4.2.1. Recalculation every 2 years

SS6/16 indicates that the PRA expects firms to carry out a recalculation every 2 years from the date of implementation of Solvency II (1 January 2016) over the TMTP’s 16-year lifetime, irrespective of any recalculations resulting from a material change in risk profile (which are discussed in section 4.2.2 below).

Is resetting of TMTP on recalculation every 2 years compulsory even if the change is not material?

The purpose of the biennial recalculations is said to be to keep alignment between the relief afforded by the TMTP and those elements of technical provisions for which the TMTP was designed. No mention is made to materiality for biennial calculations which is in contrast to recalculations for changes in risk profile for which the SS indicates that only if the solvency cover ratio changes by 5% or more would the PRA expect firms to apply for a recalculation. This could be taken to imply that the TMTP would be reset on two yearly recalculations even for small changes. Firms could though presumably discuss this point with the PRA.

Is supervisory approval required to recalculate every 2 years?

For the biennial recalculation, it is currently unclear as to whether a firm will have to follow the same process as for an application to recalculate following a material change in risk profile or whether a firm can recalculate without requiring PRA approval. The regulator’s requirements for such a recalculation should be discussed with a firm’s supervisor and could be reflected in its policy. The policy could also include the conditions under which a firm would not recalculate its TMTP at a biennial review date (e.g. immateriality triggers).

4.2.2. Material change in risk profile

Outside of the regular biennial recalculation cycle, the TMTP can be recalculated if a firm can demonstrate that a material change in its risk profile has occurred. A firm’s definition of a material change in risk profile needs to be defined in its recalculation which is discussed in section 4.3 below.

What sort of circumstance may result in a material change in risk profile?

SS6/16 states “In the PRA’s view a variety of circumstances may give rise to a material change in risk profile. Risk profile changes that may trigger a recalculation include but are not limited to the following examples:

• ∙ acquisition or disposal of business priced and written before 1 January 2016;

• ∙ material changes to the reinsurance programme for business priced and written before 1 January 2016;

• ∙ unexpected changes to the run-off pattern of the insurance obligations in scope of the transitional measure;

• ∙ a change in the firm’s use of either the matching adjustment or the volatility adjustment; or

• ∙ changes in operating conditions, including in interest rates or market prices of other financial assets leading to revised market risk exposures, or crystallisation of an insurance risk exposure, e.g. a change in projected mortality experience.”

The list above can be classified under two main headings, changes related to: a change in operating conditions (third and fifth bullets) that are generally outside a firm’s own control; and more structural changes that are likely to be initiated internally (first, second and fourth bullets).

What level of evidence would be required to support a change in risk profile needing recalculation?

A compelling argument that the recalculation would have a material impact on the entity’s balance sheet is required. The SS indicates that the PRA’s assessment of a material change in risk profile will take into account the:

1. (i) change in risk-free rate (expect 50 bps or more in the 10-year rate);

2. (ii) impact on solvency coverage ratio, and

3. (iii) impact on recalculation of solvency coverage ratio (expect 5 percentage points or more) at a legal entity level.

The changes in the risk-free rate are expected to be sustained. A sustained change is defined as one that has persisted over a significant period of time or driven by factors that are likely to persist over a significant period of time. The PRA also note that changes in credit spreads could be relevant in this assessment but this depends on a firm’s asset holdings.

The PRA expects firms to provide evidence when applying a recalculation taking into account the PRA’s own three assessment criteria described above. For example, to explain materiality of the change in solvency ratio by comparison to expected frequency and likelihood of occurring. The PRA states that it is their view that the expected frequency and likelihood of changes occurring in the solvency ratios and own funds are key indicators in determining what constitutes a material change in risk profile.

It may be appropriate for a firm to have different thresholds to those outlined in the SS, as discussed, and if this is the case these should be discussed with the firm’s regulatory supervisor. Evidence of a breach of the thresholds set out in the firm’s recalculation policy will need to be agreed internally and then provided to the PRA as part of the recalculation application.

At what level in a firm’s business hierarchy should the TMTP recalculation impact on solvency coverage ratio be assessed?

The PRA SS on the recalculation of the TMTP outlines that the PRA would not generally expect a firm to apply for a TMTP recalculation where the increase or decrease in solvency coverage ratio at a legal entity level is less than 5 percentage points.

The working party has identified a situation where assessing the impact of a pro-forma TMTP recalculation on solvency coverage ratio at a legal entity level may not be fully appropriate in certain instances. Specifically for firms with ring-fenced funds (RFF), a legal entity solvency coverage ratio does not necessarily demonstrate a change in risk profile as a recalculation of the TMTP does not impact the surplus assets available due to RFF restrictions. Furthermore, RFF business dilutes the impact on the coverage ratio of a recalculation of the TMTP. The simplistic example below demonstrates this:

Before a recalculation of the TMTP, the firm’s position is as follows:

1. (i) TMTP in respect of non-RFF business=£1,000 m, with own funds of £3,000 m and SCR of £1,500 m so non-RFF coverage ratio=200%.

2. (ii) TMTP in respect of RFF business=£500 m, with own funds of £1,000 m and SCR of £400 m so RFF coverage ratio 250%.

3. (iii) Prior to recalculation of the TMTP, the legal entity solvency coverage ratio=own funds of £3,400 m, with SCR of £1,900 m=179% (i.e. lower than the non-RFF ratio in isolation because there are no surplus assets from the RFF taken into account when determining the ratio).

Risk profile changes have occurred which could impact the TMTP for RFF business differently to that of non-RFF business such that:

1. (i) TMTP in respect of non-RFF business=£950 m, with own funds of £2,950 m and SCR of £1,525 m so non-RFF ratio=193% (i.e. reduction of 7 percentage points in comparison to pre-recalculation)

2. (ii) Say there are three scenarios for RFF business, one where the TMTP reduces (i.e. (a) below), one where the TMTP increases (i.e. (b) below) and one where the TMTP is unchanged (i.e. (c) below)

1. a. TMTP in respect of RFF business=£480 m, with own funds of £980 m and SCR of £400 m so RFF ratio=245% (i.e. reduction of 5 percentage points)

2. b. TMTP in respect of RFF business=£520 m, with own funds of £1,020 and SCR of £400 m so RFF ratio=255% (i.e. increase of 5 percentage points)

3. c. TMTP in respect of RFF business is unchanged so RFF ratio=250%

1. (iii) After recalculation of the TMTP, the legal entity solvency coverage ratio=own funds of £3,350 m, with SCR of £1,925 m=174% (i.e. reduction of 5 percentage points) regardless of whether scenario a, b or c has occurred because the level of the RFF SCR is unchanged and there are no surplus assets.

It is therefore possible to have a change in risk profile in respect of RFF business in isolation, non-RFF business in isolation or both together that may result in a change in legal entity coverage ratio that is below the PRA’s indicative view that a material change in risk profile would result in a movement of 5 percentage points or more, even though when considering the business separately, as most firms do for the purpose of risk management, the change in coverage ratio exceeds 5 percentage points. The working party view is that the legal entity solvency coverage ratio for the purpose of calculating the impact of a TMTP recalculation may require adjustment where RFF restrictions exist. Such an adjustment would likely need to be discussed and agreed with the PRA.

Can firms have triggers in their policies different to the 50 bps and 5% cover ratio in SS6/16?

The PRA expects firms to develop their own policy setting out “how the design and calibration of the triggers are related to the firm’s risk profile.” The SS also requires firms when they are making an application for approval to recalculate to consider the PRA’s indicative view of a material change in risk profile (i.e. a sustained change of 50 bps and 5% movement in coverage ratio). However, the SS also states, in the context of the cover ratio, that “the degree of change that is considered material will depend on individual firms’ risk profiles.” As a result, the SS leaves it open to firms to justify a higher or lower threshold based on their own firms’ sensitivities. This includes demonstrating that the chosen threshold would not prevent a reduction in the TMTP when circumstances imply reduced need for the TMTP, i.e. that triggers are symmetrical.

If PRA invite a recalculation should it be compulsory?

SS6/16 refers to the PRA inviting firms to recalculate the TMTP where external market-wide events lead to a change in market conditions that would be likely to cause a material change in firms’ risk profiles. However, firms are required to justify that the market-wide event had a material impact on their own risk profiles in order to carry out recalculation. It is recognised that events could affect firms differently, having material impacts on some firms and little impact for others. It follows that recalculation should not be compulsory and would instead depend on how a particular event affected different firms. Firms could though be expected to justify why recalculation was not appropriate, for example, because it was not in line with the policy or because, based on recent sensitivities, a particular event would not have a material impact. The working party would hope the overhead of demonstrating why recalculation is not appropriate would be less onerous than seeking approval for a recalculation. Notwithstanding this point, the recalculation estimate produced to support a recalculation or non-recalculation should be equivalently robust.

Firms are expected to have symmetric triggers for recalculation. They would, therefore, be expected to show non-recalculation was in line with a symmetric policy and not, for example, influenced by the likely direction of change on recalculation.

5.2.1. Segregation of business

Recalculation of the TMTP requires firms to segregate Solvency I and Solvency II technical provisions into the parts in respect of pre- and post-1 January 2016 business. This is because the TMTP only applies to business written before Solvency II came into force. Further, and depending on interpretation of the FRR limitation, capital requirements and non-technical provisions may need to be split between the two business classes. Such segregation is likely to need reporting systems to be updated to be able to identify the classes and to report them separately. Where underlying assets affect the calculations, for example for with-profits business and annuities with a Matching Adjustment applied, an additional level of asset hypothecation may be indicated.

This provides both a calculation granularity challenge, as well as the challenge of being able to reconcile and explain reported results for pre- and post-1 January 2016 business.

Issues relating to this segregation are outlined in more detail in section 5.3.2 for the unrestricted TMTP and section 5.4.5 for the restriction of the TMTP.

5.2.2. Maintaining the prior regime

Recalculation of the TMTP requires firms to maintain aspects of models and methodology from the previous solvency regime. The working party expects the full maintenance of Solvency I models, methodology and associated governance, in combination with the enhancement of granularity for pre-1 January 2016 business, would be excessively onerous for most firms to implement.

The working party therefore expects a TMTP recalculation will need simplified, and proportionate, methodologies. Proportionality is expanded upon further below.

5.2.3. Proportionality

The PRA supervisory statement on a TMTP recalculation (SS 6/16) implies that proportionate approaches may be acceptable, indicating that firms should discuss proposed methodology with their supervisors. This is helpful and leaves it open for firms to agree with the PRA what would be proportionate for their own circumstances. What would be proportionate is difficult to define and will vary from firm to firm and legal entity to legal entity. A view on proportionality is ultimately an area of judgement which needs to be agreed internally within a firm and, as outlined in SS6/16, discussed with the PRA.

