Before I turn to the topics of reinsurance and

Transcription

1 Portfolio Diversification Benefits of Insurance-Linked Securities Laura Taylor, CFA Managing Principal and Chief Financial Officer Nephila Capital Ltd Hamilton, Bermuda Insurance-linked securities (ILS), such as catastrophe bonds and other types of risk-linked securitization, have moved considerably closer to mainstream investments since their inception in the mid-1990s. ILS continue to grow in terms of market size, scope of products, and investor profile. Although typical investors are still insurance companies and pension funds, other investors, such as sovereign wealth funds, endowments, and high-net-worth individuals, are also included and all are eager to enjoy the favorable diversification benefits of this unique asset class. Before I turn to the topics of reinsurance and insurance-linked securities, I want to start by answering a question I received today: How did you end up with this asset class 12 years ago and why do you find it so interesting? I started my career in banking in Bermuda, where I was first introduced to hedge funds. It was the early 2000s, and it was a pretty exciting time for hedge funds. They had been around for 15 years or so, but a lot of innovation was occurring, a lot of capital was coming to the market, a lot of ideas and new strategies were coming to life, and efficiencies were being sought and created. While I was working with KPMG, the cofounders of Nephila Capital were in the middle of a management buyout and aiming to build an independent business out on their own. They asked me to join to help build up their team. When I did, I had my introduction to reinsurance, which is insurance for insurance companies. What fascinated me about the asset class was that, in the midst of financial innovation and all of the efficiencies that were occurring in the capital markets, the insurance industry was doing things largely the way it had done them for hundreds of years, from the beginning of Lloyd s in London in The problem was that capital was not being brought to where it was needed most. At the same time, investors were seeking to diversify returns. was a way to provide capital to solve a problem that needed solving that is, matching capital to risk in the US insurance market, where risk is overabundant, and to do so in a This presentation comes from the 69th CFA Institute Annual Conference held in Montréal on 8 11 May 2016 in partnership with CFA Montréal. way that also provided an alternative asset class for investors. To understand how reinsurance works, let s start with the homeowner. In the majority of insurance markets, when homeowners take out a mortgage, they are required to buy a homeowner s insurance policy to cover potential losses because of fire, theft, water damage, or whatever else might incur a monetary loss to the homeowner. This process is fairly straightforward for such places as North Dakota but not for some other places like California. Insurers can sell homeowners policies across the United States, and in the process, they are diversifying the probability of a home burning down in California, which is independent from the probability of a home burning down in North Dakota. Selling one policy is not a problem. If insurers have small losses, say $250,000 or $500,000, they can rely on the law of large numbers for predictability. The insurers strive for low volatility, which makes low margins acceptable. One policy for fire in California, for example, gets absorbed in the insurer s portfolio. But when they start selling more and more of those policies, insurers end up with a large concentrated risk that could be wiped out by one natural catastrophic event. Then, the risk becomes very expensive. Insurers can retain a small amount of capital for those low-volatility risks, but once they have highvolatility risks, they have to retain more capital to keep their credit ratings. The cost of that additional capital is expensive, and the risk becomes much less attractive. 16 Fourth Quarter CFA Institute. All rights reserved. cfapubs.org

2 Portfolio Diversification Benefits of Insurance-Linked Securities It is in these situations that insurers buy an insurance policy from a reinsurance company. In reality, and depending on the size of the risk, insurers may approach a dozen or so reinsurers and spread that risk among all those different reinsurance companies. The reinsurance companies are either rated or collateralized. They are typically not selling protection against small, non-volatile homeowners risks but rather the large, volatile catastrophic risks. As a result, reinsurance companies also need to create a diversified portfolio with a variety of different types of risks, such as European windstorms, Japanese earthquakes, and US hurricanes and earthquakes. These reinsurers can take on only so much in each bucket of risk. This limits their ultimate risk appetite; in some cases, they try to address this by, in turn, buying a reinsurance policy from one or more competitor reinsurance companies, which is called retrocession reinsurance. You can imagine that this process gets constrained at a certain point; risk is changing hands within the industry but not benefiting from new capital support. The reinsurer does not want to be too far away from the original homeowner because the company may lose the transparency of the underlying risk. Moreover, quantifying