The ITFA Insurance Roundtable, held on 7th September 2017, was discussed by ITFA’s insurance committee, moderated by Katharine Morton. At the time of the interview, Katharine Morton was still Editor in Chief for TFR, the Trade and Forfaiting Review. Unfortunately, Wilmington decided to stop the edition of this magazine shortly after the ITFA conference. Katharine was nevertheless so kind to transcript the interview so that the readers of the ITFA Newsletter now have the benefit of getting the insights on this important topic.
- Geoffrey Wynne, Partner, Sullivan & Worcester
- Huw Owen, Global Financial Risks, Head of London Markets, Liberty Specialty Markets
- Sébastien Heurteux, Head Trade & Insurance Syndications, BNP Paribas
- Volker Handrich, Head Trade and Supply Chain Finance, Swiss Re Corporate Solutions
- Robert Nijhout, Executive Director, International Credit and Surety Association (ICISA)
- Silja Calac, Senior Surety Underwriter, Swiss Re Corporate Solutions and ITFA Board Member, Head of Treasury and Insurance
- Moderator, Katharine Morton
Walking the fine line
The knotty subject of differentiating and choosing between risk participation agreements, surety and insurance as a way of mitigating credit risk was discussed by ITFA’s insurance committee, moderated by Katharine Morton.
Katharine Morton: Where are we now in terms of insurance policies versus risk participation agreements (RPAs) and what does everybody understand by these?
Geoffrey Wynne: We are looking at a series of potential credit risk mitigants, and, essentially, they have to fall into the section: guarantees. That’s the only real clue that sits in the CRR [Capital Requirements Regulations], and the discussion, will be around which is better – which would you prefer to have – guarantee, first demand guarantee, and, within that, surety, risk participation and insurance.
Katharine Morton: Around the world there are different definitions of what all these things mean. From a US audience, for example, what is possible to achieve with them, and what is not possible?
Geoffrey Wynne: We talk in terms of CRR, which is European, so our discussion is EU-wide. One of the strange things is that that is everybody’s law. That is not an interpretation - that is the law. So, credit risk mitigation is determined in accordance with article 194 up to article 225. And, of course, whether you are allowed to do any of this is still a little bit of an issue. So, I can look at, in the UK, whether you can be an insurer, and that’s governed by an English Insurance Act. In Germany, it is different. And, one of the debates is: are insurers moving outside of what their territory should be i.e. insurance, when they’re issuing surety bonds? Can an insurer issue a guarantee? Is every guarantee issued by an insurer ‘insurance’? And that makes a difference, because insurance law dictates both the regulation of the entity and the interpretation of it. And, then you get into an even more interesting area, which is: can an insurer issue a risk participation? Is that outside the permitted requirements? We would say no, and certainly some around this table, would say no. So, some very interesting and intricate debates, because, as you’ll see, when you say “I guarantee that if he doesn’t pay, I’ll pay,” and if you say, “and I insure you that if he doesn’t pay, I’ll pay,” it may be depending on who says it as to whether that’s insurance or not insurance.
Katharine Morton: That sets up a nice framework for where we’re at, but we’ve got the whole market around the table. Silja, why did ITFA set up an insurance committee?
Silja Calac: When I joined Swiss Re three years ago, there was no insurance community within ITFA. Swiss Re was the only member of ITFA that was from the insurance world. I knew from my previous job that banks often have quite a few problems to find their way around these questions. What insurance should I use? Which insurance gives me which benefits? Because, there is not one product which is all good and another one all bad; each product serves specific needs, and has specific advantages or inconveniences, so that was what we wanted to clarify. That’s why ITFA set up the insurance committee. Since then, many insurance companies and brokers have joined ITFA. We have already achieved quite a lot in giving clarity with our insurance guidelines and with our training efforts. One interesting question is: can an insurance company issue surety or guarantees or risk participation? For surety this seems quite clear because there is regulation associated with that, but Robert will elaborate a little more on this because this is his specialty.
Robert Nijhout: There is a bit of a thin line between what insurance is and what bank guarantees are. The determining factor is how it is regulated. Insurers are regulated by the insurance regulator, which arguably in many countries is the same as the bank regulators, in the same buildings, often in central banks. But, perhaps more importantly, insurance is reinsured and that is usually the reason why some conditions need to be applied in an insurance product because reinsurers don't like certain risks such as nuclear risk or five great power war or financial guarantees and such.