One potential way to consider proportionality is as follows.

Each of these is expanded upon below.

Recalculate or not?

A firm should define areas of recalculation which will be theoretically maintained going forward, and define those areas of recalculation which it believes on grounds of principle should not be maintained.

An example of an area of recalculation which might not be maintained, on the grounds of principle, is the maintenance of separate asset hypothecations for Solvency I and Solvency II technical provisions.

Maintaining a separate (theoretical) asset hypothecation based on what investment strategies might have been had Solvency I continued seems a step too far, and likely to be difficult to achieve in practice. However, using the asset allocation adopted under Solvency II to determine Solvency I technical provisions for recalculation is likely to lead to an erosion of the TMTP value because firms will not be purchasing assets to optimise their Solvency I technical provisions.

Where such erosion was significant, firms could presumably propose simplified methods to allow for Solvency I asset allocations in recalculation. For example, a firm may assume the asset allocation in force at 31 December 2015 applies to TMTP recalculations going forward.

Materiality framework

A firm should agree a materiality framework to apply to those areas of recalculation, which will be maintained going forward. Firms could define the recalculation materiality relative to its general materiality and risk appetite frameworks.

“Exact” calculation

A firm should define what it believes an “exact” calculation is. We use the terminology “exact” here to note that modelling in general is a simplified representation of the real world.

The working party expects a typical definition of “exact” as being the maintenance of the methodology and processes which applied at 1 January 2016. Notwithstanding this point, it is possible some firms may develop processes and methodologies after 1 January 2016, which provide a more precise value of the TMTP than those which applied at 1 January 2016.

Understanding onerous

As part of defining an “exact” calculation, firms should understand which parts of recalculation would be onerous to carry out exactly. If the “exact” calculation is not onerous then additional simplifications will not be required.

Propose initial simplification(s)

Firms will propose an initial simplification or simplifications on onerous parts of recalculation, justifying them relative to the “exact” calculation.

Understand limitations

Understand the limitations of the simplification, i.e. in what circumstances the simplification would fall outside of the firm’s TMTP materiality framework.

Part of understanding limitations of simplifications may include the use of qualitative and quantitative methods. For example, a firm may use qualitative methods to outline theoretical analysis to outline when a recalculation may fall outside of the materiality limit, and use sensitivity analysis to provide an expectation of this occurring.

Amend proposal if required

If the limitations are deemed to be too material, then the proposal(s) should be amended and assessed again.

5.2.4. Firm specific

Recalculation methodology will depend on the following factors:

Materiality of the TMTP

A TMTP which is material relative to the capital surplus (i.e. the surplus of capital resources over capital requirements) is likely to require a more complicated, more robust and more frequent recalculation than a TMTP which is less material relatively to the capital surplus.

The type and complexity of the underlying business

The recalculation methodology is likely to depend on the type and complexity of the underlying business written in the legal entity, or ring-fenced fund. For example a mono-line writer of business, e.g. annuity business, may be able to employ more accurate recalculation methods than a multi-line insurer.

Furthermore certain types of business may be easier to derive a TMTP in certain aspects than others. For example, it may be easier to split annuity policy data than unit-linked policy data.

Whether the legal entity, or ring-fenced fund, is open or closed to new business post-1 January 2016

A TMTP recalculation for a legal entity, or ring-fenced fund, which was closed to new business prior to 1 January 2016 is significantly simpler than for a legal entity, or ring-fenced fund, which is still open to new business. The requirement to segregate business written pre- and post-1 January 2016 is trivially met by there being no post-1 January 2016 business.

Similarly, where a legal entity or ring-fenced fund writes only immaterial volumes of new business post-1 January 2016 it may be reasonable to simplify the recalculation of its TMTP by ignoring the impact of post-1 January 2016 business.

Whether a TMTP is restricted by or close to the FRR comparison

Where a legal entity, or ring-fenced fund, has a TMTP that is based on, or close to, the FRR comparison test, the calculation will require a more robust derivation of the FRR comparison test components than for a firm whose FRR comparison test is less likely to “bite.”

Notwithstanding this, all firms seeking a recalculation are expected to provide details of the FRR comparison in their applications for recalculation or justify why the test would not restrict the TMTP.

Which of the two Solvency I Pillars is relevant for the FRR comparison test

The recalculation of the TMTP will depend on which of the Solvency I Pillars is the more onerous for a legal entity, or ring-fenced fund. Where one of the Solvency I Pillars dominates and this can be expected to continue going forwards it may be justifiable to exclude the other Solvency I Pillar from the calculation. This simplification is explored in more detail in section 5.4.6 and would clearly need to be discussed with the PRA.

Reporting systems and process constraints

The complexity, change agility, and structure of the reporting systems and processes used to derive the TMTP will vary from firm to firm. Some firms may have sufficient granularity in their processes, or be able to change the granularity, to derive the segregation of pre- and post-1 January 2016 as part of their regulatory quarterly reporting process. For other firms, this level of granularity may not be available without significant investment.

5.2.5. Evolve over time

The working party expects that the processes and methodology applied in recalculating the TMTP will evolve over time. Over the short to medium term, we envisage the complexity of recalculation will tend to increase. This reflects a combination of the following factors:

• ∙ Increasing materiality of post-1 January 2016 business for a legal entity, or ring-fenced fund, open to new business.

• ∙ Material levels of TMTP on balance sheets.

• ∙ Infancy issues from developing and embedding suitable TMTP calculation methodologies, processes, and associated governance. For example, issues relating to the reconciliation of the components of the FRR comparison test back to the reported values in the QRT forms. Here we note that the FRR comparison test may only apply to business written prior to 1 January 2016, while the reported values in the QRT forms will include all business.

• ∙ Technology changes potentially offering different approaches and improvements or refinements.

However, over a longer time horizon we expect that the calculations used to derive the TMTP will become simpler reflecting the reduction in the value of the TMTP as pre-1 January 2016 business and the TMTP run-off. This may involve the simplification of processes or the use of margins applied in TMTP recalculations to allow for any approximations.

Ultimately firms may decide to stop using the TMTP before the 16-year transitional period is up if the benefit becomes less material.

5.3.1. Double run-off

Condition 1 of regulation 54 of the UK Solvency 2 Regulations 2015 introduces the issue of double run-off. Specifically, it includes an equation stating how the calculation of the TMTP is performed. The equation for the TMTP, T _N is as follows:

$$O\leq T_{N} \leq \left( {X_{N} {\minus}Y_{N} } \right)\left( {1{\minus}{N \over {16}}} \right){\rm }\left( {Where \;{\rm }N{\equals}0,1,2\,\ldots\,16} \right){\rm }$$

Where X _N and Y _N are the Solvency I and Solvency II technical provisions, calculated at time N, and N is the number of years from 1 January 2016.

Therefore, at the point of recalculation, the equation implies that only the business still in force can be included in the calculation and not the entire business that was in force as at the original point of recalculation. The need to apply the run-off factor, to ensure TMTP are zero by the end of the transitional period leads to a further reduction in TMTP.

Figure 6 below shows an illustration of the issue. The top line represents the 1/16^th pa straight line run-off. The bottom line shows the impact of double run-off on the recalculated TMTP, with the capital strain illustrated by the bars.

Figure 6

Double run-off example

The working party understands that the PRA has acknowledged that this was not intentional and stated that firms could propose their own methodology if they wish to avoid this issue.

Potential solutions

The first “solution” proposed is clearly the simplest – firms could accept the double run-off issue and not try to avoid it. This could be a viable option for firms with long-dated liabilities that have a slow run-off rate.

The working party is of the opinion that while this may be a suitable option for some firms, it will not be suitable for all. A key consideration that firms may have is, if accepting the issue results in the capital position of the firm being constrained, the impact this could have on policyholder outcomes.

Three potential approaches to avoiding the double run-off issue are below:

1. Run-off TMTP evenly over the remaining part of the 16-year period

The first approach avoids the double run-off issue by negating the amortisation factor part of the calculation. This would be done by calculating the difference between technical provisions at the relevant re-calculation date but instead of using the 1/16^th pa amortisation factor it would amortise from 100% to zero over the remaining transitional period (i.e. if a recalculation were performed after 2 years the factors applied in the future would be 14/14, 13/14 etc..).

To ensure compliance with the regulation 54 of the UK Solvency 2 Regulations 2015 and ensure the presentation of run-off in the TMTP QRT is aligned to EIOPA’s log file, firms may wish to scale up the reported TMTP, as at 1 January 2016, by applying a scalar of 16/(16-N), where N is the point of calculation.

This approach run-offs the TMTP in line with the underlying business and would seem to be very appropriate to use if the business runs-off at broadly the same pace as the TMTP (i.e. 1/16^th p.a.)

It also seems to be appropriate if the business ran-off quicker than 1/16^th p.a. This is because the TMTP would run-off more quickly than the original planned 1/16^th rate. However, for business that runs off slowly this could lead to capital strains, as the TMTP is required to reduce at a faster rate to ensure they are zero by the end of the transition period. Figure 7 below shows this based on biennial recalculations of the TMTP.

Figure 7

Progression of TMTP with adjusted amortisation factor, slow run-off of business

For example, towards the end of the transitional period, the value of TMTP is required to reduce by half over the year.

An advantage of this approach, as opposed to the others proposed, is that it does not require firms to retain any further data over time. This is shown in an example below.

2. Scale the business in-force at the recalculation date

An alternative approach could be to adjust the TMTP to avoid the actual run-off of the business. This could be achieved by scaling the calculated TMTP at the recalculation date. The scalar would be set so that it reflects the actual business run-off since 1 January 2016 so that the level of TMTP is consistent with the business in force at 1 January 2016.

The choice of scalar used in this method needs careful consideration to ensure that it meets the objective of resolving the double run-off issue. For example, firms could consider that the change in risk margin since the last calculation of the TMTP is a proxy to the TMTP. However, firms calculate the risk margin on the current business in-force and therefore a scalar based on risk margin would not address the double run-off issue.

Further using a scalar based on operating conditions at the point of recalculation could be problematic. This is due to operating conditions causing the item to fluctuate in a way that does not actually match the run-off of business over time.

This leads us to suggesting simple measures such as sum assured or policy count of the business benefitting from the TMTP could therefore be appropriate. The box below provides a worked example.

However, simple measures, such as policy count, do not directly compare to the value of the TMTP on a one-to-one basis. Therefore, care will be required to ensure that the TMTP allocated to groups of homogeneous policies to ensure that the policy count proxy is suitable. This could lead to a large number of additional calculations for firms with diverse types of business.

