Two lessons from Solvency II

EIOPA’s Solvency II work is at a delicate turning point: it must continue drafting standards and guidelines while at the same time shifting attention to the challenges of implementing the Directive on 1 January 2016. All this while uncertainty remains in a number of key areas. In an extensive interview, Carlos Montalvo, EIOPA’s Executive Director talks to Gideon Benari, Editor of Solvency II Wire, about the past, present and future of Solvency II.

In Part 1 of this three-part article Mr Montalvo shares his concerns about the Omnibus II agreement, the impact of the delays and the damaging legacy of mixing technical and political matters.

Dos café con leche. We are in Madrid. Two days before New Year’s Eve. The waiter places the coffee and a little silver basket with a selection of biscuits on the round marble table. A curved wafer-thin biscuit drops onto the dark marble surface of the table. It looks like a curled up leaf. It will remain there, untouched, until the end of the interview.

An agreement on Omnibus II less than two months prior to our meeting paved the way for Solvency II to go ahead, bringing the legislative process to the home stretch. I want to unpack some of the recent history, fraught with acrimony between industry, regulators and politicians, as I believe it may help us understand what happened, but also enlighten us as firms enter the next stage of the Solvency II process: preparing for implementation and, just as important, finalisation of the technical details. Both are as challenging as the political phase.

Carlos Montalvo

© EIOPA Frankfurt am Main

Carlos Montalvo Rebuelta, Executive Director of EIOPA, is in charge of the day-to-day running of the European Supervisory Authority responsible for insurance and pensions. Before becoming its first Executive Director in 2011 he was the Secretary General of its predecessor, the Committee of Insurance and Occupational Pensions Supervisors (CEIOPS), since 2007. Prior to that he worked here in Madrid as an Inspector de Seguros, Inspector Jefe de Unidad, at the Spanish insurance regulatory authority the Dirección General de Seguros y Fondos de Pensiones. He began his career in 1995 as a lawyer.

We have both taken time out of our respective holidays to meet. Mr Montalvo is a Madrileño. This is his home ground. I am a tourist. The temperature outside is close to freezing, yet the glare of the Madrid winter sun shines brightly on our table.

Reflecting on the Solvency II process, no doubt it is the treatment of long-term business that has been the bane of the Directive. Had the problem not reared its ugly head when it did, Solvency II would have probably been in force at the beginning of 2013, instead the current due date is 1 January 2016.

The original Solvency II Directive took just over two years to get from draft to entry in the Official Journal on 17 December 2009 – quite an achievement given that it was a major overhaul of European insurance regulation about a quarter of a century overdue.

However, since then a combination of events led to a series of delays, the so-called Long-Term Guarantees (LTG) being the main culprit. It was only resolved on 13 November 2013. Sharon Bowles’, tweet late on the night of the agreement, “Omnibus 2  deal after 8 hours”, sounded like a birth announcement. One that came at considerable cost to all concerned. For industry, the cost of reorganization and financial planning; Europe, a tarnished reputation and threat of international insurance regulation; and consumers, who will ultimately pay the price in higher policy cost or lower cover.

I wonder, what lessons can be taken forward from all this?

“I would say at least two lessons,” Mr Montalvo says. “The first one is let’s avoid mixing technical with political issues, and that has been a constant of the whole LTG debate. Extremely technical issues have been discussed only under a political type of hat. The second one is that sufficient time for [drafting] regulations is needed. Too little time means lack of testing and appropriate discussion, and too much time in order to develop regulation probably creates lets say, not ideal effects in some areas.”

Lesson 1: Of the political and technical

Madrid 8 – Solvency II Wire
Madrid 3 – Solvency II Wire
Madrid 4 – Solvency II Wire
Madrid 2 – Solvency II Wire

The need to address market volatility on the Solvency II balance sheet first became apparent around 2010 during the Eurozone crisis. Solvency II introduces market consistent valuation of insurers’ assets and liabilities. As fixed income, and in particular government debt, makes up the vast portion of insurance portfolio for long-term business, the debt crisis would have had a destructive impact on the balance sheet of many insurers.

The problem is that assets used to back long-term liabilities are often held to maturity, so with the exception of default risk, that volatility that Solvency II will bring onto the balance sheet would in effect have little material impact on the insurer’s ability to fulfil its long-term obligations to policyholders.

As this became apparent, especially after the results of the QIS 5 exercise were published in March 2011, a sort of regulatory land-grab erupted in slow motion. A range of industry associations and lobby groups lined up to present alternative solutions to the problem – the merits of which varies from pragmatism to almost opportunism, and Member States often backed their respective insurance industry’s position. The decisive step in the process was elevating the main LTG measures from the Level 2 text to the Level 1 text, thus taking a complex technical issue and placing it in the hands of politicians.