the overall risk becomes increasingly difficult. Another problem is the promise to pay, which is fundamental to this relationship. The business does not involve any margin call; it relies on a trading relationship in which one insurer buys a policy from a reinsurer and trusts that the reinsurer will be there to pay if a big event occurs, the promise to pay. That trust is fine for a certain number of events, but if an event occurs that affects all companies at once, an insurance company may start to consider that it is actually holding more capital on its balance sheet than the reinsurance company does. Insurers trust the reinsurance company to be there to pay in a $40 billion or maybe a $50 billion event something large enough to cause an issue but not huge in the context of the industry capital base. But in a $200 billion event, insurers do not believe that there really is sufficient capital within those reinsurance balance sheets to pay all of their potential obligations. In that situation, demand for protection exceeds supply. The other factor at play is the rating agencies preference portfolio for a well-diversified portfolio with the risk spread out. The probability of a US hurricane of a certain intensity occurring in the same year as a European windstorm has a lower likelihood than for US hurricane risks alone. The ideal portfolio has evenly distributed risks in buckets of, for example, $25 billion each for US hurricane risk, US earthquake, Japanese earthquake, Japanese typhoons, European windstorms, cyclones or earthquakes in Australia, aviation risk, and satellite risk (which is basically the risk of a satellite falling out of the sky). But the reality is that the demand does not work that way. As Table 1 shows, the US need for insurance coverage far exceeds the other buckets. The primary reason is the large insured values in the United States, but another reason is the US proximity to hurricanes and earthquakes. The United States simply has more demand for coverage than exists in any of the other risk buckets. Each of the perils shown in Table 1 has the same probability of occurring and the same expected loss. For example, a $32 billion European wind event has the same probability as a $180 billion US hurricane risk. But because the demand for coverage of the US hurricane event far exceeds supply of capital, the reinsurer is going to be paid 11 times for that risk value, whereas the premium for the European risk is only 6 times the risk value. When the demand for coverage exceeds the supply of capital, reinsurers can charge more for that risk. Given the constraints of the cost of capital, the rating-agency constraint, and the credit risk, what happens to that excess demand for coverage? Essentially, only a small portion of the market is served and a large portion is not. When I joined Nephila in 2004, we had four hurricanes hit Florida that summer, two of which hit the same county. We had never seen that kind of frequency for hurricane landfall. A large number of hurricanes happened, but they were not of huge dollar value as far as losses were concerned. Everyone was paid for their losses, and no defaults occurred during the 2004 hurricane season (which officially runs from 1 June through 30 November). Then during the hurricane season of 2005, Hurricane Katrina hit with $40 billion claimed in August, Hurricane Rita with $5 billion in September, and Hurricane Wilma with $10 billion in late October. Table 1. Approximate Insured Loss to the Industry from 0.40% Expected Loss US: Hurricane US: Earthquake Japan: Earthquake Japan: Typhoon Europe: Windstorm Australia: All Perils Aviation Satellite Risk (%) Premium (%) Capital needed ($ billions) CFA Institute. All rights reserved. cfapubs.org Fourth Quarter

3 CFA Institute Conference Proceedings Quarterly The experience of 2004 and 2005 meant that two years of losses had occurred in a row, so a lot of capital had left the industry just because of loss payments. Balance sheets were not as strong as they once were. Moreover, rating agencies and risk managers, who had always been preparing for the big one, now realized they needed to have enough capital for multiple big ones. The industry needed more capital simply to increase the capital on the existing balance sheets to the levels where they once were, but also, the rating agency was demanding increased levels of capital. A huge need for capital existed throughout the reinsurance industry, which meant that there was not enough capital to reinsure those insurance policies. The other thing happening was that, after two years of heavy losses in a row, the nationwide insurance companies that were in the coastal US states started to pull out. This occurred not only in Florida but also in Texas and all the way up to the Northeast. So, at that time, less capital was in the market and insurance policies were not being sold. But people were still buying homes, still needing mortgages, and still being exposed to hurricane risk. What happened? Many states not just Florida stepped in and assumed the risk. The state government of Florida started selling a much larger number of US homeowners policies against hurricane risk. The model that was in place at that time, because the states also did not have enough money to insure those events, was that the states would hope to borrow (issue bonds) after the event. This method might work for a $10 billion event, a $40 