3. Recalculate liabilities at 1 January 2016

An alternative approach could be to avoid any double run-off by using the actual business in force at end 2015 to calculate the TMTP. The historic models, or more probably just the policy data from end 2015, could be retained over the transitional period.

The data, and if relevant models, would need to be adjusted, at the point of recalculation, to the operating conditions at that time The factor (1−n/16) would then apply to the TMTP calculated on the end-2015 policy data.

The requirement to update the model point files could be difficult for some firms, particularly if some of the assumptions used are duration based instead of calendar year specific. Firms would, therefore, need to ensure that their processes could deal with these types of changes.

5.3.2. Splitting business into pre- and post-1 January 2016 segments

The TMTP only applies to business that was written prior to Solvency II being implemented. Therefore, each firm will need to decide on a definition of how to classify what is pre-1 January 2016 business and what is not in the unrestricted TMTP calculation.

Potential solutions

The decision, on whether a split of business is required, and how to perform it, will depend on the level of new business written in each firm.

For books that closed to new business prior to 1 January 2016, this could be potentially simple. If no business has been written post-1 January 2016 then all business can be treated in scope of the TMTP and no split would be required. This could still be valid, on materiality grounds, for funds that are only writing a limited amount of new business, e.g. top-ups to existing contracts or annuities written due to a guarantee being exercised.

For open books, especially those that are writing a material amount of new business, it will be necessary to split the technical provisions of this business into pre- and post-1 January 2016 segments.

Each firm will need to consider both the business being written and the historic business the firm holds when determining how to split the business. Any split will bring computational and process challenges that will need to be overcome.

Potential methods to split the business could be:

• ∙ Split the business by date of entry and ensure that the administration and valuation systems can separately identify this business;

• ∙ Maintain the model point files at 1 January 2016, and update for market and assumption changes, as appropriate and use these for a TMTP recalculation;

• ∙ Allocate the business to pre- and post-1 January 2016 by using suitable proxies, such as estimates of the business written since 1 January 2016.

It follows that the method to split the business should also be appropriate to the run-off method that being used.

A key element of the calculation of the unrestricted transitional is the risk margin and this is covered below.

5.3.3. Changes to the discount rates used under Solvency I versus Solvency II

The Solvency II regime prescribes the basic risk-free curve that the liabilities of a firm should be valued on, e.g. for business written in the UK it is based on the UK swap curve. This basic curve can then be adjusted, following a successful application to the PRA, through either the MA, or the volatility adjustment.

Under Solvency I, firms had more flexibility to set discount rates based on the underlying business, and in some cases could include an illiquidity premium based on the assets it held. Therefore, in moving to the Solvency II basis, some firms could have experienced large changes in technical provisions purely from a change in discount curve used. This could be particularly significant for funds that have long-dated liabilities or guarantees in them.

Recalculation of a TMTP requires firms to maintain at least an underlying basis for calculating Solvency I Pillar 2 technical provisions. Therefore, firms are required to determine how to construct an equivalent Solvency I risk-free curve at the point of recalculation, given that its underlying business, and the assets backing it could have significantly changed since the point of first calculation of the TMTP.

This section splits out the key themes that the working party has identified as the key differences of the risk-free rate between the regimes and suggests potential solutions that could be used to overcome the issues.

5.3.3.4. Maintain the bps differences between Solvency I Pillar 2 illiquidity premium and Solvency II MA at 31 December 2015 with suitable adjustments

A number of adjustments could be made to the difference between Solvency I Pillar 2 and Solvency II at 31 December 2015 to ensure it more appropriately captures divergences between the internal view and the Solvency II MA view. In particular, the working party notes:

1. a. Adjustments to capture the changes in the Solvency II risk-free rate relative to the Solvency I Pillar 2 risk-free rate

   The Solvency II MA is applied as a non-term dependent addition to the Solvency II basic risk-free curve, excluding volatility adjustment. If firms also applied the Solvency I Pillar 2 illiquidity premium as a non-term dependent addition to the Solvency I Pillar 2 risk-free curve then it therefore follows that the illiquidity premium differential is term-dependent and can be defined as:

   $$\eqalignno{ & {\mib{Illiquidity}} {\mib{premium}} {\mib{differential}} {\rm } \cr & {\equals}{\rm }[Solvency{\rm }I{\rm }Pillar{\rm }2{\rm }risk{\rm }free{\rm }curve]{\rm }-{\rm }[Solvency{\rm }II{\rm }basic{\rm }risk{\rm }free{\rm }curve]{\rm } \cr & {\plus}{\rm }[Solvency{\rm }I{\rm }Pillar{\rm }2{\rm }illiquidity{\rm }premium]{\rm }-{\rm }[Solvency{\rm }II{\rm }matching{\rm }adjustment] $$

   Therefore any divergence in the Solvency I Pillar 2 and Solvency II risk-free curves need to be captured in the illiquidity premium differential. Examples of potential differences include the Solvency II credit risk adjustment and differences in the gilt-swap spreads if the Solvency I Pillar 2 risk-free rate was set on the gilt curve as opposed to the swap curve.

2. b. Weighted average spreads

A firm could decompose the assets held to back liabilities into a notional Solvency I Pillar 2 asset portfolio and derive the spread over the Solvency I Pillar 2 risk-free curve based on the spreads at 31 December 2015, noting that a full asset hypothecation was applied at 31 December 2015.

It could then summarise the key attributes of the portfolio at 31 December 2015 which demonstrates the contribution of each asset class to the Solvency I Pillar 2 illiquidity premium. These may include the:

• ∙ Market value of the asset;

• ∙ The modified duration of the asset;

• ∙ The “gross spread,” i.e. the spread over the risk-free rate ignoring the best estimate deductions for downgrades and defaults;

• ∙ The Solvency I Pillar 2 view of best estimate deductions for downgrades and defaults;

• ∙ And the “net spread” over the risk-free curve, i.e. the “gross spread” less the Solvency I Pillar 2 view of best estimate deductions.

An example is shown in Figure 8 below, with two possible simplified methodologies of deriving the Solvency I Pillar 2 illiquidity premium. Method 1 below derives the Solvency I Pillar 2 illiquidity premium by weighting the net spread by market value of the asset. Method 2 modifies this approach by weighting the net spread by both market value and the modified duration of the assets. Method 1 in the example below understates the contribution of the higher spread, higher duration infrastructure assets.

Figure 8

Simplified methodologies to derive Solvency I Pillar 2 illiquidity premium

The firm could similarly assess the Solvency II MA at 31 December 2015 using a similar method, but noting that:

• ∙ The assets allocated to back liability cash-flows in the MA portfolio will likely differ to those in the Solvency I Pillar 2 illiquidity premium portfolio

• ∙ The Solvency I Pillar 2 best estimate deductions will be replaced by the Solvency II Fundamental Spread

At future recalculation dates, the table and calculation above can be reproduced, but with inputs adjusted in a proportionate way to reflect the situation at the recalculation date. The adjustments made will depend on:

• ∙ Whether the assets are notionally split between those backing pre- and post-1 January 2016 business or whether the same discount rate should apply to both pre- and post-1 January 2016.

• ∙ Whether changes in the assets held in the portfolio are allowed for at the point of recalculation. Alternatively, the Firm may keep the portfolio static as at 31 December 2015. See below for more details on this.

By changing the gross spread to be set on individual asset classes, firms can use the Solvency I Pillar 2 best estimate view of deductions and Solvency II Fundamental Spreads to approximate the Solvency I Pillar 2 and Solvency II illiquidity premium differential. The working party notes that this requires firms to maintain Solvency I Pillar 2 methodology going forward, although this could be required as part of sound risk management reflecting an internal view of asset risk in any case.

In the case that a simplified approach is taken firms will need to ensure that this approach has been agreed internally and discussed with the PRA.

Firms may also wish to consider when the approach used would no longer be acceptable, e.g. the disposal or acquisition of a book of business or the transferring in of business via Part VII transfer etc.

1. c. Captures changes in asset portfolio

Firms may take a view that they should manage their business according to the regulatory regime currently in force, rather than the previous regime. Therefore, an argument could be that the level of a TMTP should not be affected by making ALM decisions that optimise Solvency II, to the detriment of the Solvency I Pillar 2 basis. This approach could lead to ignoring the new assets purchased since 31 December 2015 in the determination of the Solvency I Pillar 2 and Solvency II illiquidity premium differential.

However, there could be material changes in either the Solvency I Pillar 2 or Solvency II asset portfolio which should be reflected in a materially different TMTP. For example if a firm applies for a MA portfolio or securitises Equity Release Mortgages after 31 December 2015 then these events would likely result in a material reduction in the TMTP and so it is the working party’s view that each asset change should be taken on merit. However, the working party considers that it is doubtful that it would have been intended that increased Solvency I risks, resulting from actions to hedge the Solvency II balance sheet, should be brought into the resultant TMTP calculation.

A further option that firms may be able to utilise is the use of the allocation of assets within the IFRS portfolio gross redemption yield. This process is likely to provide insight into the actual assets held at the valuation date, although firms will need to consider whether IFRS fair value margins (effectively prudence margins) applied to both the cost of asset defaults and downgrades, as well the liability demographic assumptions, would result in a different allocation of assets, compared to the asset allocation which arises on a Solvency I Pillar 2 best estimate basis. All else being equal, the removal of prudent IFRS fair value margins is expected to increase the cash flows assets provide and reduce the liability cash flows being matched. This presents firms with an opportunity to “kick out” lower yielding assets and optimise their Solvency I Pillar 2 illiquidity premium.

5.3.4. Other factors that impact on the potential value of TMTP

This section covers the items not addressed previously that need to be considered when recalculating the TMTP.

5.4.1. Overview

As outlined in sub-section 2.2, a restriction of the TMTP applies if the financial resource requirement (FRR) of Solvency II, after the inclusion of the unrestricted TMTP, is lower than the more onerous of the two Solvency I Pillars, i.e. Pillar 1 and Pillar 2.