Things came to a head in the Autumn of 2012 when the Commission, at the behest of the Parliament and Council, asked EIOPA to conduct an impact assessment on the treatment of long-term guarantee business under Solvency II.
The following June EIOPA published a set of recommendations keeping some of the measures and introducing some alternatives. The report paved the way for the November agreement but a number of EIOPA’s proposals were watered down – some significantly – through industry lobbying of politicians.

Do you think the agreement that was reached in the trilogue provides a sound prudential framework for long-term business?

“With regards to the agreement itself nobody is fully happy. Definitely we supervisors are not. Industry is not as well. So that means probably it’s not that bad agreement at all.”

I detect a tone of resignation in his voice. “Why aren’t supervisors happy?” I ask.

“We know there was quite a lot of deviation from the EIOPA proposal in trilogues.”
“Precisely because of that,” he replies. “We thought we were addressing the problems that, rightly so, had been put on the table in order to make a better Solvency II. In our proposal we saw that some of the responses that we gave we thought they were prudentially sound and from a point of view of business they were neutral. In the end they were replaced in some cases by basically, what would be the English word for that … more tailored type of responses that probably were not thinking of the long-term future but more on the past.”

He is articulate and precise. And maintains the same steady but strident pace almost throughout, delivered with the accent and cadence of his native Spanish. His search for the right word appears to be less about vocabulary and more about diplomacy.

“We want Solvency II and we need to look forward not to look backwards. Our proposal as supervisors was certainly looking forward rather than looking backwards as the next crisis will be different than the last. That doesn’t mean of course that what has come over, what we have ended up with is only looking backwards, it is something that will be applicable. And it’s something that if rightly applied can be a positive outcome of the whole Solvency II project.”

A key concern about the LTG measures as they were originally proposed was articulated by the ESRB. It was argued that in a market consistent regime they did not provide any counter-cyclical buffers. In other words, while they afforded certain protections when market prices were considered abnormally low, they did not accumulate reserves when prices were artificially inflated.

For some these concerns remain even after the final agreement. The day after the Omnibus II deal was agreed Sven Giegold MEP wrote: “While legislators agreed for banks that values might be depressed as well as inflated, insurance regulation asymmetrically foresees that markets can only depress prices. Against ESRB advice insurances (sic) do not have to build buffers in good times in order to be more stable in bad times.”

Does Mr Montalvo think the agreed measures are pro-cyclical?

“I think we are all human and think like that. I mean in the sense that I want the best part of it, but I don’t want the worst part of it. The concern on pro-cyclicality is something that also needs to be understood in the context of the fact that we are coming both with temporary and long-term type of responses.”

He uses the Matching Adjustment, one of the key measures in the package, to illustrate his point. “If you look at the Matching Adjustment, what you want is a mechanism to acknowledge sound ALM [Asset Liability Management] not only looking backwards but also looking forwards. In other words a way to run your life business in terms of long-term assets linked to long-term liability with a number of conditions.”

The Matching Adjustment allows firms to recognize that where they have a portfolio of bonds or bond-like assets with durations and cash flows that closely match a specific portfolio of liabilities they are primarily exposed to the risk of default on these assets and not the volatile movements in market prices. It was introduced following pressure from the industry in the UK and Spain that hold large annuity books that would benefit from such an “adjustment”.

Initially there were also demands from other Member States for an ‘Extended’ Matching Adjustment to include a wider set of assets and liabilities with much looser criteria. But this was rejected by EIOPA and excluded from the final agreement.

The Omnibus II agreement sets strict criteria in terms of asset types and surrender options for the assets and liabilities that can be used for the Matching Adjustment. In the final stages of the negotiations the UK industry lobbied hard to change a key definition in the text from “fixed” cash flows to “predictable” cash flows. This was rejected by the trilogue parties.

Mr Montalvo is also concerned that if the measure is misused it will lead to the wrong perceptions that it is not good, or poorly designed. “We, of course, are more comfortable with the ones [measures] that we put on the table. We thought that they were limiting the possibility of misuse of a good measure. I think that is an important message,” he reiterates. “A misuse of a good measure can lead to the wrong perception that the measure is not good. And this is something we certainly would like to avoid.”
The second key measure of the LTG package is the Volatility Adjustment: a permanent modification to the risk-free curve used to calculate the present value of liabilities and which can be applied to a wider range of liabilities. Here too the EIOPA proposal was modified – and significantly so. The agreement states the modification should be 65% of the risk-free adjusted yield on a “notional portfolio” representing an insurer’s typical portfolio of assets. EIOPA proposed only a 20% modification.

