Primary tabs

Allianz SE

Ulrike Stinglwagner
We understand that the question on the KID itself would only be “What are the risks and what could I get in return?” as stated in Art. 8 (3) (d) of the PRIIPs Regulation. The “Key Questions” listed in section 3.3.1 of the CP therefore should only serve for purposes of pre-clarification. Materially, it is important that the presentation of risk and return characteristics of the product is adequately balanced and includes a sufficiently clear positioning of a product with respect to its risk/return trade-off. This means that neither risks nor rewards should be inadequately stressed, be it regarding the indicators or in the wording.
Generally, the proposed wordings could be adequate starting points. The major challenge for the disclosure is, however, that the product types covered by the PRIIPs Regulation are very diverse in nature (e.g. a short-term derivative contract vs. certain long-term unit-linked life in-surance contracts). While these products share some common characteristics (which trigger application of the PRIIPs Regulation), they are designed to meet very different customer needs and therefore are not considered direct substitutes by the customers.

For example, market risk is the predominant risk for insurance-based investment products, since they are expressly designed for longer time horizons and marketed as such. Credit risks often are less relevant (especially, when guarantees by the insurer are involved, which are already adequately addressed by prudential regulation, i.e. Solvency II, and insolvency guarantee schemes). In addition, the limited liquidity of an insurance-based investment product may be an explicit feature rather than a risk which supports the long-term investment orientation of the customer.
The different product types covered by the scope of the PRIIPs regulation make it nearly im-possible to derive adequate unified risk measures. This calls for a differentiation of KID tem-plates for different PRIIPs types without impairing the goal of general comparability. In addition, several different measures are appropriate to measure the risks for the different PRIIPs types.

Several examples can illustrate these points, the risks for unit-linked or index-linked insurance-based investment products in many instances can be modelled very similar to UCITS funds. By contrasts, more “traditional” insurance-based investment products may contain substantial guarantee and/or profit participation components which do not lend themselves to such measures. In addition, existing secondary markets for some products, assets or indicators may permit the calculation of certain measures for some products but not for others, e.g. because they are issued and backed by financial intermediaries. Furthermore, the underlying customer objective should be adequately reflected: for products which are expressly designed and mar-keted for longer time horizons, liquidity aspects should not be presented too prominently.

For modelling the market risk/reward measures, meaningful historical measures are not avail-able or may not be adequate for many products, namely many insurance-based investment products. This invalidates the general suitability of the UCITS KIID indicator. Prospective measures seem more adequate to derive risk classes. The Expected Loss for a given Value-at-Risk seems the most appropriate measure among those proposed, a similar but even better choice may be CTE (conditional tail expectation). Both measures capture the expected loss under a worst-case development, which is highly relevant as well as comparatively easy to explain to the customer. The specific risk/reward classification should be performed using a probabilistic modelling approach based on a sufficiently large set of economic scenarios (which do not have to be displayed to the customer). This can ensure the comparability of different measures.

Overall, it seems that the most adequate approach would be to differentiate specific measures, presentation requirements and templates by PRIIP type under a common framework for presentation to promote general comparability.
Any participation, reimbursements or attribution of costs to the customers should be deducted from costs. Typical (but as yet unknown) costs, should be estimated and included in the cost indicator. By contrast, truly contingent costs, e.g. which depend on decisions of customers, should be disclosed separately and not be integrated in the cumulative cost indicator.
It is essential that retail investors understand the performance scenarios properly. Deterministic modelling with several different assumed returns seems most adequate to illustrate the possible pay-outs. Historic data often may be misleading or not even be available for many products (especially those with new structures and/or long time horizons). Therefore forward-looking scenarios seem more adequate. If no clear reference value is available, it is advisable to choose an even number of scenarios to avoid misinterpretation of the scenario in the middle as the most probable one.