Surety, unlike credit insurance, is legislated as well as, regulated by the regulator. You can start a credit insurance company without having to comply with a certain law. But surety is legislated and regulated on a national level. If there is no surety legislation, you cannot sell a surety bond. It depends on the country on how they define it, but in Europe it is almost always defined as an insurance product, therefore regulated by the insurance regulator and subject to Solvency II in Europe and similar regimes in jurisdictions around the world. But, of course, insurers will do as much as they can to satisfy the clients' demands. If a client wants an on-demand product which is unconditional, then insurers will try to issue that. Another big difference between banks and insurers is that insurers don't have collateral, they just have underwriting. So, if they issue an unconditional on-demand guarantee to just anyone, you don't have any security and you're also in breach of your Solvency II regulations, etc. So, surety companies are very selective to whom they issue these types of products, but they are available. That’s perhaps the attraction for a lot of clients despite the conditions that some surety products have, vis-à-vis bank guarantees; the fact that they don't have to collateralise it and don't have to block credit lines or other ways of guaranteeing the collateral, gives them a lot more freedom. It’s horses for courses, as Silja said.
Katharine Morton: Where do banks sit on this? Are you using insurance products as a way of just getting away from Basel II considerations?
Sébastien Heurteux: Risk Participation Agreements [RPAs] are a dream product, it is on-demand and it is simple to explain internally and to credit committees. The wording is simple. It limits scope for wording negotiations. I would say RPAs are more automated in their way of working. Now to simplify the product to such extent, such cover will only works between two parties which have a high level of trust. The RPA is a dream product but it is not accessible or offered to everyone. At the industry level, Should there be a default by the Insurer in a situation of claim, this would taint the reputation of the product. In the wording of an insurance policy, terms and conditions are detailed in a greater manner. .
As far as BNP Paribas is concerned, we use both products – RPA and insurance policies – in a very “fungible” manner. The RPA is a more efficient and user-friendly insurance policy, but at the end of the day, we contract an RPA in a similar way we contract an insurance policy.
A point raised was the RPA’s jurisdiction-specific feature. What many banks want is a single tool used worldwide to cover its risks and/ or get capital relief. Insurance Policies have more of a worldwide reach in their availability across jurisdictions. The more the covers are differentiated in their nature or form between business lines, regions or territories, the more banks will be exposed to operational risk. The key risk for an Insured to lose the benefits of an insurance policy is often assimilated to the Insurer going burst ie a Credit Risk. We should not underestimate the risk of the Insured not fulfilling correctly his obligations viz. the Insurer: an Operational Risk.
I would say there’s a sense of uniformity that we need to have in a Global Organisations to make the risk cover work optimally. Today RPA remains very specific to certain Insurers, to certain underlying asset class, to certain geographies. Until RPAs are offered in a more widespread manner, it will only be used in an opportunistic manner.
Katharine Morton: So, is that an issue of documentation, is this an issue of just different types of risks that are covered, or different jurisdictions in terms of doing things on a one-to-one basis?
Silja Calac: What Sebastien said is important. Risk participation does not give the same level of comfort to the insurance company as a traditional insurance policy does. Therefore, with a traditional insurance policy, an insurer would cover up to 90%, 95%, sometimes even 98% of a risk while under a RPA, one would usually not cover such a large share.
Volker Handrich: I agree. Our main product is risk participation. What does the client want? In the EU, for CRR compliance, capital relief is a big topic. RPA is a great product because it’s unconditional hence it gives the capital relief and it’s a very simple approach. If there’s a non-payment, we come in. There are some challenges, one is on the underwriting side, and this is where there is a place for both insurance as well as RPA.
Because we give this broad cover, this guarantee type of cover, we would not do 90-95% of a transaction. It’s a question of ‘‘skin in the game’’. Thorough underwriting is much more important. On the retention side, we are much more selective and usually would like to go for 50/50 risk sharing - we want to see a lot of ‘’skin in the game.’’
The other challenge is that whereas usually the RPA is based on English law, as Geoff pointed out, from a regulatory perspective, and from a tax perspective, you still have to respect local rules, and there the treatment can go in your favour or against it. The UK Insurance Act is another question. The regulatory question is ‘’can we use the same wording in every country?’’ And the answer is ‘no’. In the EU, yes we can, typically, but in the US for example, we cannot. In Singapore, we have to adjust it. In other countries, like China, you have to get every word change approved by the regulators.
So it can become a challenge from a tax perspective too. Sometimes it’s an advantage in so far that IPT [Insurance Premium Tax] is not applicable under the RPA in certain countries, but again, you have to check in every country.
Katharine Morton: What is Liberty’s view on these products?