To my mind the decision to award the Nobel Prize for economics in 2013 to economists with diametrically opposed views of market-consistency perhaps reflects the state of the world post crisis. A quarter of a decade of dominant economic theory was proved wrong and yet we somehow know that the truth is in there. If we only knew how to see past the noise, markets do provide the true value of assets. The solution proposed in Solvency II is only partially satisfactory because once you deviate from strict market consistency, where do you stop?

These questions exist in an awkward space, a bit like that biscuit lying, upturned, on the table casting a faint shadow on the dark shiny surface.

I pick up on his earlier point about the “misuse” of a good measure. His concerns are already materialising. It didn’t take long for companies to start considering which of the two measures would yield the biggest benefit: the Matching Adjustment or Volatility Adjustment. I discovered this when I was speaking at an event in Paris and someone pointed this out to me at the speakers’ dinner the night before. That was on 28 November. The Omnibus II agreement was reached on 13th.

All the LTG measures do not have Member State options so they can be used by any firm across the single market, subject to some supervisory approvals. The Matching Adjustment and Volatility Adjustment are mutually exclusive; you can either use one or the other for any portfolio of assets and liabilities. What companies have started to do is to test their portfolio against these measures to see which will be the most capital efficient.

“My concern is not linked to regulatory arbitrage,” he says. “It is linked to the fact that we have been hearing, in particular when linked to such things as the Matching Adjustment, that this makes full sense from a business perspective. Now when you are starting to consider whether this is better than another approach that was aimed to address volatility then what do you think in terms of how embedded it is in your business? And that is basically our concern.”

In the final rounds of the Omnibus II trilogues, provisions were introduced to monitor the use of the measures at a European level. At national level supervisors are given a lot more discretion in approving their use. The concern, articulated by Mr Montalvo, is that firms will use the measures as a way to gain maximum benefit when the intention of the regulator is to shield long-term business from short-term volatility.
The use of any measure should be linked to the nature of the business. The business should not be changed to gain maximum benefit from using a specific measure, he argues.

“By definition look at the name ‘Long-Term Guarantees’. It should be a long-term measure rather than a past driven type of measure. And it should be applicable throughout the Union. Not as a way to see how I get the better benefit at this moment of time, but as a way to see does this fit with my business model? And I think that is the most important part of the whole thing. It should be part of the business model, it should not be a way to impose change on what you had been doing right.”

So whose job will it be to monitor that?

“It is going to be a shared role. It is going to be the main responsibility of the national supervisor, but then EIOPA will ensure that there is a consistent application throughout the Union.”

The dual responsibility will help reconcile the problem that while Solvency II is a European wide regulation, many of the long-term products are unique to specific countries. “Nobody knows the local market better than the local supervisor, but nobody can provide a better European perspective than EIOPA,” he says.
EIOPA has at its disposal two sets of tools to address this problem: supervision and regulation.

“When it comes to supervision, Colleges [of Supervisors] will play an essential role. But also EIOPA is gradually enhancing its role, for example when it comes to joint on-site visits or when it comes to getting an understanding of how supervision is undertaken.” EIOPA has also been working on a supervisory handbook for over a year, which will help ensure consistent responses to similar problems.

From a regulatory perspective EIOPA will conduct peer reviews between national supervisory authorities and will also monitor that the rules and guidelines are implemented consistently without lowering the bar or ‘gold plating’, which Mr Montalvo politely notes is “not necessarily ideal” for a single market.

Lesson 2: Time … too much time

Hotel Villa Magan 2 – Solvency II Wire
Madrid 12 – Solvency II Wire

Mr Montalvo’s second takeaway is that too much or too little time spent drafting legislation could have undesired effects. The Solvency II process was protracted due not only to the LTG and the crisis but also to a number of other factors such as European Elections and the Lisbon Treaty.

In 2009 changes had to be introduced to the European legislative process to reflect the Lisbon Treaty. Omnibus II, the amending Directive to Solvency II, was originally only supposed to adapt these changes (essentially the formalisation of EIOPA), but the timing meant it could be used to change the Level 1 text to reflect some of the LTG concerns that were becoming apparent. This resulted in reopening some agreed issues thus creating one of Solvency II’s more damaging legacies.

“When you open [agreed issues] you create automatically a feeling of mistrust, which is also I would say an important lesson out of this,” Mr Montalvo laments. “We can perhaps say that we are not in the level of trust of the different parties that is ideal in order to best work together. And one of the reasons has been precisely this reopening of things that some of us already thought were discussed and closed.”
The issues that were reopened range from broad measures to specific calibrations and, as is often the case in acrimonious situations, it is difficult to identify who did what, when and where and whose fault it is and why. Who is to blame depends largely on who you ask. It makes for good TV drama but, what is important, is the outcome.

