Thinking inside the box: parametric risk and ILS
The definition of parametric insurance is reasonably straightforward—a non-traditional product that offers pre-specified payouts based upon a trigger event—but there is work to do in explaining its potential value, according to a debate at ILS Bermuda’s Convergence conference. The panel discussion was titled “Thinking inside the box”.
The panellists were Steen Byskov, senior financial officer at The World Bank; Guillermo Franco, managing director & global head of cat risk research at Guy Carpenter; Raveem Ismail, head of parametric underwriting at ASR Re; and James Doona, managing director of Capital Partners at Munich Re America. The moderator was Jeff Manson, SVP underwriting and head of global public sector partnership at RenaissanceRe.
“We can make that happen because all of the analytical work is front-loaded.”
Ismail stressed how a parametric product differs from traditional insurance cover in that a payout directly follows an event, meaning many of the intervening steps are obviated.
“We can make that happen because all of the analytical work is front-loaded, so that the right metric is chosen and the right threshold at which it triggers is formulated. Otherwise, the industry will end up paying everybody or nobody,” he said.
Parametric products can thus be extremely broad and are no longer confined to huge natural disasters, entire countries being the buyers and very large reinsurers being the sellers.
“For any event that can be objectively verified, we can make it a parametric product. For example, a cyber and terror parametric product for small businesses is entirely possible,” he added.
The World Bank’s clients are emerging market sovereigns and for them a rapid payout through parametric insurance is especially helpful, said Byskov.
“This is a great solution for the type of risk that they are exposed to. They have to come in quickly after disaster cleanup, get the basic systems working again, and they need the quick cash,” he said.
“These products have started to appear in the retail market.”
A growing market
Manson noted that parametrics can “bridge the gap” between the client and the capacity provider. Franco agreed, adding that parametric insurance had evolved since it appeared in the late 1990s.
“Back then, parametric solutions were niche,” Franco said. “They were only deployed as cat bonds and they were only used by very sophisticated financial institutions. In the last five years, we’ve seen something that we did not predict would happen this fast, which is these products have started to appear in the retail market.”
Guy Carpenter was already active in parametrics five or six years ago, he said, when it had a couple of client meetings per year to develop these types of solutions. That number has grown to more than 20.
“We not only see public sector entities, who are big governments, but also large corporations, distributors, banks and investment managers. Many more entities are coming to learn about what is feasible within the world of parametric risk transfer,” he said.
Their interest is growing because of losses that are “appearing out of nowhere”, such as business interruption.
“That’s where parametric is stepping in as a ‘catch-all’,” he said, adding that Guy Carpenter is responding to this increase in interest by going beyond its traditional role of an intermediary by advancing parametric technology.
Ismail credited brokers with helping clients articulate their needs.
“Risk capital still needs to be educated because, when they hear of any innovation or new technology, they immediately think it is mediation. We have to very carefully explain that for growing the parametric market, brokers might be even more useful than in the indemnity market because who else is going to help the client understand how to change their buying habits or, if they’re buying for the first time, that there’s a new product which they could buy with minimal intervention and their loss can be recovered.”
Doona highlighted the twin roles of insurers, “as an underwriter and a risk taker”, and the need to understand what matters to the people who are buying coverage. There are two extremes, he said, between a focus on the immediate aftermath of an event and the issues that go beyond the physical damage it caused, and portfolios of small risks, each of which would be uninsurable on its own because of a lack of specific information or it can’t be administered efficiently.
“Those two different extremes make you realise that, it’s not so much you can’t do it, but that you can’t get your hands around it administratively or there’s an indirect financial problem,” he said.
Byskov highlighted how, uniquely, the World Bank issues its clients with cat bonds off its own balance sheet.
“Our unique role here is the fact our clients engage directly with us,” he said.
“The World Bank is owned by its members, who are 189 governments from around the world. It means that, when we engage with our clients, they have comfort from knowing that they own us and that they’re represented on our board. That helps to create the trust of these emerging markets sovereigns in thinking about transferring risk to the capital markets.”