billion event, or even a $70 billion event, but for a $180 billion event? A state would be unlikely to be able to step in to pay those losses. Another less prominent phenomenon was that people were self insuring. Snowbirds would go down to Florida for six months of the year, and many were on a limited or a fixed income. They were lucky enough to own two homes, and the majority of their retirement was, frankly, wrapped up in those large assets. They could not afford the price of a hurricane insurance policy. They did not have a mortgage remaining, so they self insured. They figured, Well, if the hurricane hits, we are out of a house. Obviously, this was a less than ideal situation for everyone. This industry needed two things. It needed capital with an appetite for cat (i.e., catastrophe) risk and diversification. And it needed a solution to the creditrisk problem mentioned earlier: How could buyers of protection be sure that they would be paid? Enter insurance-linked securities (ILS). Insurance-Linked Securities In their simplest form, ILS are an efficient way to bring capital to the US insurance market, particularly the insurance market for US catastrophic risk to property that is, protecting homeowners or businesses or buildings from natural catastrophes, such as hurricanes and earthquakes. Figure 1 is a typical ILS structure for a catastrophe bond, or cat bond. Cat bonds first came to the market in the early 1990s, following Hurricane Andrew, because of a similar disruption that led to the continuing need for capital at that point and the continuing credit-risk problem. These bonds were used as a way to sell coverage at levels exceeding the amounts in which typical reinsurance was being purchased at the time. Catastrophe risk is extremely capital intensive and thus extremely expensive for insurance companies to hold. Demand far exceeds supply for protection from catastrophe risk, so there is a healthy risk premium for the capital. These bonds also address the credit-risk concern arising not only from the big one but also from multiple events. Figure 1 shows that a special-purpose vehicle (SPV) is set up to sell a reinsurance policy to a US insurance company. The policy says, If you lose, say, $500 million from a $50 billion earthquake, I will pay you $500 million. The SPV does not have a balance sheet, nor does it have any assets, so it issues notes to its investors. Those investors buy $500 million worth of notes, which go into a trust. They are invested in cash or cash equivalents, typically US Treasuries or the equivalent, that roll over every three months. The policy is often in effect for a three year term, but the calculation of losses occurs every year. Figure 1. Typical ILS Structure for $50 Billion Earthquake Risk US Insurance Company X% Premium Contract SPV (T-bills or equivalent) Notes Coupon T + X% Investors Earthquake ($ notional) No Earthquake ($ notional) 18 Fourth Quarter CFA Institute. All rights reserved. cfapubs.org

4 Portfolio Diversification Benefits of Insurance-Linked Securities If an earthquake occurs, the US insurance company draws down the trust and gets paid its money. If there is no earthquake, investors get their money back. The US insurance company can view that trust. The company can see whether it is truly invested in T-bills and that the funds are secure and will be there if needed. This kind of cat bond accomplishes two things. It solves the credit-risk problem, and it also spreads the risk more widely than in previous structures. Because of the investor element, the industry has now gone from the relatively small constellation of reinsurance companies to a much broader universe of capital providers, including pension funds, sovereign wealth funds, and endowments. High net worth (HNW) individuals are also buying these instruments. These investors already hold a large portfolio that benefits from diversification through stocks, bonds, commodities, and other instruments but are still seeking uncorrelated assets. Because endowments and pension funds and other institutional investors can take concentrated risks as a result of their diversification elsewhere, they can go to where they are needed most and where the risk is priced the best, which is US hurricane and US earthquake risk. For these sophisticated investors, losing $500 million to one earthquake might result in losing 50 bps of annual return a very small percentage of their portfolios. Compare such a situation with that of a reinsurance company, where a $500 million loss could mean the end of the company. Risk is more easily absorbed on a much larger balance sheet (investors) than on a relatively small balance sheet (only reinsurers). ILS Performance for Investors How do these securities perform? We know that structural inefficiency creates a case for positive expected returns over time, but is it truly diversifying? How do the bonds perform in a traditional portfolio? Table 2 shows the correlation of the Eurekahedge ILS Index with indexes of a variety of asset classes. Over the 10-year time period measured, the correlations are basically zero. With those low correlations, the question now is how the ILS Index actually performed during periods of financial crisis or market instability in broader markets. ILS has proved to be a historically significant diversifier from broader financial markets. A financial crisis does not cause a hurricane, so a company like