The FRR comparison test introduces the need to be able to construct the entire Solvency I balance sheet, available capital, and capital requirements. This includes:

• ∙ Solvency I Pillar 2 technical provisions;

• ∙ Other liabilities on the Solvency I Pillar 2 balance sheet (accounting liabilities including deferred tax liabilities);

• ∙ The Solvency I Pillar 2 capital requirements: both the Individual Capital Assessment (ICA) and any additional Individual Capital Guidance (ICG) add-ons imposed by the PRA;

• ∙ Value placed on assets by the Solvency I Pillar 2 regime. There is some debate about whether differences in the valuation of assets can feed directly in to the FRR comparison test, and this is discussed in sub-section 5.4.3. Asset values may also be relevant to the extent that they need to be stressed to derive the capital requirements and differences in asset values in Solvency I Pillar 2 and Solvency II may generate a deferred tax liability difference;

• ∙ Solvency I Pillar 1 technical provisions, i.e. mathematical reserves;

• ∙ Other liabilities on the Solvency I Pillar 1 balance sheet (accounting liabilities including deferred tax liabilities);

• ∙ The Solvency I Pillar 1 capital requirements: the Long Term Insurance Capital Requirement (LTICR), for realistic with-profits insurance legal entities the With-Profits Insurance Capital Component (WPICC) and for firms which aren’t realistic with-profits insurance legal entities the Resilience Capital Requirement (RCR);

• ∙ Value of Solvency I Pillar 1 assets, noting again that it is not entirely clear whether asset valuation differences may have a direct impact on the FRR test, but noting further that differences in asset values in Solvency I Pillar 1 and Solvency II may generate a deferred tax liability difference.

Overall the working party believes there are two main considerations here:

• ∙ The Solvency I regime’s models, processes and appropriate levels of governance that need to be maintained going forward;

• ∙ The valuation under the Solvency I regime including a split by pre- and post-1 January 2016 business. This potentially introduces an additional requirement beyond that required under Solvency I, although the PRA has clarified in PS11/17 that they expect the FRR test should be applied to all business, regardless of when it was written, at the recalculation date.

The remaining part of this section outlines the following:

• ∙ Sub-section 5.4.2 explores whether the interpretation of the FRR comparison test and in particular whether it applies to only business prior to 1 January 2016, to all business at the recalculation date, or whether it should only apply for the initial TMTP calculation and not to recalculations.

• ∙ Sub-section 5.4.3 outlines differences in the Solvency I and Solvency II FRR that could lead to distortions in the relative levels of policyholder protection, which firms may wish to consider in their TMTP methodology.

• ∙ Sub-section 5.4.4 describes an ambiguous area relating to whether the FRR comparison test should be applied before or after the 1/16^th run-off. The issues outlined in this section are similar to the “double-run-off” issue outlined in section 5.3.1.

• ∙ Sub-section 5.4.5 outlines issues concerning the segregation of pre- and post-1 January 2016 business.

The remaining sub-sections outline specific issues and potential solutions that firms may employ in relation to the recalculation of the FRR comparison test.

5.4.2. Does the FRR test take into account all of a company’s business or only the pre-2016 business?

The FRR test ensures that the financial resources under Solvency II are not lower than the financial resources that would have been held if Solvency I were still in force.

As outlined in section 5.4.1 above, the PRA has clarified in the policy statement PS11/17 that they expect the FRR test should be applied to all business, regardless of when it was written, at the recalculation date, i.e. option 2 below. However, firms are able to apply the FRR comparison test to only business written prior to the introduction of Solvency II if they can demonstrate that the outcome would not be materially different to a full calculation.

This gives firms a choice over how they implement and interpret the FRR comparison test calculation, namely:

1. 1) The cap applies to just pre-1 January 2016 business at the recalculation date.

2. 2) The cap applies to all business, regardless of when it was written, at the recalculation date.

3. 3) The cap only applies at 1 January 2016.

1) The cap applies to just pre-1 January 2016 business at the recalculation date

This interpretation is that the test only applies to the business still in force at the recalculation date that was also in force prior to the introduction of Solvency II. This is equivalent to an interpretation that the business written after the introduction of Solvency II does not fall within the capped amount, in line with the TMTP calculation as at 31 December 2015.

The practical challenges associated with this interpretation should not be underestimated. In particular, this requirement in full requires the segregation of technical provisions, asset investment portfolios and accounting liability trial balances by pre- and post-1 January 2016. Therefore the working party views the PRA’s clarification on this matter as a positive step.

Practical challenges relating to this definition are outlined in more detail in section 5.4.5.

2) The cap applies to all business, regardless of when it was written, at the recalculation date

An alternative interpretation is that the FRR comparison test would imply that all business, including that written after 1 January 2016, needs to be valued on a Solvency I basis and a Solvency II basis to determine whether the FRR comparison test “bites.”

On balance the working party did not view this interpretation as being in line with the original intention of the TMTP, but rather viewed the link between business written prior and after 1 January 2016 as an argument to support a greater degree of approximation in TMTP recalculations. Subsequent to forming this view, the PRA has clarified in PS11/17 that this is how UK firms should apply the FRR comparison test.

The working party has retained this section of the paper, so that the reader can understand the thought process the working party followed in coming to this view.

On the assumption that Solvency II requirements for new business are more onerous than Solvency I requirements, for a firm open to new business, the restriction arising from the FRR test would become less likely to “bite” if applied to all business in an undertaking, as compared to a FRR test applying to business written prior to 1 January 2016. Over time this would likely result in a TMTP which transitions towards unrestricted technical provision differences. This is viewed as beneficial from both a practical and explanatory perspective as it reduces reliance on approximations and detailed knowledge of the Solvency I regime. However, the impact of new business depends on a firm’s recalculation methodology and how the relationship between Solvency I and Solvency II has been determined, in particular with regard to any approximations used.

It is also noted that the Solvency II regime’s capital resources and requirements are derived at an undertaking level, or within a ring-fenced fund of an undertaking. There is not a concept of separate capital resources and requirements for business written prior to or after 1 January 2016 within an undertaking. This means the Solvency II regime recognises offsets (e.g. directionally opposite exposures, MA test requirements, etc) and diversification benefits within an undertaking across business regardless of the prior to and after 1 January 2016 segregation.

This means that the financial resources of business written prior to 1 January 2016 are inextricably linked to the financial resources of business written after 1 January 2016. The segregation of the financial resources across these lines would potentially introduce treating customer fairly consideration around the fairness of treatment of policies written prior to and after 1 January 2016. These considerations could be difficult to reconcile where a TMTP arises on differences in valuation methodology, i.e. contract boundaries treatment, rather than genuine differences in a contract’s commencement date.

3) The cap only applies at 1 January 2016

A further interpretation could be that the cap is required to be calculated at 1 January 2016 but not for subsequent recalculations. The PRA has made clear in SS6/16 that this requirement remains.

5.4.3. Differences between Solvency I and Solvency II to consider in the FRR test

The Statutory Instrument requires that the TMTP be limited so that the “financial resources” required by Solvency II (after allowing for the impact of TMTP) be at least as great as the financial resources required by the Solvency I regime. The Statutory Instrument requires that the financial resources are no less than those which would be required to be maintained in accordance with GENPRU 1.2.26 R. The Excel template the PRA requires firms to complete when making a TMTP application defines the FRR to be the sum of:

• ∙ Mathematical reserves/insurance liabilities/technical provisions

• ∙ Liabilities other than insurance liabilities

• ∙ Capital requirements

There are, however, a number of differences between the Solvency I and Solvency II regimes that are not captured by this definition of FRR. By not capturing the following differences firms’ FRR may not be comparable across regimes and this could result in inappropriate outcomes from recalculating the value of the TMTP.

The working party believes that firms should identify whether the issues outlined below apply in the recalculation of the FRR. If they do, then firms may wish to consider modifications in their TMTP methodology to remove these distortions. These modifications will need to be agreed internally and discussed with the PRA.

Differences in regimes, particularly in respect of with-profits funds

Under the Solvency I Pillar 2 regime the contribution of with-profits funds to a proprietary company’s consolidated position typically allows for the present value of future shareholder transfers. These future shareholder transfers vary in nature including the following types:

• ∙ Those future shareholder transfers related to distributions, which have been fully allocated to asset share,

• ∙ Those future shareholder transfers related to distributions that are planned but not fully allocated to asset share (planned enhancements), and

• ∙ The shareholder’s share of further distributions of the with-profits fund’s estate which are neither planned nor allocated to asset share.

This adverse impact is in part a consequence of the Solvency I Pillar 2 regime having no single presentation of the contribution of with-profits funds to a proprietary company’s consolidated position and the working party understands that a number of approaches have been taken across different companies. Examples of the different approaches taken in Solvency I Pillar 2 include:

1. 1) Including a present value of shareholder transfers asset in a shareholder fund. This may include some or all of the different types of shareholder transfers.

2. 2) Excluding the with-profits fund’s capital resources and requirements, and only allowing for the shareholder’s value in the with-profits fund not captured in the shareholder transfer asset, e.g. the shareholders’ share of the with-profits fund’s estate.

3. 3) Notionally increasing the value of liabilities of a with-profits fund such that the value of assets in excess of liabilities targets the shareholders’ value in the with-profits fund not captured in the shareholder transfer asset, e.g. the shareholders’ share of the with-profits fund’s estate.

4. 4) Notionally write down the value of available capital of a with-profits fund such that its value targets the required capital of the with-profits fund plus the shareholders’ value in the with-profits fund not captured in the shareholder transfer asset, e.g. the shareholders’ share of the with-profits fund’s estate.

Under Solvency II, the initial FRR of the firm includes contribution from ring-fenced funds including with-profits funds. This could lead to an adverse impact on the shareholder of a proprietary firm, even where the ring-fenced funds have surplus capital resources over capital requirements. Examples of this include:

• ∙ Differences in the treatment of planned enhancements under Solvency I Pillar 2 and Solvency II, to the extent they impact the overall legal entity FRR.

• ∙ Differences in the presentation of capital requirements. Solvency II capital requirements include the notional Solvency Capital Requirement of ring-fenced fund, whilst some firms presented their ICA and ICG capital requirements including only the shareholder’s interest in the with-profit fund.

• ∙ Differences in the shareholder value in the with-profits fund. Solvency I Pillar 2 tends takes a broader view than Solvency II. For example, Solvency I Pillar 2 may include the shareholder’s interest in the estate, while Solvency II shareholder value is limited to the share of bonuses recognised in the technical provisions.

Any differences on the asset side of the balance sheet

Examples of this include:

• ∙ Valuation differences, for example due to accounting rules;

• ∙ Presentation of accounting assets and liabilities, which have the same value on assets net of liabilities;

• ∙ Differences in asset inadmissibility;

• ∙ Differences in the value of participation assets. Noting Solvency I Pillar 2 would generally be based on an economic look-through valuation and Solvency II will reflect accounting rules.

Differences in treatment of ancillary expense companies

Where a firm uses an ancillary expense company, which charges fees to the legal entities for expense services, and this ancillary expense company is not a subsidiary of the legal entity, this can result in differences in the Solvency I Pillar 2 and Solvency II FRR. This difference results from the use of expense assumptions in Solvency I Pillar 2 technical provisions which “look through” to the underlying expenses incurred and contrasts with Solvency II technical provisions which are valued based on the fees charged.