Transparency and speed are two features of parametric insurance that are helpful in a political environment, he said, when the alternative is having to ask parliament to authorise spending.
“We are all in favour of a solid, democratic process but it’s not always that quick, and so this feature of rapid payouts is helpful. We’ve done about three and a half billion dollars-worth of transactions in this space with a handful of clients, and we think there’s a lot of scope for more to be done.”
“Our unique role here is the fact our clients engage directly with us.”
The need for scale
Manson asked Franco to describe the parametric market and he replied that, frustratingly, “I don’t have a number for the size of it”.
He said: “What I do know is that we are just scratching the surface and we have a long way to go. You could argue that potentially all of the protection gap is addressable with a parametric solution over time. The issue then is how we scale up the business, rather than how big can it get.”
Educating risk managers on what parametric solutions exist is needed, he added.
“Risk managers who have always transacted with traditional indemnity products may think this is a risky move for them, so for them to be confident that they’re making a good decision, maybe they need to start with buying small limits and to identify exactly what niche they’re going to target.”
The conversation “goes both ways”, however, and knowledge of parametrics among entities in the capital markets is not homogeneous.
“Parametric solutions have some very key advantages for clients: they pay fast, they pay transparently and they are a catch-all. There are advantages for the capital markets too: they provide a certain outcome, so there’s no trapped collateral issues, and they protect them from social inflation because the limit is what the limit is, and you’re paying based on a hazard measure. And they’re easier to analyse, because you don’t care about the exposure, since the exposure is the event. And we have many models and many sources of information to estimate, what are the probabilities of experiencing a certain earthquake or certain tropical cyclones.”
The issue with parametric products, he added, are “synthetically making them less affordable than they could be”, and technical basis risk. The latter will “naturally disappear”, however, as parametric technology evolves.
“Earthquake instrumentation used to be expensive and now you can buy yourself a three dimensional seismometer for 200 bucks based on a Raspberry Pi instrument, and install it in your house and then connect to the internet and share your data. So the amount of information that we have to make parametrics better and better is only going to grow. So this issue of technical basis risks that is real, that exists, is actually something that will naturally disappear.”
Ismail stressed that many challenges with parametrics had already been overcome.
“For example, a UK company called FloodFlash installs a widget on the side of the insured’s building. If the water level comes to a certain level in the widget, you get paid, and that’s already connected to the internet,” he said.
There is the matter of “instrumenting the insured” being considered an intervention.
“Normally, a client would expect to lose a business opportunity when they expect that level of intrusion and yet there is a psychological reassurance that insurers have from having the widget on the side of the building,” he added.
The point is that such products reduce the basis risk.
“Once you’ve instrumented the insured, the basis risk is negligible for whether something has happened.”
As regards a challenge that still exists, parametrics are normally used as triggers from large events, rather than for predicting them. Citing UK civil service analysis, he said that, for every dollar spent before a disaster, it’s worth $30 after a disaster.
“The trouble is that the metrics for pre-disaster triggers are much more difficult, and often a bit more ethereal, but there are efforts around anticipating events. Parts of the world which are subject to maximum volatility, such as from climate change, could do with having exactly that mechanism, and I don’t see any other form of insurance except parametric being able to do something like that.”
Parametrics are also a way to remove the bugbear of social inflation, which threatens to push “a whole tranche of insurance underwater commercially, come the next renewal”, he said.
“The rates simply will not compete with the current interest rate environment. So I think that this is going to be an upcoming challenge, which isn’t directly to do with storms or anything like that. It’s just to do with the political and economic situation, you find yourself in.”
“The more resources that are available right away, the better off everyone’s going to be.”
Too poor to join
Byskov said an unfortunate irony for clients of the World Bank is that they are often countries too poor to participate in the risk market though they would benefit the most from it. They face a trade-off therefore between relying on international aid following a disaster, and purchasing insurance that would offer more immediate assistance.
“We are trying to bring in donors to purchase insurance and we did a transaction last year, where the premium and the transaction cost were donor-funded. It’s a way to bring our clients into the market,” he said.
