Model sophistication


Model sophistication

Catastrophe modelling firms are increasingly sophisticated stakeholders in the ILS space, driving innovation, diversification and a greater understanding of risk transactions.

The role of catastrophe modelling firms in insurance-linked securities (ILS) transactions continues to evolve and deepen, as the market increases in complexity and looks for still greater insights into the evaluation and pricing of risk.

“While modelling companies continue to deliver detailed descriptions of the models and any assumptions used; analyses which estimate how often investors may lose principal; on average, how much principal investors may lose; and descriptions of the drivers of these loss estimates, modelling firms’ involvement has grown increasingly sophisticated,” said Rhodri Lane, manager, ILS at AIR Worldwide.

Modelling firms are also now involved in explaining modelling methodologies in rated ILS transactions, in the provision of remodelling tools to investors in order to provide a detailed review of transactions as they come to market, and in providing comprehensive portfolio management tools and transaction details.

“AIR’s CATRADER platform has become the industry standard for ILS, and the remodelling service AIR provides through this platform helps investors make informed, measured investment decisions both as transactions are marketed and for long-term portfolio management,” says Lane.

As Peter Nakada, managing director, capital markets at RMS, explained: “It used to be that our role was only to offer expertised risk analysis, working with the issuer to model the transaction and publish the risk analysis in the offering material.

“Now the focus is increasingly on providing services to investors in order to help them understand risk-adjusted returns,” said Nakada.

He added that software such as RMS’s Miu—the company’s ILS portfolio management platform—is enabling investors to understand cat bond risk more qualitatively and placing cat modelling firms front and centre in the transactional process. “Our role is more like that of an analyst, with the kind of analysis we do very similar to what you get when investing in equity or corporate bonds.”

Devil in the detail   

Helping to spur development is the increasing sophistication of ILS investors. As Nakada made clear, ILS fund managers are looking for ever more detailed model data to help them with portfolio optimisation. “Dedicated funds are moving to detailed loss modelling. Aggregate loss modelling is in many instances no longer sufficient.”

Lane added that “catastrophe models have contributed to attracting all forms of capital into the space and the sophistication of the market has as a result elevated over time”. He explained that new and experienced investors alike are growing increasingly sophisticated in the ILS marketplace. Models are helping them to achieve this: AIR makes available to its ILS-investor clients a semi-annually updated comprehensive database of modelling files and transaction details of all active catastrophe bonds, which supports portfolio management and assists secondary market trading opportunities, explained Lane.

“A key concept is if risk can be estimated, it can be priced, transferred, and traded. Different sources of capital—be it traditional or convergence capital—leverage catastrophe models to help them formulate their respective views of risk. Having different views, experiences, strategies, and capital structures is what makes a fluid and dynamic market, and we at AIR feel catastrophe models support this objective,” said Lane.

Nakada said that the trend towards more sophisticated investors has proved beneficial for RMS, which launched its own cloud-based analytical platform RMS(one) in April. The system enables investors to access detailed company loss data, with controlled access limiting the availability of sensitive company information, explained Nakada. Using an open platform approach, companies and investors will be able to ‘plug and play’ different parameter assumptions and even alternative risk models to assist with risk assessment and portfolio optimisation.  

The open platform approach is intended to encourage other modelling firms to input their view of risk on to RMS(one), said Nakada, with this approach helping to establish “one view of exposure, but multiple views of risk”. The hope is that working from a single platform will simplify the risk assessment process for investors.

He who pays the piper

Another important evolution in the ILS space is a move towards greater emphasis on an ‘investor pays’ approach, said Nakada. “An ‘issuer pays’ approach to ILS has been one of the key factors holding back growth in the marketplace,” he said, describing the move towards an investor-centric approach to ILS transactions as a healthy development for the industry. “The development is important because of the moral hazard inherent in the ‘issuer pays’ model. We saw this in the financial crisis with the rating agencies.

“Under the current approach, issuers get to choose the cat modelling firm and bear all of the costs of the modelling. This arrangement creates the temptation for the issuer to pick the modelling firm that is going to characterise their risk as the lowest. In addition, the modelling firm could well feel pressure to provide results that are acceptable to the issuer.”

Nakada said that the industry would benefit from an approach in which the emphasis is upon investors to pay and partially pay for the modelling of transactions. “Such an approach has the incentives the right way around. Investors need to be comfortable that they are getting an accurate view of risk,” he said. Investors can also bear some of the financial and administrative burden of the transactions, particularly in light of levels of investor appetite in the space. This would in turn help to bring more transactions to market, particularly among smaller issuers.

