Overall, the key questions presented in the discussion paper appear to address the basic information needs for retail investors. Nonetheless, we would like to submit two suggestions:
• As regards the uncertainty of returns as a dimension of risk, we believe that it is also relevant for retail investors to receive an answer to the question how long it can take that the request for payout will be fulfilled. The market experience shows that some products, especially insurance contracts, where early withdrawal also entails additional costs tend to retain investors’ money for a considerable period of time, sometimes several months. This is certainly an important aspect of risk, especially given the fact that early cashing in is generally prompted by urgent liquidity needs not expected by the investor at the time of concluding the contract. Therefore, it should be duly taken into account for the purpose of depicting risks at the point of sale.
In relation to the table 3, this question should fit into the last line as it is linked to the recommended holding period and early cashing-in of the contract.
• The questions “How much can I win?” and “How much am I likely to win?” in table 3 evoke images of gambling. In order to provide for a more neutral presentation, the term “get” should be used. (i.e. “How much am I likely to get?”)
In general we agree with the identification of market, credit and liquidity risk as the main risk categories for PRIIPs.
Regarding the concept of these risks, we see certain problems for investment funds. The notion of market risk encompassing credit risk of the underlying assets of a PRIIP may be of value for streamlining the risk concepts for the purpose of retail-digestible information. However, it must be noted that such approach to market risk and in consequence also to credit risk displays considerable differences from the notion applied for the purpose of risk management/calculation of investment limits for investment funds and most importantly, described in the fund prospectus in accordance with the UCITS Directive. In any event, it must be avoided that investors obtain contradicting information on risks relating to an investment in the fund prospectus and the PRIIPs KIID. Therefore, it is important to work towards consistency of risk disclosure under different EU frameworks.
If available, historical data should be used for the purpose of risk measurement. Modelled data are less reliable and robust because they can easily be manipulated or at least influenced by manufacturers to provide an overly positive picture of a product’s risks. If historical data cannot be used (e.g. in case of new products or a significant change in the investment policy), modelling should be allowed provided the existence of stringent methodical standards and proper supervision by the authorities.
As regards measuring of market risk, we would suggest using the SRRI methodology based on historical volatility. Volatility of past returns has proven a reliable measure in the UCITS world and many investors are already familiar with it. The methodology is engaging from the investors’ perspective because the SRRI assesses the most relevant aspect in respect of the market risk: the volatility of investments. The results are fairly comparable and ensure discriminatory output for different investment strategies. Finally, it should be emphasized that the supervisory authorities are familiar with the SRRI. This should facilitate proper supervision in respect of the presentation of market risks.
In case of new PRIIPs or products with insufficient performance history, (partial) modelling based on consistent methodical standards should be allowed in order to calculate the distribution of possible returns. This approach would be also in line with the current practice under the UCITS Directive.
According to our assessment, proper approach to an understandable and comparable indicator of credit risk should be based on qualitative measures. Fixed clear qualitative criteria could lead to a reliable, robust and stable indication of credit risk. In these terms, we could envisage a scale ranging from products basically shielded from the relevant obligor’s insolvency (such as UCITS or many retail AIFs) over those protected by a deposit guarantee or investor protection scheme or offered by institutions subject to prudential supervision to products fully submitted to the obligor’s credit risk.
With regard to the liquidity risk, the key aspect from the viewpoint of retail investors relates to the possibility and timeframe of disinvesting.
In this light, liquidity risk should be assessed based on the specific characteristics of the exit arrangements (qualitative measures). Only after establishing the applicable exit arrangements, appropriate quantitative measures as suggested in section 3.4.3. can be applied. For instance, the absence of a secondary market for a PRIIP (aspects (i) and (ii) of the exit arrangements) does not automatically result in a high liquidity risk. Most investment funds offer their investors daily redemptions opportunities at a price calculated on the basis of the fund’s NAV. Such redemption opportunities should qualify as “organised liquidity facilities” (aspect (iii) of the exit arrangements), but cannot be assessed on the basis of the proposed quantitative criteria which reflect specificities of secondary market trading.
Thus, different quantitative criteria should apply in case of products not traded on secondary markets in the first place, in particular investment funds and insurance contracts. In relation to those, liquidity risk should be primarily measured against (1) the time it takes to disinvest and (2) the price of such disinvestment (e.g. exit charges or costs of cashing in early). As regards the relevant timeframe, a scale based on exact periods could be stipulated in order to allow for proper comparability of liquidity risk (e.g. (i) one trading day, (ii) one month, (iii) half a year, (iv) a year and (v) more than a year).