Another challenge, Franco said, is the quality—or even existence of—a global reporting agency for certain perils.
Assimilating data on earthquakes is “easy” and also free, he said, thanks to the work being done in that space by the US Geological Survey (USGS) which has the support of the US Department of Interior. But other perils, such as flood, lack a global reporting agency.
“That’s a tragedy because the parametrics are relatively straightforward to build for flood,” he said. “There are some private companies who are offering themselves as reporting agents and that is good, but we don’t have an equivalent of the USGS. So when we talk about scaling up, there first need to be agencies analogous to the earthquake group in the USGS that would give us parameters for other perils.”
Doona said there is a “tremendous amount of information” available thanks to big data, satellites, remote sensors and telematics, which in some cases enable action before there’s an event.
“We have a programme through which we are paying people to evacuate in advance of hurricanes in the Gulf of Mexico. It’s a small amount of money for the recipient, but it’s enough to get them out of harm’s way. That is going to mitigate loss creep and mitigate losses. The message to them is, ‘Get your expensive electronics out of your house and get out of town for two days’. We can sell people the notion that it’s not enough to let an event happen and then rely on the kindness of strangers, but rather to be prepared, which puts people in a different mental state. Also, the more you can do in the immediate aftermath of an event, the more resources that are available right away, the better off everyone’s going to be.”
He added that, the use of remote sensors to detect the risk of a flood event early enough to get people out of harm’s way, and then to try to evaluate what can be done to mitigate the effects of such an event in advance, are examples of how useful parametrics could be.
“Parametrics negate the issue of social inflation.”
Basis risk
Basis risk is an “empty notion”, Doona said, for people with no insurance cover.“Providing them with something is more important than providing them with the perfect thing. We have to understand, what it is that’s driving people because it’s not the same from client to client. The issue is to structure a cover that is going to be responsive. I often say, it’s better to buy less cover that’s highly responsive than to buy a larger cover that’s not. If you’re talking about rapid response funding, or getting people out of harm’s way, then it has to be a very responsive type of cover.”Manson noted that the protection gap is about $160 billion, which will increase with climate change, and thus will increasingly show the divide between the Global South and the Global North.Doona replied: “Is the cover you need better as $50 million in the hands of government, after a certain event occurs, or as $20,000 in the hands of thousands of people? Those are two very different things and you have to figure out what is going to work best. For the most part, we like the idea of getting a little bit of money into a lot of hands more than the idea of getting a lot of money into the hands of some organised entity that’s going to ‘solve the issue’. There is speed from big data, since everyone’s got cell phones, and the more you can automate and make payments administratively very simple, I think that’s a key issue.”Manson said there was hope from the fact that technological developments “inconceivable” five years ago meant five years from now they would enable even better parametrics.“There are clear benefits to parametrics: speed, transparency, and it’s easy to understand from a buyer’s perspective. Traditional indemnity products are more expensive and, as a capacity provider, there’s a lot more uncertainty in that product than people realise. Indemnity is a ‘bread and butter’ thing and so the uncertainty of it gets ignored to a degree, relative to the benefits of a parametric product,” he said. Parametrics negate the issue of social inflation, he added, meaning that “recovery is the recovery and not three times the recovery”.
Franco said there was a need for better understanding of “first world basis risk”.“We have conversations with clients and we measure basis risk with a catastrophe model, but people tend to fixate on the differences. They say, ‘There’s a 10 percent difference in this payout versus the model’ because they don’t realise that the model estimate has 50 percent uncertainty.“And so, we have to put basis risk in context which is a labour of love and something that’s ongoing, but entities that trade in parametrics need to understand that the basis risks that we calculate, with the help of models, are an estimate and one that is not homogeneous throughout the insurance tower.”Indemnity solutions also have basis risk, he added, because there is uncertainty in the contract clauses and that can lead to a delay in payouts.“This business of basis risk has to be nuanced and part of the job we as an industry have, is to start establishing best practices. The parametric space has been jambalaya of approaches. We are all trying to do a good job but there is work to be done on clarifying the rules of the game.”