Differences in capital resources derivation

Examples of this include:

• ∙ The treatment of subordinated debt, including the Solvency II transitional arrangements in relation to subordinated debt issued prior to 1 January 2016;

• ∙ Differences in presentation of with-profits fund (covered above).

5.4.4. Is the FRR test applied before or after the 1/16^th run-off reductions?

The statutory instrument is unclear on how the FRR comparison test applies at a point other than the original point of calculation.

One interpretation of the rules is that the FRR comparison test is static and is applied post the 1/16^th reductions that are performed each year. This would mean a level TMTP for each year until the TMTP value reduces below the FRR comparison test set at outset, before reducing linearly once the amortised value of TMTP falls under the cap value.

An alternative interpretation however is that the FRR comparison test should also reduce over time, meaning that if the cap bites at outset that it is the capped value that the 1/16^th deductions get applied to.

SS17/15 provides a little more detail and states that: “The PRA’s default assumption is that once any limit on the amount of the TMTP has been determined at outset, the limitation will not need to be assessed again unless the TMTP is recalculated at either the firm’s or the PRA’s initiative. Assuming the TMTP is not recalculated, it is expected to run-off linearly each year from its starting amount until it reaches zero.”

The first part of this statement implies that at the point of recalculation the FRR comparison test is reset as part of the calculation routine. It is not clear however whether the FRR comparison test should be applied pre or post the amortisation factor, or indeed if actions that may have been taken to address the “double run-off” issue outlined in section 5.3.1 for the unrestricted TMTP can also be applied to the FRR comparison test.

It is the view of the working party that the run-off applies after the application of the FRR comparison test. The working party suggests that approach taken should be agreed with the PRA and should form part of the firm’s recalculation policy. This may also be an area that should be brought to the attention of the firm’s Audit Committee if there is uncertainty as to how the rules should be interpreted.

It is also important that firms consider how the TMTP runs off in their capital planning. Whilst the value of TMTP in, say, 2017 may depend upon how the rules are interpreted, the value will ultimately run-off to zero in 2032: so a slower run-off in the early results will result in a faster run-off in later years.

5.4.5. Segregation for pre- and post-1 January 2016 business

As outlined in section 5.4.1 above, the PRA has clarified in the policy statement PS11/17 that they expect the FRR test should be applied to all business, regardless of when it was written. In addition to this, the PRA has also clarified in PS11/17 that:

Nevertheless, a firm may be able to use just the business written before the introduction of Solvency II as part of a proxy calculation (for the business written after), as long as it can demonstrate why using this methodology is appropriate and does not result in a materially different outcome to a full calculation. Footnote ⁸

This clarification is helpful because it allows a firm to determine the FRR comparison test based on either all the business in a legal entity, or, subject to appropriately demonstration, only to business written prior to 1 January 2016.

For firms that wish to apply the FRR comparison test to business written prior to 1 January 2016, the recalculation of the TMTP requires firms to segregate the Solvency II and Solvency I balance sheet and capital requirements into the part only applying to business in force as at 1 January 2016 (“pre 1 January 2016 business”).

This requirement, if applied as accurately as theoretically possible, is very onerous for firms because it requires further development to reporting systems and processes. In addition to this, new methodology may be required to consider how the segregate the balance sheet and capital requirements, in addition to the technical provisions outlined in sub-section 5.3, challenges include:

• ∙ Where a split of pre/post 1 January 2016 business is required, how are the assets split? This includes both unit-linked assets (and associated stresses) and non-linked assets (and associated stresses).

• ∙ How do you allocate surplus assets and assets backing capital requirements to pre- and post-1 January 2016 business segments?

• ∙ How do you split accounting liabilities such as reinsurance payables, accruals for commission payments, provisions for reconciliation mismatch, unearned investment income and gains between pre- and post-1 January 2016 business?

• ∙ How do you split deferred tax liabilities between pre- and post-1 January 2016 business?

• ∙ How do you allocate diversification benefits between pre- and post-1 January 2016 business? This includes diversification within pre-1 January 2016 business and diversification benefits between pre- and post-1 January 2016 business.

It is noted that it is only the relative differences in the FRRs, which generate differences in the value of TMTP. In other words if a liability has the same value in Solvency 1 Pillar 1, Solvency 1 Pillar 2 and Solvency II, it will not result in a difference in the TMTP.

Potential solutions

The problems outlined above are difficult to solve and the working party view is that any resources employed solving these problems could be better employed elsewhere. Therefore, the working party believes that firms should be able to pragmatically allow for these issues without a full recalculation, including additional pre- and post-1 January 2016 business split. We believe there are two basic solutions which can be employed here:

• ∙ Recalculate the FRRs as at 1 January 2016 using models, data and processes in force at that date and appropriately allow for the run-off of the business.

• ∙ Recalculate the FRRs as at the recalculation date, and then attempt to split into pre- and post-1 January 2016 business. The splitting of the technical provisions was discussed in sub-section 5.3.2. Suitable proxies may be needed to derive an approximate split of the capital requirements or, alternatively, firms may propose not to split the capital requirements/liabilities other than insurance liabilities and instead split only the technical provisions.

5.4.6. Maintaining Solvency I models

Overview of issue

Maintaining models for the Solvency I regimes could create additional complexity for firms. Carrying out runs of the Solvency I Pillar 2 capital model, and maintaining Solvency I Pillar 1 processes, alongside the firm’s Solvency II model has the potential to create significant extra work for reporting teams.

It is unclear the extent to which firms are expected to keep their Solvency I Pillar 2 models up to date. For example, a firm’s Solvency II internal model may be calibrated using the latest available data and therefore only be updated when additional data is received. For example, when the latest year’s mortality experience becomes available. The model may also be updated to take account of new research and the development of new modelling techniques.

As well as maintaining Solvency I Pillar 2 capital models, firms may need to maintain records of how prudential margins are incorporated into Solvency I Pillar 1 technical provisions. Firms may have used different calculation models to determine their Solvency I Pillar 1 technical provisions and different processes to extract and aggregate the results.

This requirement is slightly mitigated from the requirements of IFRS 4 for insurance business which may use aspects of the firm’s historic Solvency I Pillar 1 PRA valuation reflecting the evolution of IFRS 4 from a modified statutory solvency basis. However, with the future implementation of IFRS 17, firms may not be able to rely on IFRS valuations for determining the Solvency I Pillar 1 technical provisions in the future.

Potential solutions

In this section, we outline potential solutions to reduce the onerousness of maintaining Solvency I models.

Ignoring FRR comparison test

If the FRR restriction does not bite at 1 January 2016 and is unlikely to bite in the future, it could be assumed that the cap would never bite and not recalculated. In particular firms could apply to amend the FRR definition in TMTP calculation to make this more likely as outlined in section 5.4.3.

This has the clear benefit of being simple to maintain going forward. However, this proposal may not be appropriate for firms which have an FRR restricted TMTP or have a TMTP which is close to the FRR restriction.

Overall it is unlikely this approach would be permitted by the PRA, reflecting the requirement to quantify the FRR comparison test in recalculation applications.

Recalculate Pillar 1 or Pillar 2

For legal entities where one Solvency I Pillar is materially higher than the other then it would seem reasonable for firms to only assess the impact of the most onerous Solvency I Pillars in future recalculations.

Firms may wish to include qualitative and quantitative arguments, including sensitivities and the rolling forward of Solvency II day 1 calculations, to support the limitations and reasonableness of this approach. The approach may also include triggers where the approach should be reassessed, for example material acquisitions of pre-1 January 2016 business, change in reinsurance or disposals of pre-1 January 2016 business.

Deriving Solvency I Pillar 2 ICG capital requirements

If a firm’s Solvency I Pillar 2 and Solvency II capital requirement processes are sufficiently aligned then a firm may wish to develop simplistic way of adjusting the Solvency II SCR to derive an approximate ICG capital requirement. In the case that a simplified approach is taken firms will need to ensure that this approach has been agreed internally and discussed with the PRA

This approach would need to consider the following factors:

• ∙ Alignment of aggregation methodology between Solvency I Pillar 2 and Solvency II SCR. If this is possible, this means the firm can use the Solvency II processes for both ICA and Solvency II capital requirements in its recalculations.

• ∙ Remove binary differences in regimes, e.g. Solvency I Pillar 2 cost of closure. These are the differences in capital requirements that are clearly outside the scope of the other regime.

• ∙ Firms may wish to preserve the value of Solvency I Pillar 2 binary items, not included in Solvency II SCR, like cost of closure to reflect the value as at 31 December 2015 to avoid having to maintain Solvency I Pillar 2 specific methodology going forward.

• ∙ Derive conversion factors or recalculate capital requirements for material risks, to estimate the ICA based on the Solvency II SCR.

• ∙ Create process for reviewing the appropriate factors. The two yearly expected recalculation would appear to be the natural review point.

• ∙ Maintain 1 January 2016 ICG level of capital requirement (see next section for further details on this point).

• ∙ Use the Solvency II SCR process, with adjustments outlined above to recalculate the ICG in future TMTP recalculations.

Deriving Solvency I Pillar 1 mathematical reserves and capital requirements

If a firm needs to recalculate the Solvency I Pillar 1 FRR in its TMTP recalculation, then adjustments to the Solvency II SCR are less likely to be appropriate than they were for ICG. This is because there are fewer similarities between Solvency I Pillar 1 and Solvency II, than there are between Solvency I Pillar 2 and Solvency II. Therefore firms are required to develop a simplistic way of recalculating the Solvency I mathematical reserves and capital requirements. The Solvency I capital requirements are the Long Term Insurance Capital Requirement (LTICR), for realistic with-profit insurance legal entities the With-Profits Insurance Capital Component (WPICC) and for legal entities which are not realistic with-profits insurance legal entities the Resilience Capital Requirement (RCR).

There are a number of potential simplifications, which firms may consider to reduce the onerousness of recalculating the Solvency I FRR. These include the following:

• ∙ Derive sensitivities at YE15 and use these sensitivities and movements in observed market indices to estimate the Solvency I Pillar 1 FRR in future TMTP recalculations.

• ∙ Consider whether to recalculate all aspects of the Pillar 1 capital requirements. In particular aspects relating to With-Profits funds such as the WPICC might be de-scoped for future recalculations where the TMTP primarily relates to shareholder owned sub-funds.