The risk/return classification should be performed based on a probabilistic modelling approach (see Q4).
We understand that it is important to achieve the most important objectives of the PRIIPs KID, in particular to provide the customer with adequate information and to unify presentation across products and EU countries to ensure comparability and a level playing field. In many instances, these goals may be at least in partial conflict with each other, which requires careful balancing. For example, the products under the Regulation’s scope have substantially different dura-tions/maturities and aim at fundamentally different customer needs.Overall, the most adequate approach would be to differentiate specific measures, presentation requirements and templates by PRIIP type under a common framework for presentation to permit general comparability. In particular, different measures and presentations should be considered (and formally declared) equivalent based on the type of PRIIP. While this may slightly limit the comparability of certain PRIIP types, the alternative of forcing one methodolo-gy and presentation onto every product (“one-size-fits-all”) would not provide the customer with adequate information and therefore could unintentionally mislead customers. It would be helpful to receive guidance about the different PRIIP types and adequate measures and presentation.
The products under the Regulation’s scope have substantially different durations/maturities.

It is therefore necessary to set a timeframe in line with the features of the product. A strictly unified approach for all products would not be adequate and could possibly confuse and mis-lead the customer and therefore should be avoided.

For example, for insurance-based investment products a comparatively long timeframe (possi-bly up to several decades) would be typical while for many derivative contracts maturities rare-ly exceed a few months.
Yes, performance scenarios should contain absolute amounts.
The presentation in the performance section should be consistent with the presentation on costs (in the cost section of the KID) as well as with the risk classes.
It is essential that the retail investor gets a clear understanding about the risk/reward trade-off of the product. Therefore, the adequate number of scenarios as well as the range between those scenarios depends on the risk/reward characteristic of the product: riskier products re-quire more scenarios with a wider spread to permit an adequate presentation. In particular, the number and range of scenarios should also be differentiated by PRIIP type (see also Q7).
Since risk and reward are typically strongly correlated, a unifying presentation of both aspects is adequate. In addition, it is most important that the retail investor clearly understands the trade-off between risk and reward.

Care should be taken that the aggregation of single risk indicators (in particular for market, credit and liquidity risk) into a summary risk indicator should not mislead customers. For example, the long-term nature of many insurance-based investment products is an explicitly marketed feature of the product and should not be misinterpreted as liquidity risk. The most adequate response would be the differentiation of measures and presentation for different PRIIP types under a unifying concept (see also Q7 and Q55).
Risk and reward scenarios are typically strongly correlated, therefore a unifying presentation of the features is needed (see also Q12 and Q14).
Performance and risk/reward indicators should be consistent and not presented separately. Specifically, the performance presentation could entail several scenarios (in annualized per-centage and EUR terms) regarding gross performance (before costs), costs (based on reduc-tion in yield measure) and net performance (after cost), see also Q20. The risk indicator and narratives should be presented in close proximity to these performance scenarios in the KID.
We understand that the question posed in the KID would only be the question in Art. 8(3)(f) PRIIPs Regulation: “What are the costs?”, therefore the “Key questions” in Table 10 of the CP merely act to help in the clarification of the discussion (see similar point on risk in Q2).

In addition to the questions listed, the net return/performance (after cost) should also be con-sidered (see also Q14).
It is important that the information on costs is consistent with the presentation of the perfor-mance scenarios.

In addition, the premium for biometric (and other) risks should not be considered a cost (as proposed on page 54 and 55 of the CP), since the customer receives additional benefits for these premiums, which may be substantial.
-
Cost structures vary substantially, especially across different PRIIPs types (e.g. derivatives, UCITS, insurance-based investment products, etc.). Even more important than a consistent presentation across different product categories is the consistent presentation for each single product (PRIIPs category or type).A specific challenge for insurance-based investment products is that they often contain premi-ums for biometric risk transfer, which in no case should be accounted for as costs (as is pro-posed on page 54 (Table 11) and page 55 of the CP), since the customer receives additional benefits for these premiums which may take various forms (benefits in case of death, disabil-ity, longevity etc.). In any case, the KID should provide the possibility to present relevant bene-fits from biometric risk transfer.