Huw Owen: I’d just like to pick up on a couple of points. From our perspective, we’d make a distinction between the two RPA products we’ve been talking about. The RPA that’s issued by the bank is one type of product that’s probably underpinned by contract law. Some of the RPAs that are issued by insurance companies, we feel that they are insurance policies.
This is probably the underlying topic of conversation: what is an insurance policy and when is an insurance policy an insurance policy? I appreciate that this is a tricky area due to there being no specific definition of a contract of insurance under UK/EU law but arguably from an English law perspective, if the product is issued by an entity that’s an insurance company, and it has the characteristics that common law has shown to define an insurance policy [consideration, insurable interest, fortuity then it probably is an insurance policy.
Katharine Morton: If it walks like a duck, and quacks like a duck, maybe it’s a duck. Why is it important?
Huw Owen: Potentially this is semantics because it’s just arguing about what you call it. But it also goes to what the underlying law is and any disclosure obligations there might be. So, in the absence of anything specific in the policy that strips away some form of duty of disclosure then arguably there is a duty to disclosure by proxy because it’s an insurance policy to the provisions of the Insurance Act.
I agree that there is a place for the different products and ultimately what we’re all focused on is giving a client the best product – the product that they want. So, the question is – ‘’does it work for the client?’’ And, if we think that both products can qualify as a guarantee under CRM [credit risk mitigation], then maybe that’s doing what it needs to for the client.
One potential area of improvement of insurance is its usability. There are two aspects to the usability in terms of understanding some of the conditionality in the document. We’ve come an awful long way. Geoff might disagree, but I spend a lot of time renegotiating with our clients – in a way negotiating with Geoff and his team – and I think those wordings are very good, and they’re very good from a client’s perspective, and we’ve really narrowed down the operational risk that banks are running. I know Geoff might probably would go a little further, but they are good templates.
The other issue where insurance probably needs to focus and really where the insurance RPA’s strength lies, is on the claim payment. The potential concern for banks is when they come to claim, they could be facing an entirely different team within the insurance company who are making a determination on the claim often using external loss adjusters. In this context I can understand why any insurance RPA’s self-certification claims process and payment in 10 days, ask-questions-later type of approach, can look more attractive as a product.
Some of the big insurance providers engage their claims function throughout the underwriting and wording negotiation process, so that there’s no risk of reunderwriting when you get to a claim situation. When we are assessing a claim, we can respond very quickly and are focused on making timely payments on valid claims, so that we meet all of our obligations. Because, the worst thing we can do is make a mess of paying a valid claim, that is not good for anyone.
There are clearly some advantages to the private credit insurance market’s willingness to consider up to 90% indemnity and its risk appetite both in terms of structures (very little that the market won’t consider) and jurisdictions. But the willingness to offer high indemnities came about from a partnership with clients where insurers were relying on material information provided to by clients.
Geoffrey Wynne: I agree. Insurance has now come a long way and is a lot better. Under the Insurance Act the concept of the insured being able to discharge its liability on disclosure by making a fair presentation of facts, does make a lot of difference. But, they’re different creatures. Insurance is insuring against designated risks, it so happens that when you’re dealing with credit insurance, non-payment insurance, that that’s the same risk that the other products cover with the guarantee and the RPA. What’s interesting is that there’s a waiting period for payment with insurance. We have accepted that 180 days, the conventional waiting period, is still timely for regulatory purposes, for CRR purposes.
The requirement to act as if uninsured in that period, to keep working at it, has a slightly different relationship, and I was interested that Volker used the phrase “skin in the game”. It is quite interesting because whilst the insurer says, “you’ve got to have a bit of skin in the game,” actually what the insurer says is “you’ve got to work to get this money back, even though you’ve insured it. I’m looking to you, the insured, to offer me something I can understand that I can assess.” So, I’ll assess the risk based on what you’ve told me. And it just so happens that I’m quite likely to have a bigger and better appetite for risk because my view, cynically, if you like, is that if I charge premiums of 10%, so long as I don’t have to pay out more than 10%, I make a profit. A bank looks at every single loan it makes and expects it to be repaid. So, it can’t go to its credit committee and say, “let’s do these 100 deals but I’m certain that 10 of them are going to fail.” The credit committee would say “which of the 10 are going to fail? We’ll get rid of those.” The balance, the skin in the game point in risk participation is essentially that the guarantor and the insured will work together.
