2021; Lane Financial LLC

[3]. Trade Notes_Quarterly Market Performance Report Q4 2021, December 31, 2021; Lane Financial LLC. For a more complete discussion and description of the Lane Financial analysis we recommend visiting their website (lanefinancialllc.com).    

[4] By financial leverage we simply mean hypothecating the portfolio with additional debt, whereas by structural leverage we are referring to arrangements a portion of the tail risk is reassumed either by the counterparty or by a fronting reinsurer who does not require 100% collateral for each dollar of exposure written. This is typically accomplished by the fronting reinsurer requiring collateral for the portfolio expected loss for an agreed return period (usually between 1 in 200 and 1 in 500 years) and then taking on any residual tail risk for a fee.   


Contributes to
Unstructured Learning Time

CPD Accredited