In addition, care has to be taken when considering certain potential cost components, in par-ticular for insurance-based investment products:
- surcharges according to method of payment (e.g. monthly instead of annual): these sur-charges should be conceptualized as an interest payment on a deferred premium payment and therefore should not be included as cost
- early redemption fees should only be treated as costs when they accrue to the insurer and not (as typically the case and in line with actuarial principles) directly benefit the community of policyholders (i.e. the remaining insureds)
- look-through costs can only be disclosed to the extent they are known to the insurer, but not to the extent the underlying vehicle (e.g. fund) itself does not have to disclose them.
Insurance-based investment products are not traded, thus implied market estimates of costs for the products are not available. In addition, look-through costs for some assets held may often be not available for the PRIIP manufacturer.
Yes. In addition, the RIY disclosure should be accompanied by a disclosure of gross average annual return (i.e. before cost) and net average annual return (i.e. after cost). See also Q14.
-
Yes, it is important that performance and cost scenarios are consistent. Costs should therefore be presented for all (relevant) scenarios. If additional information is relevant to understand costs, those should be explained. See also Q14 and Q20.
-
There are different challenges to a standardised cost disclosure across different PRIIPs, in particular
- some PRIIPs may contain specific components which do not lend themselves to easy com-parison with investment components, e.g. premiums for biometric risk covers of insurance-based investment products, which also provide additional benefits (in case of death, disability, etc.), see also Q17 and Q20 or early redemption fees which are accrue to the community of the policyholders instead of the PRIIPs manufacturer (see Q16).
- some cost components may be subject to different disclosure requirements for different products (e.g. exemption of UCITS funds from disclosure of transaction costs)
- some cost components are difficult to compare depending on the contract structure or rea-sonable assumed duration (e.g. calculation of RIY measure requires assumption of a contract duration) which may vary substantially by product.

Even more important than a consistent presentation across different product categories is the consistent presentation for each single product (PRIIPs category or type) in order to avoid any inadequate bias on the customer decision based on the KID.
Performance and risk/reward indicators should be consistent and not presented separately. Specifically, the performance presentation could entail several scenarios (in annualized per-centage and EUR terms) regarding gross performance (before costs), costs (based on reduc-tion in yield measure) and net performance (after cost), see also Q20. The risk indicator and narratives should be presented close to these performance scenarios. In addition, care has to be taken to treat certain specific aspects adequately (e.g. biometric risks in case of insurance-based investment products should not be treated as costs), for details see Q16.
It is important, that any cost disclosure provides the retail investor with relevant information for his or her decision making. Contrary to this intent, a very detailed cost breakdown may not lead to relevant clarification but to more confusion and misguided decisions. For insurance-based investment products we consider the comprehensive reduction in yield (RIY) as the most suitable starting point for cost disclosure. This comprehensive measure should be com-plemented by separate disclosures of contingent costs where appropriate (see Q28).
It does not seem adequate to fully integrate market, credit and liquidity risk indicators into one summary risk indicator. In particular, not all risk dimensions are equally relevant for all PRIIPs types (see also Q3 and Q4). Customers should not be misled by an inappropriate integration of risk indicators. In consequence, the most adequate approach would be to differentiate specific measures, presentation requirements and templates by PRIIP type under a common framework for presentation to permit general comparability.