Nephila Capital performed quite well in 2007 through 2009, as can be seen in Table 3. The majority of our investors in were hedge funds of funds. They were finding, as many investors found, positive correlations between different asset classes, different currencies, different countries, and different instruments. They had a need for liquidity. Therefore, although the ILS Index was performing well, Nephila Capital experienced large redemptions throughout that period because it could provide liquidity. Most of our contracts expire twice a year, at the end of May and the end of December. We had redemptions that actually exceeded our gates, 1 but because we could provide that liquidity and still protect our investors who were remaining in the funds, we were able to provide those redemptions to our investors. After the 2008 financial crisis, when things began to calm and investors were looking to reallocate in 2009, pension funds started coming to us directly. They had previously allocated through funds of funds but now wanted to allocate to the ILS asset class directly. They could see the benefit of that noncorrelation. For example, at the end of 2009, Nephila Capital had about $2 billion in assets under management (AUM). By the end of 2012, AUM had grown to $8 billion. We grew dramatically in a short period, 1 A hedge fund s gate provision traditionally refers to the right to limit withdrawals on any specified withdrawal date to a maximum percentage of the fund s net assets. Table 2. Correlation of Eurekahedge ILS Index Monthly Returns with Returns of Various Indexes (1 January December 2015) US Stocks (S&P 500 Index) US Bonds (Barclays US Aggregate Bond Index) Hedge Funds (HFRI FoF Index) Commodities (S&P GSCI) ILS Index Table 3. ILS Index Relative Performance during Crisis Periods in Financial Markets Crisis Event Total Return Period ILS Index (%) US Stocks (%) US Bonds (%) Hedge Funds (%) Commodities (%) a 2015 Market volatility 1/Jan/ /Dec/ Yuan devaluation/equity market volatility 1/Aug/ /Sep/ Financial crisis 1/Jul/ /Feb/ a S&P Goldman Sachs Index (S&P GSCI) CFA Institute. All rights reserved. cfapubs.org Fourth Quarter

5 CFA Institute Conference Proceedings Quarterly and we were not the only ones growing in the industry. New types of structures were being born; new instruments were being introduced. People were finding different ways to allocate and help solve that capital need I mentioned. Table 4 shows how ILS has performed in a traditional portfolio. As you can see, with a 5% enhancement from ILS for the period, return went up. In addition, volatility went down. From Niche to Mainstream Mainstream might be an overstatement for the ILS asset class, but the class is certainly no longer a niche industry. We no longer have to explain the idea of reinsurance to clients. We have investors who come to us. They know the space. They may have already purchased some cat bonds on their own, and now they are looking to make an allocation to a manager. The industry has grown in quite a few ways. The first area is different types of instruments to solve different kinds of problems. The other ways relate to the types of investors now coming into the industry and the geographical spread. Panel A of Figure 2 lists some of the structures that are available, and those listed are just a few. The arrow indicates the extent of management. Panel B lists various instruments, and the arrow indicates strength of the barriers to entry. As far as structures go, typically, investors will allocate to catastrophe bonds first. The bonds are easy to understand and represent a somewhat passive way to be introduced to the asset class. Although the universe of available cat bonds is small, the variety of structures and instruments is growing. For example, sidecars are investment vehicles that participate in a slice of a reinsurance company s portfolio, rather than just what risks are packaged into cat bonds. When capital was coming into the space during , reinsurance Table 4. Measure Benefits to a Portfolio of Traditional Assets (1 January December 2015) Traditional Portfolio Portfolio Enhanced with 5% ILS Total return 69.7% 73.8% Annualized return 5.3% 5.6% Annualized volatility 8.0% 7.8% Growth of $1 million invested at inception $1,696, $1,738, companies were also trying to bring capital to their balance sheets and create ways to execute their strategies for different needs. Typically, sidecars are where the reinsurance companies cede a portion of their portfolio of risks to an SPV. A sidecar has a limited life; three years is a standard term. As a side note, hedge fund reinsurance (called hedge fund re or hotrod reinsurance) is a very different strategy from insuring property cat risk. On the underwriting side, it usually refers to the predictable law of large numbers type risks but with the addition of a more aggressive asset strategy to complement it. In other words, it is a classic noncatastrophe reinsurance strategy but with the traditional low-risk asset engine lifted out and replaced with a higher-octane investment plan. This is distinctly different from reinsurance hedge funds that are dedicated to property cat risk, or marine risk, or whatever the strategy might be. In these funds, a manager is actively managing the portfolios of risk. On the instrument side in Panel B of Figure 2, cat bonds are the most broadly invested. The majority of the reinsurance market today is still traded over the counter. There are the cat bonds you can buy through brokers, obviously, but most are tailored products for insurance companies that need that coverage. Figure 2. Today s Variety of ILS Structures and Instruments Catastrophe Bonds Low Sidecars A. Structures Hedge Fund Actively Managed Hedge Funds High B. Instruments Catastrophe Bonds Industry Loss Warranties Country- Weighted Industry Loss Quota Share Retrocession Traditional Nonsyndicated/ Private Transaction Low Barriers to Entry High 20 Fourth Quarter CFA Institute. All rights reserved. cfapubs.org