Silja Calac: I fear we’re starting to mix things up. When you say that you believe that when insurance do risk participation this might be insurance, I’d say of course it is all insurance, but there are different insurance products. When Swiss Re does risk participation, we actually do it under the surety regulation [Class 15 as per SI No 359 of 1994 -European Communities (Non-Life Insurance) Framework Regulations, 1994]. This is clearly a product which is separate from the comprehensive non-payment insurance product which we are comparing here, and it is not regulated by the UK Insurance Act, for instance. But, it is a product which insurance companies are allowed to do, and which is regulated as such. Both these products have their specificities and can be of advantage to a bank or not. I would like to avoid confusing readers here, not that they end up thinking it’s all the same, and some insurance companies are just twisting regulations.
Volker Handrich: The thing is do we talk about the law dimension? Is it a regulatory dimension? Is it a tax dimension? That’s where the difficulties start. The credit insurance product, on a global level, is the more standardised product, because there is a more established market, whereas RPA works very well in certain countries or regions where you have the capital relief, but in other countries you have to adjust on a local basis and that’s where the challenge lies - standardisation versus a better product.
We’re all within insurance regulation and we all have a value proposition. Ultimately as a bank you have to decide for which client and which situation is the right product. Insurance and RPA are both unfunded. If a bank has very high funding costs they might go for funded syndication instead. They might opt for a very different route if capital relief is the big driver. Then it depends on the bank’s internal capital model. Most follow the advanced approach, some follow the standard approach, and depending on the model they use, maybe RPAs are better but then maybe in a different situation you have a US bank that has sufficient capital which may prefer the lower retention or maybe even a better price from credit insurance.
Sébastien Heurteux: If we put aside law, legislation, regulations etc, and we take a very operational approach. In both cases whether it is Liberty issuing Insurance Policies or Swiss Re issuing RPAs, we are facing Insurers. An Insured should have the exact same approach in terms of disclosure and answering questions when presenting a file. Similarly Insurers seem to go through similar underwriting processes. So, regardless of the type of cover, an RPA or an insurance policy, the placement process and interaction between the Insured and the Insurer look very similar.
Going one step further in the situation of a claim, the willingness of an Insurer to pay overrides the type of cover. In the examples that we have experienced or that we know of, I would say insurance policies and RPAs have always been timely paid indistinctively from their nature. As a track record, in terms of the efficiency of the product, we are all good.
At the end of the day, what we want is the cake. If we have the cherry on the cake, it’s not the cherry that will feed us, it’s the cake. Now, the cherry will always look good….
What I hear is that between an RPA and an Insurance Policy each bank has a sensitivity of its own. It is often ruled by their own internal systems, understanding of the products, methodology and/or applicable regulation.
The bank regulators have not tackled the insurance product per se when laying out the regulation. Now the Credit Insurance is commonly treated as an extension of what the regulator defines as Guarantee when tackling Credit Risk Mitigants. This somehow can blur the picture or set undue constraints to the efficiency of Credit Insurance. That is why I don’t think there is a right or wrong.
Katharine Morton: Where are we at with the Insurance Act, what’s changed and what will change?
Geoffrey Wynne: The Insurance Act presented a very good opportunity. It essentially rewrote insurance law, which had been more than 100 years, very insurer-biased was the perception of the market. Therefore, there were lots of reasons why insurers were able to escape liability. The new Act tried to close loads of those gaps. It did not remove the responsibility from the insured to tell the story properly but said you could do it at the beginning in one presentation. The problem has been two fold. One, the Act allowed opting out of certain clauses which we believed the insurance market was using a little bit too much and one might say abusing, in other words they didn’t have to think about the point, they opted out to go back to where the old law was. And the second thing, where ITFA in fact was trying to step in, was saying, “Given the fact that we have a really good Act, why have we not got some standardised wording in the policies, or indeed, why don’t we have standard policies, a bit like the BAFT risk participation agreement?” In other words, why are insurance companies still issuing their own policies?
Katharine Morton: Maybe Rob would like to make a comment from the ICISA perspective.
Robert Nijhout: First of all, the insurance market is an open marketplace, subject to antitrust regulations. So if insurance companies all collude to draft one policy wording, and the market had to swallow that, that’s unacceptable. But perhaps more important is we have a fiercely competitive market, the market is extremely soft, and the underwriters have to compete very hard for their business. With this soft open market, there is downwards pressure on price. And if you don’t want to compete on rates, on premium rates – and I wish we could charge the 10% that Geoff mentioned earlier – you have to compete on something else, which is the quality of your product.