• ∙ For Internal Model firms the margins for adverse deviation (MADs) and provisions for adverse deviations (PADs) could be converted into appropriate levels of risk, i.e. persistency margins are equivalent to a 1-in-20 “up” level lapse rate event, i.e. the lapse rates have increased to a level associated with a 1-in-20 year event, where this level of probability is deemed to be equivalent to the Solvency I Pillar 1 prudent assumptions.

• ∙ Firms may wish to derive the portfolio gross redemption yield (PGRY) based on a simplistic mechanical calculation rather than maintain a Solvency I asset allocation. This could be based on the IFRS calculation.

• ∙ Firms could simplify the derivation of the Solvency I “Paid up” test where the sterling reserve for unit linked contracts is valued as the most onerous of the premium paying and paid up reserves. They could assume all contracts are paid up or all contracts are premium paying.

In the case that a simplified approach is taken firms will need to ensure that this approach has been agreed internally and discussed with the PRA.

5.4.7. Reviewing Individual Capital Guidance

Overview of issue

Under the previous Solvency I Pillar 2 regime, firms used their own models to determine their capital requirements; these capital requirements were known as the ICA. The PRA then had the ability to require firms to hold additional capital over and above the ICA: with ICG capital add-on issued and the total capital being termed “individual capital guidance” or ICG.

In CP47/16, the PRA has indicated that upon request from firms, or if it believes the ICA basis is not maintained appropriately, it will consider conducting a proportionate ICG review. However, the working party expects these ICG reviews to be less frequent than they were under the Solvency I regime, and these reviews would predominantly apply following a large transactional change, e.g. a Part VII transfer of business.

When carrying out the FRR test, firms must make an assumption about how the ICG changes over time as the business runs off. There are a number of issues that arise upon recalculation:

• ∙ What assumptions are used for the ICA calculation? e.g. those used as at 1 January 2016, or should they be updated?

• ∙ If the calculation methodology has changed in Internal Model, should this be reflected in the ICA?

• ∙ How does a strengthening or a weakening of ICA get reflected in the overall ICG?

The approach taken to deriving ICG add-ons will need to be discussed with the PRA. It seems unavoidable that firms will need to make changes to their ICA models, to ensure that they remain appropriate for the entirety of the period that they are being used as part of TMTP recalculations.

Potential solutions

A very conservative approach could be to assume that the ICG does not change: but this would seem to overstate the capital requirements, since the risks associated with the business should reduce as the business runs off; noting the considerations in section 5.4.4 re whether or not the linear run-off is applied before or after the FRR test.

An alternative approach might be to assume that the ICG add-ons are a fixed percentage of the ICA: so it changes in response to changes in the ICA. Indeed, the working party understands that the PRA communicated some firms’ ICGs to them in terms of a percentage of ICA. However, this approach does not reflect any change in the mixture of risks. For example, an ICG might have been imposed owing to shortcomings in how a firm modelled longevity risk. If the firm then used reinsurance to remove its longevity risk then is it still necessary for it to allow for this ICG? Similarly, if a component of the ICA is strengthened that had previously attracted ICG it would seem reasonable to write down the capital add-on (subject to PRA approval).

The working party is aware that a number of firms believe that the alignment of assumptions and methodology used in the Solvency I Pillar 2 and Solvency II capital requirement should result in a removal of any ICG add-ons given these are deemed to be sufficiently strong for the purposes of Solvency II. It is not known by the working party whether the PRA shares this view.

What is known is that the PRA has indicated in CP47/16 that firms should maintain consistency between the Solvency I Pillar 2 and Solvency II assumptions and methodologies and these assumptions and methodologies should be reviewed on a regular basis and changed, when appropriate, to reflect changes in operating experience and the firm’s risk profile.

The PRA has also indicated in CP47/16 that upon request from firms, or if they believe the ICA basis is not maintained appropriately, they will consider conducting a proportionate ICG review. In CP47/16, the PRA further indicates that the principle of consistency of assumptions and methodologies does not require the equalisation of all assumption and that the aggregate strength of the valuation basis will be considered in the PRA review of a firm’s ICA and the PRA determination of the ICG.

In the scenario where an internal model firm has aligned its Solvency I Pillar 2 ICA and Solvency II SCR assumptions and methodologies, to the extent permitted under the relevant regulations, and the firm’s internal model has been approved, it is not clear what rationale remains for including any ICG add-ons going forward. In particular, the PRA will have had to consider the aggregate strength of the Solvency II SCR assumptions and methodologies as part of its internal model approvals process, with any individual areas of the model calibrated below the PRA’s benchmark indicators offset against other areas of conservatism within the model.

Overall the working party believes any decision regarding the removal of ICG add-ons is beneficial from the perspective of simplifying TMTP recalculation, but the working party recognises the need for care to ensure policyholder protection is not inappropriately compromised. The decision to remove any ICG add-ons provisions will need to weigh up the original rationale for the add-on compared to its relevance going forward, the materiality of the add-on, the overall strength of the Solvency I Pillar 2 basis and the benefits to both the firm and the PRA from the reduction in the associated administrative overheads.

6.1.1. Overview

Immediately after the inception of Solvency II, hedging and in particular interest rate hedging came under review. Insurers were confronted with the catch 22 situation as to whether to manage their Solvency II balance sheets prior to or after allowance for the TMTP. Further Solvency II introduces a complex conundrum of what metric is best to hedge. For example, protecting your Solvency II coverage ratio is likely to lead to implications on IFRS profits and volatility which needs careful consideration especially when also looking forward to the further added complications of IFRS 17 changes.

Recalculation of the TMTP can be viewed as a tool for hedging the interest rates exposure arising from the Solvency II risk margin. However, there are issues regarding the efficiency of this “free hedging” approach given PRA approval process and triggers for reset. In particular where there is a delay between applying to recalculate the TMTP and receipt of approval, it is not clear how any instability will be dealt with in the interim. While firms might develop internal monitoring mechanisms to more dynamically offset increases in the risk margin with increases in the TMTP, publically reported numbers would not be able to reflect a recalculation without approval. It is therefore unclear what firms may choose to disclose in this regard.

The presence of TMTP, and firms’ ability to recalculate it, should be embedded within firms’ general ALM frameworks. The working party has identified the following areas, which impact the ALM framework of firms and its interaction with TMTP recalculation.

Which metrics to optimise?

Firms need to be clear which metrics they are optimising to and how any hedging action taken interacts with these metrics. These metrics include, but are not limited to, the following:

• ∙ Solvency I Pillar 2 technical provisions,

• ∙ Solvency II technical provisions,

• ∙ Economic balance sheet,

• ∙ Regulatory balance sheet,

• ∙ IFRS balance sheet,

• ∙ Maximising economic surplus capital (surplus of economic balance sheet over economic capital requirements),

• ∙ Maximising regulatory surplus capital (surplus of regulatory balance sheet over regulatory capital requirements).

Whether to hedge the risk margin?

Firms need to consider whether to actively hedge the risk margin, or rely on the TMTP recalculation to provide a free hedge of the risk margin arising on business written prior to the introduction of Solvency II. In particular, firms need to consider whether they want to hedge the additional exposure from movements in interest not deemed to be a material change in risk profile, timing issues arising from the TMTP recalculation process and the risk margin arising on business written after the introduction of Solvency II.

Longer term firms will need to consider how to match the exposure from the risk margin arising from new business written after the introduction of Solvency II, which the TMTP does not cover.

MA and illiquidity premium

Firms need to consider what implications the TMTP has on their MA governance, ALM activity and processes. A firm will need to determine whether to pool asset allocations across liabilities between those written prior to and those written after Solvency II, or to allocate assets separately. These issues are considered in more detail in section 6.2.6 below.

A firm may also wish to use the TMTP to “buy time” to apply for a Solvency II MA or take MA optimisation action, such as the securitisation of Equity Release Mortgage assets or to avoid the forced sale of ineligible or sub-optimal assets.

It is clear that this is a valid approach for assets held pre-Solvency II given the associated illiquidity premium would have already been captured through the previous regime on the same assets. However, for investment into new asset classes, and when considering further investment in existing MA ineligible asset classes, should it be deemed appropriate that the same transitional benefit could apply for those that do not have sufficient regulatory approvals in place?

Equity Release securitisations

The Standard Formula (SF) treatment of securitisations is extremely penal (essentially a 100% capital charge for long-dated assetsFootnote ⁹ ) and as such for a SF insurer under the Solvency II regime it is very difficult to continue accessing asset pools that require securitisation to be made MA eligible.

In the case of Equity Release, many of the niche annuity providers who have historically been leading players, and therefore have a large proportion of their annuity liabilities backed by the asset class, do not yet have approved internal models.

Without allowing the TMTP to act as a temporary mechanism to allow the continuing investment in the unstructured form of the asset, such firms are unable to continue to write the same volumes of new equity release assets going forwards whilst maintaining solvency coverage.

FRR comparison test implications

The FRR comparison test has a significant impact on management actions that a firm wishes to implement including hedging approaches. The firm will need to consider what impact the TMTP recalculation will have on the overall expected benefit of the action.

While a management action might be the most sensible course of action if considering the Solvency II regime in isolation, factoring in any loss of TMTP may in the short term reduce the desire of firms to carry out such actions and could in some cases potentially counter-intuitively worsen the overall Solvency II balance sheet net of TMTP.

6.1.2. Which metric to optimise?

Firms will need to decide on the appropriate hedging strategy for their balance sheet. The strategy may consider Solvency I and/or Solvency II key metrics. It may be undesirable to continue to hedge to Solvency I Pillar 1, for example due to the existence of the LTICR and the WPICC. Further, there may be unintended consequences of hedging to Solvency I Pillar 1 which firms should carefully think through. A clear hedging policy can facilitate this decision making.

As the TMTP runs off over time, the Solvency II technical provisions, after allowance for the TMTP, become more weighted towards Solvency II technical provisions, prior to allowance for the TMTP. As a result, optimisation strategies that reference a more favourable Solvency I treatment should allow for the run-off of that benefit.

If the aim of the optimisation is to minimise the Solvency II technical provisions after allowance for the TMTP, then on inception of Solvency II firms are still incentivised to optimise the Solvency I Pillar 2 position given Solvency II technical provisions, after allowance for the TMTP, are effectively the Solvency I Pillar 2 technical provisions.

Hedging policies therefore should be very clear on which metric or metrics are being optimised. In general the working party notes that a focused policy is more likely to achieve its objectives than one which attempts to target multiple objectives.

The remainder of this section outlines the metrics which firms may wish to optimise to. In all cases, the relative risks, costs, and benefits of implementation need to be assessed.