6 Portfolio Diversification Benefits of Insurance-Linked Securities The middle boxes are where industry events could trigger a loss. For instance, Hurricane Katrina triggered that $40 billion in losses; if that happens, then the ILS company pays. These instruments have something of an index option type structure. Farther to the right in Panel B are the instruments that are typically considered indemnity-type contracts. In these instruments, the insurance company or reinsurance company suffers a certain amount of loss, and then the (retro)cedent s layer is triggered based on the event losses and the companies pay those losses. Most transactions are syndicated, so multiple reinsurance companies are involved. The larger the reinsurance company, the more tailored the contracts it can write to a trading counterparty outside the syndicate. The company can create solutions just for that party, which is also known as a private deal. The breadth of investors has also grown. When Nephila Capital first started, funds of funds and hedge funds were the majority of our investor base. That has changed. Now, the investor base is mostly pension funds and endowments, which are now allocating directly to reinsurance. Family offices and HNW individuals are also allocating to the space. The asset class has also seen expansion geographically. Canadian pension plans were actually early adopters of this space. Our largest investor in 1999 was one of the Canadian pension plans, and it is still with us today. Even in 2004 and 2005, and then in 2006 when the market hardened, quite a few more allocated to us as well. We have had good partnerships with the Canadians. The use of ILS has spread throughout Europe and Australia. In addition, the Japanese pension funds allocate to this space. US defined benefit plans turned to this area a little later. Some corporate pension plans and public pension plans have begun allocating to ILS in the past three years or so. Conclusion The US insurance market (and other markets globally) have a need for capital, and this market was not being served by the existing structure in this century. So, ILS providers stepped in. ILS products allocate capital efficiently while providing positive returns for investors returns that offer true diversification because they are not correlated with returns of the traditional asset classes. CE Qualified Activity 0.5 CE credit 2016 CFA Institute. All rights reserved. cfapubs.org Fourth Quarter

7 CFA Institute Conference Proceedings Quarterly Question and Answer Session Laura Taylor, CFA Question: Should investors be concerned about conflicts of interest? ILS funds that are managed by reinsurers don t seem to have any skin in the game. For example, are insurance companies laying their bad risks with insurers onto investors? Taylor: That is a legitimate concern. In their due diligence, this concern is where most investors spend the majority of their efforts. My understanding is that reinsurance companies work to solve this problem through alignment of interests. They will invest in that sidecar, that other vehicle, alongside investors. Also, sometimes reinsurers bring in independent parties to verify that they are following their allocation policy and treating investors fairly. If you are thinking of making an allocation to this space, I recommend that you listen to quarterly earnings calls in which the CEOs typically talk about their alternative capital vehicle. You can get a feel for the fairness and how they are addressing it. Question: What are your thoughts on the capacity of this asset class? Can you talk about the future for it? Taylor: Following the credit crisis, a lot of capital came into the reinsurance market, which is about $450 billion, so it felt the impact of that capital. The reinsurance market is easier to enter than the insurance market. The US insurance market is trillions of dollars and has not experienced the same amount of capital coming into it. The risk in that huge US insurance market is still not being transferred. Recall how in Florida, a person is required to get a hurricane insurance policy to qualify for a mortgage. In California, you are not required to buy an earthquake policy. A young man from California I spoke to at a conference said, Earthquake insurance is just not worth buying. It does not satisfy my needs, and it is superexpensive. It will never trigger, and the market is not being serviced. There is absolutely opportunity in that market, but it would be a hard market to enter because of large barriers to entry. That market will take time to develop, but it will require large catastrophe capacity. This is also true on other geographies (e.g., such emerging markets as China) and risks (temperature and rainfall). Question: What is the best way for retail investors to access ILS? Are there vehicles or funds they can