For example, for insurance-based investment products, the market risk component could ade-quately be measured and classified by a summary risk indicator and supplemented by a narra-tive explanation. Credit and liquidity risks should be presented using suitable narrative expla-nations, except for those cases where a clear measure would be available (i.e. some unit-linked products). In any case, early cancellation / surrender should not be treated as a liquidity issue but as an unintended and unforeseen incident to be covered under the KID explicit early cancellation question (i.e. “How long should I hold it and can I take money out early”). See also Q40.
Cumulative costs should be shown as annualized percentage reduction in yield (RIY) over the duration of the contract. This measure is comprehensive, stable, robust, consistent and com-parable among different PRIIPs and comprehensible for most retail investors.
The premium for biometric (and other) risks should not be considered a cost (as proposed on page 54 and 55 of the CP), since the customer receives additional benefits for these premi-ums, which may be substantial.
-
It would be helpful to have a further clarification on the products to which the criteria of the comprehension alert would apply. A too broad application risks losing the cautionary intent of the alert. Specifically, for insurance-based investment products, the comprehension alert should only apply to those products, where the retail investor bears a substantial (investment) risk, typically certain unit-linked investment products.
A classification according to legal type is an adequate starting point.
No.
The description should be kept very brief, yet be sufficiently clear to describe the general ob-jective of the product. This may be important, when very specific objectives should be achieved.
No.
The definition of customer type should to give the customer broad orientation regarding the potential fit with its demands and needs, however, the customer type can and should not sub-stitute necessary personal advice by a distributor (currently under discussion under IMD2/IDD). In addition, the customer type does not meet the definition of “target market” with respect to the rules regarding the ESA JC principles regarding Product Oversight and Governance (POG), which addresses product manufacturers and distributors, but not necessarily end customers
For insurance-based investment products, insurance benefits typically form a substantial and integral feature of the product. Depending on the exact scope of the coverage, they provide substantial risk transfer and benefits regarding biometric risks (benefits for death, disability, longevity, etc.) and market risks. It is therefore essential, that these benefits can be presented with adequate prominence and not just as costs (see in particular Q16 and Q17). A lack in the ability to adequately present these benefits could create a substantial distortion against the overall objective of a fair unbiased presentation of PRIIPs product performance.
No. In particular, insurance-based investment products typically have a specified duration (which may be life-long).
No.
No. For insurance-based investment products, the retail investor should be informed about the surrender value for different holding periods. Also, certain key indicators or ratios (e.g. surren-der value / cumulated contributions) can be presented in order to comply with the space limita-tions of the KID.
No.
Yes.
Yes.
We agree that the article would mainly cover unit-linked and hybrid life-insurance products.
-
There does not seem to be a simple solution which addresses all issues raised. In any case, it is important that any modification of the rules does not systematically bias the PRIIPs disclosure regime for or against a certain PRIIP type or category.
-
If the underlying investment itself is no PRIIP, a full “look-through” disclosure may be impossi-ble. For example, the “ongoing charges” disclosure requirement for UCITS funds does not include transaction costs. It should be made clear, that in this case the disclosure requirement for the PRIIPs manufacturer is limited to the disclosure requirement of the underlying fund.
The KID should be reviewed and updated regularly. However, since the KID constitutes pre-contractual information and many products (such as insurance-based investment products) are not publicly traded, KID updates should only be necessary for products which are still mar-keted. Any information for in-force business should be governed by ongoing reporting re-quirements (e.g. pursuant to IMD2/IDD for insurance-based investment products).
-
-
No.
The KID is pre-contractual information (i.e. has to be provided before the conclusion of the contract). No additional requirements regarding timing seem necessary.
-
Yes. We understand that it is important to achieve the most important objectives of the PRIIPs KID, in particular to provide the customer with adequate information and to unify presentation across products and EU countries to ensure comparability and a level playing field. In many instances, these goals may be at least in partial conflict with each other, which requires careful balancing. For example, the products under the Regulation’s scope have substantially different durations/maturities and aim at fundamentally different customer needs.

Overall, the most adequate approach would be to differentiate specific indicators, presentation requirements and templates by PRIIP type under a common framework for presentation to permit general comparability. In particular, different modelling approaches and presentations should be considered (and formally declared) equivalent based on the type of PRIIP. In addition, to the extent permissible the content should be limited to relevant information for the PRIIPs type (i.e. no disclosure about insurance benefits in structured deposits). While this may slightly limit the comparability of certain very different PRIIP types, the alternative of forcing one methodology and presentation onto every product (“one-size-fits-all”) would not provide the customer with adequate information and therefore could unintentionally mislead customers. Separate templates by PRIIP type could therefore be very helpful to provide guidance about adequate presentation.
Generally, different PRIIPs product types should be reflected in different specific disclosure requirements, including modified templates (see also Q55).

For some products (e.g. insurance-based investment products) single (lump sum) and regular payment schemes generally exist with similar importance (which also varies by Member State). Therefore, in particular the KID template for insurance-based investment products should permit appropriate presentations of both alternatives.
-
-
In the KID costs should be calculated on an ex-ante basis. Depending on the degree of per-sonalization of the KID, the calculations should be made on reasonable assumptions.
+49 89 3800 2576