6.2.1. Background

For liabilities with sufficiently predictable cash-flows, for example annuity business, both the Solvency I and Solvency II risk-free rates could be adjusted by either the illiquidity premium or the MA. These items are conceptually similar, capturing the excess spread over the risk-free rate expected to be earned on assets “bought to hold” to maturity, but there are some important differences between the two. The working party outlines these key differences in section 3.3, with the potential recalculation options discussed at length in section 5.3.3.

From an ALM perspective, the working party has identified the following considerations for firms with a MA portfolio.

6.2.2. Buying time to apply for the MA

A number of firms did not have approval for the use of a MA at 1 January 2016 when Solvency II came into force. For firms in this situation, it is clear that the use of the TMTP essentially buys time to conclude the MA application process.

It is noted a change in the usage of MA is deemed to be a material change in risk profile which is likely to result in a TMTP recalculation.

6.2.3. Non-use of MA

Firms may have chosen not to apply for the use of MA, where it previously applied an illiquidity premium under Solvency I Pillar 2. In this situation, the firm is likely to minimise technical provisions through managing on a Solvency I Pillar 2 basis and seek to recalculate the TMTP if the optimisation is sufficiently material to trigger a risk profile change.

6.2.4. Differences in Solvency II and Solvency I Pillar 2 hypothecation

Hypothecation is likely to be different under the Solvency II and Solvency I Pillar 2 regimes reflecting the following issues:

• ∙ The MA uses EIOPA Fundamental Spreads while Solvency I Pillar 2 illiquidity premium uses an internal view of downgrade and cost of default risk. Therefore, it is possible some firms will have only allowed for cost of default risk in the illiquidity premium.

• ∙ There could be differences in the cash-flow matching tests used to assess the appropriateness of the hypothecation that impacts on the assets allocated to the Solvency I Pillar 2 illiquidity premium portfolio versus the Solvency II MA portfolio.

• ∙ The MA methodology is more restrictive which could lead to constraints in the calculation. For example, for some assets there is a requirement to allocate these in whole assets to a portfolio, rather than being able to allocate only a proportion of the asset.

• ∙ Differences in asset eligibility. This is outlined below.

6.2.5. Differences in eligibility

Under Solvency II certain assets may not be eligible, or may require transformation before they can be included, e.g. Equity Release Mortgages, where they were eligible under Solvency I Pillar 2.

Some firms may have decided not to restructure ineligible assets and continue to hold them under Solvency II, taking the difference through the TMTP, rather than the MA. However, the working party notes this is likely to lead to more penal capital treatments.

Overall the working party is of the view that the more penal capital treatment is likely to outweigh the use of TMTP, increasingly so as the TMTP benefit runs off over 16 years. Therefore firms are likely to be using the TMTP benefit to “buy time” to dispose of these assets or enact the necessary management action, e.g. an equity release mortgage restructure.

As above, a change in the usage of MA is deemed to be a material change in risk profile which is likely to result in a TMTP recalculation.

6.2.6. Pooled asset allocation

A number of firms have been granted approval for new business written after 1 January 2016 to be included in existing MA portfolios). For firms set up in this way, the additional requirement to cash-flow match pre- and post-1 January 2016 business separately for the purpose of the TMTP is onerous.

One option is to assume the MA and Solvency I Pillar 2 illiquidity premium is the same for business written prior to and after the introduction of Solvency II. This assumes that the MA/illiquidity premium benefit is pooled across these two buckets of business. This pooled matching MA/illiquidity premium benefit is then only applied to policies written prior to the introduction of Solvency II in the TMTP recalculation

The alternative option is for firms to allocate assets separately for business written prior to and after the introduction of Solvency II. This may be consistent with the design of the firms MA and separate MA portfolios have been created for business written prior to and after the introduction of Solvency II, or equivalently a firm may have immaterial volumes of new business following Solvency II. However where firms are not doing this already for the purposes of their MA calculation it is not clear whether sub-allocation within a MA portfolio is required and to what extent the MA cash-flow tests at a sub-MA portfolio level would apply.

6.2.7. PRA cash-flow tests

In order for the PRA to assess that asset and liability cash-flows across all MA portfolios in the industry are sufficiently well matched, it has developed three standardised tests. These tests are designed to monitor consistency across the industry and are not replacements for a firm’s own view of matching.

These PRA cash-flow tests outlined on the PRA website under the following link.

http://www.bankofengland.co.uk/pra/Documents/authorisations/siiapprovals/macashflow.pdf

Implications of cash-flow tests on TMTP

For firms with MA portfolios which include business written following the introduction of Solvency II if TMTP recalculation requires the separation of assets to pre- and post-1 January 2016 buckets and the consideration of cash-flow matching at this sub-MA portfolio level, there are likely to be implications beyond the practical challenges of maintaining the sub-portfolios.

Annuity business written after 1 January 2016 is likely to be younger than business written prior to 1 January 2016 and may also experience greater levels of future improvements in longevity than the business written prior to 1 January 2016. There may also be differences in business mix, which include but are not limited to, the following differences:

• ∙ The mix of individual annuities and deferred annuities;

• ∙ The mix of annuities enhanced following underwriting and “standard” annuities;

• ∙ The mix of annuities which pay level income (or income which increases at a defined percentage) and those which pay out income linked to inflation, e.g. RPI linked or LPI linked annuities.

• ∙ The mix of individual annuities and bulk purchase annuities.

• ∙ Mix differences arising from the different types and associated magnitude of bulk purchase annuity scheme written, e.g. whether the scheme arises from a company associated with “blue-collar” workers or “white-collar” workers.

This means that the liability cash-flows of these two blocks of business are unlikely to have the same profiles including amounts, durations, and convexity. This has the following implications in the context of TMTP recalculation:

• ∙ The sum of the MA benefit derived from the sub-portfolios is unlikely to equal the MA benefit from the portfolio as a whole. Here MA benefit denotes the monetary value of the reduction in BEL arising from the use of a MA.

• ∙ It is possible that no combination of allocating existing assets across the two sub-portfolios, to meets the PRA cash-flow tests for each sub-portfolio, exists. Given this scenario new assets would be required to “fill the gaps.” Trading assets to “fill the gaps” could also lead to a reduction in MA benefit and/or solvency I Pillar 2 illiquidity premium.

The working party is of the view that the costs of monitoring and the theoretical requirement to change asset mix introduced by this interaction of TMTP recalculation and MA requirements would not be beneficial to the industry as a whole.

6.2.8. Re-hypothecate to maximise the TMTP benefit

If a firm has chosen to assign assets separately to the sub-portfolios backing business written before and after Solvency II, the overall aim will be to minimise the post-TMTP value of the Solvency II technical provisions, in addition to hedging other metrics, for example optimising regulatory surplus capital. The working party stresses that any actions taken to “minimise” and/or “optimise” technical provisions should be consistent with a firm’s given level of risk appetite and in addition should comply with all regulatory requirements.

The working party expects that firms are likely to first consider minimising Solvency II technical provisions prior to TMTP, and then maximise the value of the TMTP, given those Solvency II technical provisions. However, the working party also notes the potential opportunity for firms to allocate differently across Solvency I Pillar 2 and Solvency II in order to maximise the TMTP benefit.

Where the asset allocation choice is compatible with sound risk management, the working party is of the view that it is perfectly reasonable for firms to do this. However, the working party believes firms should avoid any asset allocation choice that has been overtly determined to maximise the TMTP, to the detriment of sound risk management.

6.2.9. New assets or new liabilities post Solvency II implementation

Firms need to consider whether new assets can be allocated to business written prior to the introduction of Solvency II, or included in pooled asset allocation across business written prior to and after the introduction of Solvency II.

There will be practical challenges associated with calculating the Solvency I Pillar 2 technical provisions if the company invests in a new asset class. Firms’ basic business acumen and risk management implies that they should have a thorough understanding of the likely return and the risks associated with a new asset class prior to investment. However it may be the case that the existing Solvency I Pillar 2 models do not adequately capture the risks associated with a new asset class. In this case, the firm may need to make a judgement as to what its approach would have been had it owned the asset when the Solvency I Pillar 2 regime was in force.

A suitable Solvency I Pillar 2 methodology for the asset’s spread and an appropriate allowance for defaults and downgrades should be no more onerous than other considerations required for allowing for a new asset class, including Solvency II methodology considerations, operational challenges, modelling challenges, risk challenges, and any required amendments to the firm’s MA application.

However, the firm is likely to be required to demonstrate the suitability of an approach. Therefore, the working party’s view is that any material changes in asset classes, or investment in new asset classes, and their impact, is likely to be required information in the TMTP recalculation application to the Regulator.

Furthermore, one possible ALM strategy might be to select assets not eligible for MA, whose capital requirements are similar to eligible assets but whose spreads are wider, then take MA benefit through the “back door” by means of an increased TMTP. However, this strategy will only work in the short term as the TMTP runs off over time. In addition, there will likely be a resulting capital impact of holding assets ineligible for the MA on the balance sheet.

A firm may acquire new liabilities that have different risk profiles to the business already on the books. Examples of this include a Part VII transfer of business or accepting reinsurance from another life company. This once again raises the question of whether a firm should base its decision on whether to enter this transaction based on the Solvency I treatment of the product, with its implications for the FRR comparison test and TMTP, or the risks as assessed by the Solvency II regime.

As with assets, there are practical issues if the risks associated with new liabilities are not adequately captured by the Solvency I Pillar 2 models. In the case of pre-2016 business that has been transferred from another company, it may be necessary to incorporate parts of the Solvency I Pillar 2 model of the firm from which the business was acquired.

6.2.10. FRR comparison test implications

The FRR comparison test introduces a number of potential implications. Overall the extent to which this test impacts on a firm’s hedging will depend on:

• ∙ Whether the TMTP is based on the unrestricted technical provision differences or on the FRR comparison test,

• ∙ The difference between the unrestricted technical provision differences and the FRR comparison test and hence the likelihood of the FRR comparison test “biting” in future recalculation,

• ∙ Which is the biting Solvency I Pillar (Pillar I or Pillar 2) and the likelihood of the dominant Solvency I Pillar remaining dominant.

The FRR comparison test has a significant impact on which hedging choice or broader management actions that a firm wishes to implement. The firm will need to consider what impact TMTP recalculation will have on the overall expected benefit of the action.

6.3.2. Aims of reset

Where a reset of the TMTP is an important consideration in relation to the solvency position of a firm, it is important that the process for recalculating is the TMTP is transparent, reliable, and not onerous.

In the event of changes in TMTP methodology or a recalculation resulting from a structural change, e.g. a TMTP recalculation as a result of a Part VII transfer, the working party believes a full TMTP application would be required. This reflects the nature of the changes being implemented, which may require a more detailed explanation to the PRA.