use? Taylor: Some UCITS (undertakings for collective investment in transferable securities) funds are available in Europe, and some US retail vehicles are available. Those are a very different space from the one my firm is in. We do not have a retail product. We work only with institutional investors who really understand the risk that they are allocating capital to. Question: What has been the level of returns in the past, say, three or four years in the market? Taylor: Anyone reading up on the reinsurance industry recognizes that a softening has taken place in the market, but a risk premium still exists for US insurance risk. European windstorm and Japanese typhoon risks were the first market to soften and then such markets as marine and aviation risks. For a fund that was making around 10% 12% earlier, returns are probably closer to 7% 9% now. Question: In the past, ILS worked well, but what is the probability of catastrophes in the future in light of the outlook for global warming and more severe weather? Taylor: A lot of research experts in the industry can address this question better than I can. But I can tell you that we look at the conditions for hurricanes every year. In some years, despite global warming, the Atlantic Ocean is not very warm and El Niño or La Niña will occur. Those conditions will either help or hurt you, and you can adjust your strategy accordingly. We also have a general weather strategy. It is not catastrophe based but more about rainfall, snowfall, that type of thing. We are spending a lot of time working with wind farms, for instance, and selling them cover so they can get the financing they need to start their projects. Question: This is a catchall question: What can go wrong? Taylor: There are two ways to answer that question. The first is what can go wrong that you could expect. This asset class is one where you spend a lot of time with investors talking about exposure and potential losses. When the wind blows and a hurricane makes landfall, there is a chance investors are going to lose money. For an investor who is thinking about allocating assets to this space, I would spend most of the time understanding the exposures how much are you going to lose in a particular type of event in any particular state? Once the investor understands the issue, the investor can ask the management firm what it is 22 Fourth Quarter CFA Institute. All rights reserved. cfapubs.org

8 Q&A: Taylor doing as a part of its underwriting strategy based on the market and climate conditions. Then, the investor should consider data quality. For all the companies underwriting US insurance, you receive data on the addresses, types of house, types of roof, and whether the structure is made out of wood or made out of stone, and so on. The approach is to put the science of risk (hurricanes) together with the underlying exposures (properties) and then predict what the losses will be and the insured values of the losses. How up to date are the insured values? What assumptions is the company making so that it knows in an event of such and such a type it is going to lose such and such amount of money? The other part is what can go wrong that you do not expect. We saw in 2007 that, although the majority of the asset class was fine, some incidental and unexpected correlation occurred. The way cat bonds were previously structured is that the investment bank, which was involved in structuring the bond itself, had a swap in place between the SPV and the investment bank. The investment bank would guarantee a LIBOR return on those assets. If the return was more than LIBOR, the investment bank would keep it. If it was less than LIBOR meaning there were losses in the portfolio the investment bank would step in. Like other asset classes, the quality of those assets deteriorated over the years, and we found that in most cases where portfolios experienced losses, the investment bank was able to step in and make the investors whole. For a few bonds that Lehman Brothers was involved in, the assets were not there. Those investors were paid pennies on the dollar, and the investment bank was not there to step in. At that point, a bit of a shift took place in the industry. Investors made it clear we weren t willing to take on this asset risk, and the structures went back to those money market funds or US cash and cash equivalents. If diversification is truly what you are looking for, make sure the ILS company is doing what it says it s doing and investing in cash and cash equivalents for protection. Question: What fees are involved in this product and what are the typical ranges for them? Taylor: A typical fee structure for an alpha, or actively managed, strategy would be about 1.5% annually with a 15% performance fee. I am unsure about what the retail products charge. If you are new to the space and want to allocate to a beta structure, there would be no performance fee and the annual expense would be about 1%. Question: What other factors specifically, negative comments made by Warren Buffett about pricing in reinsurance are driving the pricing of securities? Taylor: Demand is a big part of the price softening as well as data quality. Insurers that do not have good data are charged more. Historically, the reinsurance market has experienced a lot of cyclical trends. A market would soften, but then an event would happen, and a hardening would occur. And there were a lot of