Quality is defined in the policy wording, in a way that the customer likes to see, and whether that’s a global wording for a multinational or whether that is a very specialised wording for a particular deal with structured finance and a long tenor. So, the market wants that difference. Having said all that, it’s in the interest of the insurers who are dealing with banks with a symbiotic relationship, and we depend on one another to come up with a product that is recognisable for banks. So, you could think of creating guidelines or parameters or something in that realm where you say, “if the policy complies with these characteristics, then it’s more acceptable for banks,” but then we would say, “can the banks also have a single opinion on what they think of insurance companies?” And that doesn’t happen either. So, the first thing we need to do is educate one another, because I don’t think insurers appreciate the limitations or the specificalities that banks have to deal with and vice versa, so it would be good to have a bit more openness on both sides.
Katharine Morton: That’s what IFTA’s up to, isn’t it?
Silja Calac: As Geoff mentioned, we have thought about drafting a non-payment standard policy. At least in the insurance committee, we think you could standardise a non-payment insurance policy to some extent, and that’s also our aim. The guidelines were a start. The time isn’t yet ripe to go too far, but Huw would you agree in the longer term we should aim to come to more standardisation?
Huw Owen: If you could reduce the amount of time we spend on conference calls with the lawyers that would be a bonus! Potentially the biggest losers of a standardised product will be the brokers. This is their leverage on what they sell, and maybe some clients have got themselves in a position where they’ve been using the market for a very long time and they have a very good wording themselves, so maybe they wouldn’t be that keen to use a more standard product.
But maybe, it’s inevitable. It’s a bit controversial but certainly we still spend a lot of time on wording negotiations. Most banks are moving to the point where they have one policy template, but there are 50-odd markets now, so to get the approval of 50 markets takes quite a long time. The point about the opt-outs and carve outs – yes, I think that insurers have had the use of carve outs and we have narrowed it down to a couple of clauses that we feel that if we didn’t have the carve outs the clause would be more or less toothless. I’m talking specifically about Section 11 – causation – but the Insurance Act work was very welcomed, it definitely redressed the balance more in favour of the client. It was long overdue. The market responded relatively well and we got there in the end. We came to a very good landing, again with some lawyers’ help on certain issues, like defining what constitutes a reasonable search and knowledge for our banking clients. Arguably we all could have reacted slightly better, but we got there in the end. And one thing it did demonstrate is that you’ve got 50-odd markets, but they are relatively collaborative, and that’s one of the defining positives around our market.
Geoffrey Wynne: What’s interesting was that there was a long gap from when the Act came into effect and from when the Act was passed, and that time was not used well, and that’s been a criticism. The point is, there was no real meeting of the minds. And, it was interesting that the RPA, the risk participation market, having gone for years with lots of different forms, finally woke up in 2006/2007 to the idea that it would work to have a standard form of wording, and therefore it was collaborative in a way that you would probably have said five years before then that banks would never have agreed. What banks have done is, they all have their little idiosyncrasies and, 10 years down the road, the idea is to re-tread where we’ve got to. And in some ways, Huw you’re right that, in the insurance market, the combination of a lot of insurance companies a, lot, lot more Lloyds underwriters and a lot of brokers has meant that nobody saw it really in his interest not to put in his two pennies worth on the wording. Whereas the banks in RPA terms said, “actually, we don’t want to waste time on those long phone calls,” and now on many risk participation that we’re involved in there might be one or two exchanges about one or two tiny issues in the drafting. The lawyers would actually like the insurance market to get to that point. We’re not saying “standardise totally”, but we are saying, “Why do we have to rewrite this? Why isn’t Clause 1 the same? Why have I got to look and see whether Clause 2 has been put into Clause 12 and is also in Clause 8?” And I think that the RPA has sort of overcame that.
Sébastien Heurteux: In the past, the policy wordings for credit insurance were worked-out on a bilateral basis between the Insured and the Insurer. One party was proposing a template that was amended to reach an acceptable format to both. At that time, there were a limited number of insurance companies and Insureds. Over the past few years, the market has grown very rapidly with now over 60 insurance companies. At the same time lawyers have started to be used for the drafting of insurance policies.
With the new Insurance Act (and CRR requirements), the involvement of external law firms has increased even more rapidly. . With a rather small number of lawyers truly knowledgeable about credit insurance, these very few law firms came out to be instrumental in setting, common ground between Insureds and Insurers on policy wordings.
Because of both the credit insurance market growth and the legal and regulatory environment changes, the role of law firms has suddenly become essential. The limited number of knowledgeable law firms advising Insureds and Insurers has resulted in more uniform templates used throughout the credit insurance market.