However where a TMTP recalculation is a result of changes in operating conditions (including economic conditions), the working party believes that TMTP recalculation should be a dynamic process, based on principles and tolerances agreed upfront between firms and the PRA. The overall aim of this proposal is to minimise the overheads on the PRA and on firms from an excessive number of reset applications, notwithstanding the need to create a process sufficiently robust such that reliance can be placed on this process in lieu of recurring TMTP applications.

In the case of a structural recalculation, e.g. a recalculation as a result of a Part VII transfer, firms should submit applications for recalculation in advance of the structural change occurring. The experience of the working party thus far is that any approvals granted by the regulator are contingent on regulatory approval for the structural change taking place within a certain timeframe. This flexibility from the regulator is helpful for firm planning, although firms should be prepared to submit any application to the regulator well in advance of the structural change taking place. The approval for these applications is up to 6 months, in line with prior TMTP approvals.

6.3.3. Principles of a dynamically recalculated TMTP

Under the working party’s proposal for a more dynamic TMTP, key principles of the dynamic recalculation should be set in advance and agreed both within a firm and with the PRA. The working party envisages that these details should be included in a TMTP recalculation application and should leverage off a firm’s existing recalculation policy. Many of the principles which would be outlined in this application process would be similar to those outlined in the firm’s recalculation and this is explored further in section 4.3 of this paper.

The table below summarises the key principles the working party believes a firm should outline in advance. These principles are also expanded upon further below in this section.

6.3.4. Which material risk profile changes?

Firms should outline which types of material risk profile changes are in scope of their dynamic TMTP recalculation and which material risk profile changes are not in scope of their dynamic TMTP recalculation.

As outlined above, the working party would view recalculation as a result of operating changes, including changes in economic conditions, to be in scope of a dynamic TMTP recalculation and view more structural changes as out of scope of a dynamic TMTP recalculation.

6.3.5. When to recalculate the TMTP?

Firms should outline their model for a dynamically recalculated TMTP, i.e. TMTP which is fully dynamic, a TMTP that is recalculated every year-end or a TMTP which is recalculated as a result of a specific tolerance being breached.

The working party believes that too frequent recalculation of the TMTP would be onerous for firms and make it harder to track what the value of the TMTP is at any given point in time. Indeed the necessary data required for a robust TMTP recalculation may not be available on a daily, monthly or quarterly basis.

Although not fully dynamic, the working party believes a process where the TMTP is recalculated quarterly in arrears, subject to a specific tolerance being met, could be the most beneficial model balancing the practical considerations and avoiding the need for a parallel version of the TMTP for solvency monitoring and dividend planning.

The working party also notes the specific recalculation tolerance should ideally be similar in materiality to the 5% change in pro-forma solvency coverage outlined in the PRA supervisory statement on TMTP recalculation.

6.3.6. How to recalculate a dynamically recalculated TMTP

The calculation of the TMTP is very complex and involves the calculation of the Solvency I and Solvency II technical provisions and FRR for business written prior to the introduction of Solvency II. Therefore the working party believes a TMTP recalculation which is available within quarterly reporting timelines would be onerous to achieve for many firms, in addition to the effort required to complete other aspects of quarterly reporting.

The working party believes a model of recalculating the TMTP in arrears of a private disclosure, or only at certain times of the year to support public disclosures would help mitigate this. The approach taken should be outlined in the application and a firm’s TMTP recalculation policy.

6.3.7. Governance for TMTP recalculation

A firm should outline the governance process it follows to ensure that the dynamically recalculated TMTP is robustly derived. This process should outline:

• ∙ Who is responsible for the calculation of the TMTP;

• ∙ How the TMTP calculation is checked and reviewed;

• ∙ What documentation of the calculation will be produced, both in terms of the recalculated value and how the firm complies with the principles outlined in advance;

• ∙ Whether the calculation will be audited externally;

• ∙ Who will ultimately sign-off the updated TMTP;

• ∙ The role of the firm’s audit committee, and;

• ∙ Agreement with the PRA on what and when changes in TMTP will be communicated by the firm.

6.3.8. Review of effectiveness of dynamic TMTP calculation

A firm should regularly review the effectiveness of its dynamically recalculated TMTP, to ensure it remains fit for purpose and to consider whether it wishes to make amendments in light of experience. These amendments are likely to result in a new TMTP application to the PRA.

6.4.1. Overview

As we finalise this paper, we are now in the midst of the 2016 year-end, which marks an important milestone to insurers and how they communicate the impact of the TMTP to stakeholders including the analysts and investor community.

For year-end 2015, the majority of firms disclosed the value of their Solvency II surplus and coverage ratio including the impact of TMTP, but did not disclose the quantum of the TMTP in these announcements. Furthermore, firms’ Solvency II surplus sensitivities, and in particular those sensitivities to interest rate movements, included the assumption that the TMTP would be recalculated to mitigate the change in risk margin.

For year-end 2016, the impact of TMTP on the Solvency II technical provisions, own funds and the SCR will be disclosed publicly as part of the QRTs and the Solvency and Financial Condition Report (SFCR). It is expected that firms will seek to pre-empt these public disclosures and may wish to mitigate the risk of misinformation to the investor community by including information on the impact of the TMTP in their year-end 2016 external market announcements.

As Sam Woods described in his speech in July 2015: It is important that analysts and investors understand that Solvency II adjustments and transitional measures are a legitimate part of the regime. Transitional measures do not distort the solvency reality: they are designed to ensure a smooth transition to the new regime, avoiding disruptions in the market and allowing a certain period for companies to fully recognise the impact on old books of contracts that have been underwritten in a different regulatory framework.

In the remainder of this section, we outline the key issues which the TMTP working party believes firms should emphasise and the investor community should consider.

6.4.2. Impact of recalculation

The recalculation of the TMTP is important in ensuring firms ensure their accounts represent a “true and fair” position. It is important to not overstate or understate the benefit of the TMTP. As businesses and operating conditions evolve, so does the theoretical TMTP, and maintaining a static TMTP will no longer be appropriate.

A key example would be that firms which recalculated the TMTP in the summer of 2016, to account for the low interest rate environment, will likely have a larger TMTP than at the point of Solvency II inception. As interest rates increase, this TMTP is no longer representative of the genuine move from Solvency I to Solvency II. Without recalculation, this could lead to an overstatement of the Solvency position and a lack of connection with the amount of dividend which could theoretically be paid out. In the example above, as interest rates rise, the risk margin would fall, but without the recalculation, the TMTP would be larger than the risk margin and not run-off in the same way.

6.4.3. Recalculation strategy and policy

A firm should ensure that its TMTP strategy and/or policy, including recalculations, is adequately reflected in its external disclosures, which include those made under Pillar 3 requirements, Report and Accounts and less formal investor presentations. Clear TMTP communications are particularly important as the PRA has stated that the TMTP can be treated, in effect, as Tier 1 equity (as stated in Sam Woods speech) on a firm’s Solvency II balance sheet and thus it supports not only the solvency position of a firm, but also the payment of dividends to investors.

6.4.4. Transparent and consistent disclosures

Consistent, transparent, and clear TMTP disclosures will support investor confidence across the industry and would, therefore, be of benefit to all firms holding a TMTP on their Solvency II balance sheet. Given the importance of these external communications, it seems appropriate, although not compulsory, for a firm to reflect its communication strategy in its recalculation policy. It is also advisable that firms should also review industry-wide disclosures so that they can develop their own best practice, where appropriate. We would expect this approach would lead to a convergence of practice to develop over reasonably short space of time.

It is uncertain how analysts will react to TMTP and in particular the quantum of TMTP on insurers’ balance sheets. Managing risk of misinformation will be the key. We expect to see more disclosures from firms around the TMTP in YE16 as firms look to pre-empt publically available SFCR and QRTs. There are advantages and disadvantages to being first to move in public disclosures.

Disclosures may vary between parents and subsidiaries, but ultimately they should be of a sufficient quality to provide investors with information they require. In particular, it is important that firms provide the appropriate level of information in group disclosures.

The regulator has provided a number of recent communications on expectations for recalculation of the TMTP, namely SS6/16 and CP47/16. These communications suggest that a dynamic TMTP is not a feature of recalculation. In the latter communication, the PRA has suggested that the Audit Committee should form an independent view on whether or not the firm meets conditions 1–3 and 5 set out in regulation 54. This correctly places reliance on the firm and their Audit Committees to ensure the calculations are reasonable. Accounts are expected to be a true and fair representation of the position of the firm. An Audit Committee may view the “correct” value of the TMTP at the reporting date to be materially different from the value derived at the last recalculation date, less any run-off. The view of the “correct” value will depend on the risk appetite of the firm.

In this situation, the working party believes that it is reasonable for firms to have the discretion to “write down” the value of the TMTP where they believe it to be overstated relative to the value derived at the last recalculation date, less any run-off. It is not clear which methods will be permissible to communicate the TMTP difference between the “correct” value and the approved value. Possible options include a direct write down of the TMTP, possibly through an increase to the rate of run-off of the TMTP, or an explicit accounting or technical provision. In the event that a firm believes the “correct” value is an increase to the last approved recalculation, it is important for firms to not overstate the TMTP. It is considered unlikely the results could be adjusted upwards, but more appropriately managed through the messaging of the results in the notes to the accounts for public disclosures.

Firms could consider the appropriateness of clearly stating within their external disclosures:

1. 1. Key aspects of the firm’s TMTP recalculation policy. For example, the firm may wish to disclose its view on whether TMTP should be fully dynamic;

2. 2. Whether a TMTP has been allowed for, and the level, on its Solvency II balance sheet;

3. 3. When the TMTP was last recalculated and when the next regular review is expected;

4. 4. The expected future run-off period of the TMTP and its expected annual run-off and whether this run-off is applied (or disclosed) in the year-end balance sheet. Firms may wish to compare the expected future run-off of the TMTP with the expected future run-off of the risk margin;

5. 5. The level of TMTP recalculation allowed for in any sensitivity results disclosed and whether the recalculations are in line with approach outlined in their internal policy;

6. 6. How their dividend plans interact with, and how reliant they are on, the TMTP;

7. 7. Whether they consider the TMTP run-off will have any significant impact on future dividend streams;

8. 8. Other impacts on the balance sheet from TMTP (e.g. deferred tax);

9. 9. Whether solvency coverage requirements are met without the TMTP and if not the need for a phasing in plan;

10. 10. Any voluntary write down of the TMTP due to a material change in conditions since last approval to recalculate; and

11. 11. Notes to the accounts explaining the Audit Committee’s true view of the TMTP, in the event this is different to that published.