Nakada explained that the Tradewynd Re transaction issued by AIG was the first deal to take a more ‘investor pays’-type approach. “The important innovation in that transaction was that AIG provided us and the other modelling firms with detailed exposure data and allowed us to provide portfolio management insights to our clients that you don’t typically get in a standard risk analysis,” he said. This in turn enabled the client to understand how the deal would fit within their portfolio, “and ultimately it made that deal more affordable to issue for AIG”. There are a number of similar transactions in the works, he said.

Whose finger is on the trigger?  

Addressing developments in trigger type within the ILS market, Lane said that it is worth considering that it is very much a sponsor’s market at present. “Minimising basis risk is typically a key goal for most issuances and indemnity triggers achieve this goal. Furthermore, pricing between an industry index trigger and indemnity pricing has tightened, so there is little incentive to take the basis risk.” It is this buyer’s market that appears to be one of the lead drivers behind rising interest in indemnity transactions.

Lane said that insurers able to provide detailed exposure data are tending to pursue indemnity structures. He added that repeat issuers are also favouring a move towards indemnity structures. “The key is being able to thoroughly describe the exposure data in such a way that the modelling represents the actual risk in as robust a fashion as possible,” explained Lane.

Lane is, however, agnostic when it comes to trigger type. As he explained: “AIR is not looking to drive the market from one transaction type to another, rather we are there to support what our clients’ needs are for their risk financing programmes.” He said that when the market hardens, there may yet be a change in the preponderance of trigger types.

Nakada has a rather different view of trigger development. He said that insureds will ultimately pursue indemnity transactions, even if some will remain loyal to parametric triggers that deliver ease and speed of payout. “Ultimate end demand is for indemnity,” he said.

Nakada said that indemnity triggers will continue to proliferate thanks to two key drivers. First, it is getting easier for investors to model and understand indemnity risk. Much of this is down to developments led by the modelling companies themselves. Second, he sees a natural evolution from parametric to indemnity-based transactions as data granularity is improved, with parametric triggers the forerunners to more complex transactions in emerging geographies.

He suggested that such an approach may well favour more sophisticated investors too, saying: “Those with detailed modelling capabilities will have the ability to understand indemnity transactions and in that way gain competitive insights into ILS transactions.” He said that those investors who don’t adopt detailed modelling will tend to be those that are too small to purchase the necessary software and will likely become “following markets” in the sector as a result.

Turning to the development of ILS transactions in emerging markets, Nakada believes that parametric triggers will predominate in such geographies, but as markets develop this will change. “For perils such as Asian flood where exposures and vulnerabilities are less well-known, it makes a lot of sense that early ILS coverage is parametric. I suspect that China flood may well develop a robust parametric market,” said Nakada.

The development of emerging market ILS has encouraged RMS “to take a more agile approach—releasing hazard models in advance of good exposure and vulnerability models so that we can support things like parametric transactions,” explained Nakada, but he believes that there will be an inevitable tilt towards indemnity deals.

Lane agreed that parametric transactions would likely be the forerunner of more sophisticated ILS triggers in emerging geographies, adding that in many instances “the data simply does not support indemnity modelling”. He added: “The challenges associated with bringing any new risks, regardless of trigger type, to market is that investors will scrutinise the initial risk analysis in the same manner they do with a more familiar transaction—probably more so. Is the data there?  Can it be modelled appropriately? And, does the risk being transferred lend itself to the trigger type?”

Future innovation

Trigger sophistication and creativity, meanwhile, continue to develop apace. In 2013 combined trigger transactions were introduced. These spanned business segments, such as a transaction that extended an indemnity triggers for the issuer’s insurance business and an industry index trigger for its reinsurance business, said Lane. Also in 2013, variable reset provisions “which allow for flexibility of coupon payment during the life of the transaction” were introduced.

Lane said that bonds have also been placed lower down sponsors’ reinsurance towers, with one transaction in 2013 even securitising a working layer. He added that he expects this development to continue as cedants “continue to diversify their reinsurance programmes”. Looking ahead, Lane predicted a “proliferation of direct placements from non-insurance entities”, particularly in areas such as US flood where state provision is drawing back and where model granularity is improving. He also suggested that an expansion of region-peril combinations in emerging markets based in parametric triggers is likely. 

ILS modelling, RMS, AIR, cat bonds

Bermuda Re