the two extremes. US insurance companies that were pulling out of the coastal states found there could be no certainty of pricing over the reinsurance they wanted to buy. Of the $450 billion reinsurance market, about $70 billion now is ILS. It is much easier for capital to come in after a large event and solve a capacity problem or leave if the return hurdles that investors need are not met. So, the market hardens after the event. Now, the hardening will probably be a short period because we are in a secular trend. We called it a novelty premium for the asset class in the early days, and over time, that has naturally gone away as the investment is more broadly accepted. We still believe, and our investors still believe, that the current risk premium (after the loss of novelty) is sufficient for the risk they are taking, although others may disagree. We closed our funds about three years ago. We had been growing, but when we hit $10 billion, we decided to stay steady, and we have been investing in other ways to access risk and grow our funds. We are opening up now. Opportunity absolutely still exists in the US insurance market. Question: How could someone benchmark managers? Is there an established industry benchmark that someone could look at in terms of comparing managers in the space? Taylor: Comparisons would be difficult. There is only the Eurekahedge ILS Index. People in the industry are trying to create indexes to solve this problem. The tough part is that investors don t know how good their investment managers are until an event occurs: They only see returns but not risk. Only after an event can they validate whether they lost the amount of money the manager said they were going to lose or whether the loss exceeded expectations. If you are about to allocate to multiple managers, you need to ask the managers to show you their performance, their exposure, and their out-of-the-box or unadjusted (third-party vendor) modeling. That information will give you a baseline for comparing managers and for ascertaining how the risk looks before the adjustments are made so you can evaluate risk-adjusted returns. Question: What will the prospect of rising interest rates do to ILS allocations? Taylor: We have an investor summit every two years, and we ask whether increased interest rates 2016 CFA Institute. All rights reserved. cfapubs.org Fourth Quarter

9 CFA Institute Conference Proceedings Quarterly will mean you change your allocation to ILS. The answer has always been the same that if you assume there is no other change in expected core returns, there would not be a hardening of the market. As a matter of fact, if interest rates go up, the interest earned on the underlying assets goes up, and we are invested in cash and cash equivalents, so there is some natural flow of yield through to investors (though that depends on the investor s assets and the managers themselves). Most of our investors are allocating strategically; they are looking for the diversification benefit. Of course, they want to be paid for the risks they are taking, but for them, they are seeking non-correlation, and this intent doesn t change if interest rates change. Question: A concern is that once the industry has a big loss event, people will flee and not come back to the market. What are your thoughts on how this asset class is evolving in that regard? Taylor: We asked our investors the same question. Most respondents said that they would increase their allocation. If there is a hardening of the market, then that means you are getting paid more for the exact same risk that you were just taking. Any rational investor, especially an institutional investor, would allocate more money at that particular time. Once pricing softened again, they might take that tactical allocation back, but they are making investment decisions like everyone else. A lot of our investors, too, have suffered losses. We have investors who have been with us since 1999 or 2005, investors who lost money in Hurricane Sandy in There was an earthquake in Japan not too long ago. New Zealand suffered: There were three earthquakes there all in one short period of time. Investors have experience in paying losses. There was a European windstorm at the end of last year. Investors understand that like all financial products, this one can have losses, and they are staying and continuing to allocate to the space. We actually raised money in 2011 because there was a short period of time when people needed more Japanese cover. We then returned that money once the market stabilized. Whether they allocate to us or someone else, I think the capital has found the need for it and gets paid for it. It is here to stay. Question: Is part of the attractiveness now driven by low yields, or are allocations strategic? Taylor: It may vary for a retail product versus a pension fund, but for our investors, the majority of allocations are strategic. We do have long standing relationships with some investors who allocate to us only after a big event. They stepped in in 2006 after 2005 and 2004 hurricanes. They also stepped in in Even though no natural cat event had occurred, the market hardened because the insurance companies and reinsurance companies had lost money on the asset side of the business. Suddenly, there was a hardening of rates in the market because of a financial event. 24 Fourth Quarter CFA Institute. All rights reserved. cfapubs.org