Lobbying activity has been ubiquitous in Solvency II. Industry has found ways to make its views heard beyond the public consultations. Where one path was blocked, it pursued another. It is widely recognised that the LTG measures were moved into the Level 1 text due to intensive industry lobbying, and later during the trilogues much of the efforts moved to the Council.

The extend of industry activity can be gauged from comments made by Karel Van Hulle, the then Head of Insurance and Pension Unit at the Commission, at the EIOPA Stakeholders Group meeting on 18 October 2012. According to the minutes, he noted (emphasis added): “[The] EP [European Parliament] and COM [Commission] are both convinced that Solvency II is necessary and urge to go ahead with LTG, but some uncertainty remains in the Council as Ministers get heavily lobbied.” The following month even members of the group raised concerns about lobbying activity in the meetings.

As Hugh Savill, Director of Regulation at ABI said in his opening address to the ABI Solvency II conference in March last year, “Insurers pride themselves on being long-term investors. We are also long-distance lobbyists; and it is just as well. Because Solvency II was first proposed in 2005 and it looks like we are now getting towards the end eight years later.”

I want to understand to what extend industry has influenced the rules and seek a way of phrasing my question so as to get a meaningful answer: in modern regulation there is a dynamic balance to be found between input from industry and regulators. Do you feel that balance as shifted too much towards industry?
Mr Montalvo is quick to respond. “Two things here. My immediate response would be, yes. But if you ask industry they will say it has been shifted too much to the supervisors. And we would probably use the same argumentation: the crisis.”

“They will say, well there is an overreaction from the supervisory community and then the political level wants to punish the financial sector; that would be the industry approach. Our approach is, because of the crisis, because of the need for example to keep financing economy and so on, the political level is willing to listen more to industry than to supervisors at least in the field of insurance.”

“Am I worried that industry is doing that? No. It’s their job. Am I worried that sometimes the political level has listened to industry more than supervisors?” There is a long pause, “Yes”.

“But again, ask industry the same question and you will get exactly the same answer with let’s say a shift in terms of the balance that you were referring to.”

Despite this, Mr Montalvo is an advocate for maintaining industry involvement in the process. “But of what I am really convinced is that we do need industry and we need supervisors in order to create regulation.”

He leans back in his chair. I suspect none of the other guests at the café of the Hotel Villa Magna where we are sitting are engaged in such intense conversation. If they are, it is more likely to be about how to spend this splendid sunny, Sunday morning rather than a forensic examination of European insurance regulation.

As mentioned before, one of the consequences of the delays and mistrust has been the shifting of many of the LTG measures from Level 2 to the Level 1 text (of the Omnibus II Directive). This effectively turned a technical matter into a political one. Introducing such specific measures appears to contradict the purpose of Level 1, which is supposed to be principles based, with details provided in the Level 2 text.

The only direct reference to market-consistency in the original 2009 Solvency II Directive text is in paragraph 3 of Article 76 on General Provisions: “The calculation of technical provisions shall make use of and be consistent with information provided by the financial markets and generally available data on underwriting risks (market consistency).”

The political implication of hoisting the LTG to the Level 1 text was to sideline EIOPA, as it is by definition a technical body. The outcome is alienation.

“The point is you’re talking about, let’s put them [supervisors] aside because we don’t trust them so we need someone else to come with regulation. Then supervisors feel alienated out of the whole process, with regards to the rules that they have to apply in order for them to do their job. So if you would be a supervisor what would you do?”

“I mean you have to do your job. You feel in the same way, and we were talking about this before, that companies say, ‘I have to run my business, I have to do business and make money, I need to be part of the process in one way or another, and I am you could say successfully so through legitimate lobbying’. Supervisors also need to be part of the process. When you try to put them aside what you get is alienation between the rules to be applied and their use for supervision, and that leads to no good. That is not good.”

“Because as legitimate as it is from industry to get everything they think they need from the political level, it is for supervisors to feel reassured with the way they are doing their job.”

Like a glacier scarring the land as it slowly moves across it, Solvency II has left its marks on Europe’s insurance regulatory landscape as it edges to its conclusion. Yet Mr Montalvo firmly believes that industry and supervisors should be working together in a constructive way to draft the rules. “But when you enter into this situation you find ways as industry to put the regulator aside in terms of the rules that are going to be applied to you, and you find ways as regulator in order to try to perform your task and your work the best you can. And none of the two situations is good.”

In Part 2 Mr Montalvo discusses the EIOPA Guidelines and the challenges of meeting the 2016 implementation deadline.


Part 2: Getting ready for Solvency II

Part 3: Transparency and legacy