It is difficult to include contingent costs such as performance fees in the aggregated cost disclosure. As a performance fee is conditional upon the occurrence of certain triggering events, it may or may not be charged in the years to come. Therefore, it could be misleading for investors if the overall charges disclosure purported to illustrate the possible amount of the performance fee. In our view, performance fee could possibly play into the presentation of overall costs only if such costs were calculated on the basis of performance scenarios and accounted for performance events which trigger the performance fee (cf. our suggestions relating to Q22 and 24 above).
The aim should be to provide investors with performance scenarios that are based on realistic, reliable and robust methodologies. Performance scenarios should, in particular, be engaging, understandable and comparable. They should also be carefully tested with consumers to ensure that the presentation is intuitive to retail investors.
Performance scenarios including or based on probabilistic modelling are presumably more engaging for retail investors because they enable investors to assess which returns they can expect. In contrast, performance scenarios that show possible outcomes without any implications as to their likelihood can only provide for a general understanding of the structure of a PRIIP, but do not transmit any information relevant to assess profitability and risks of this investment. Therefore, we have a preference for performance scenarios that include or are based on probabilistic modelling. However, probabilistic scenarios should only be favoured if the underlying methodologies and/or assumptions can be standardised with a sufficient degree of confidence in order to ensure comparability of information provided in the PRIIPs KID.
With regard to the relevant methodologies, the most critical aspect should be (the access to) data. Historical data provide a reliable and robust basis for calculations. Data produced by modelling of market instruments (which underlie a product’s performance) can easily be manipulated or at least used for overly positive presentations of PRIIPs. The competent authorities will hardly have the capacities to ensure proper supervision of data produced by modelling of market instruments. As a consequence, use of historical data should be preferred if it matches with the underlying methodology for performance scenarios.
Where historical data are not available or deemed not representative for future outcomes, performance scenarios should be based a representative portfolio model, target asset mix or a benchmark in case there is a valid benchmark for a PRIIP. This would ensure consistency with the approach used for the SRRI calculation under the UCITS Directive (6).
(6) For details, cf. CESR’s guidelines on the methodology for the calculation of the synthetic risk and reward indicator in the Key Investor Information Document from 1 July 2010 (CESR/10-673).
We recommend that historic data are used for performance scenarios provided that this approach is consistent with the eventual calculation methodology. As set out in our reply to Q6 above, data produced by modelling of market instruments (which underlie a product’s performance) can easily be manipulated or at least used for overly positive presentations of PRIIPs. Since the future is unpredictable and a modelling approach in some way tries to foresee future developments, such an approach could result in inconsistent KIDs because each manufacturer would produce different results when applying a modelling approach.
In case historical data are not available or deemed not representative for future outcomes, modelling in line with the blueprint used for the SRRI calculation by UCITS should be allowed. In this regard, we recommend specifying the details of modelling in order to reduce the risk of diverging outcomes which would impair the comparability of the PRIIPs KIDs.
Timeframes for performance scenarios should be linked to the cost disclosure. Based on our assessment of the PRIIPs market, we recommend accounting for performance results over 1 year, 3 years, 5 years, 10 years and the product lifetime/recommended holding period if longer. In case this differentiation is deemed too complex, at least investment periods of 1 year, 5 years and 10 years (or the product lifetime/recommended holding period) should be covered.
In terms of presentation, we have a preference for a graph covering a rolling period similar to the example on the left top of page 43 (cf. our answer to Q13 below). This would necessarily account for the mentioned timeframes and indeed present the expected performance over time, thus avoiding an undue focus on particular time points.
Because of the fact that people make perceptual mistakes in interpreting percentages and that consumers are prone to a ‘small numbers bias’ which makes them underestimate the actual impact of small percentages on returns, we recommend using monetary figures. The figures could be marked as an example for an investment. €5,000 or €10,000 could be a reasonable amount for such an example. If the minimum investment is higher, the minimum investment should be used. If the maximum investment is lower, the maximum investment should be used. Since monetary figures can only be provided on the basis of an example, percentage figures should be provided, too. This would also ensure comparability between different PRIIPs. Providing both monetary and percentage figures should require only little additional space which is well spent as the information on performance is key for consumers. At best, monetary and percentage figures should be presented in combination in the suggested graph (cf. our reply to Q13 below).
We are convinced that performance scenarios need to be presented net of costs in line with the performance presentation standards applicable to UCITS. Gross scenarios are useless for investors as they visualise hypothetical outcomes which cannot be achieved in a product. Investors must be able to easily grasp the PRIIP’s yield prospects without being forced to make their own calculations by referring to other KID sections. Moreover, gross scenarios might induce investors to underestimate the effect of costs on the performance. For instance, certain products with surrender values falling below the invested capital in the first years will look more appealing if performance presentation disregards the effect of costs.
In these terms, however, it should be noted that it is not possible to make accurate predictions of future costs for many PRIIPs, including investment funds. Product costs depend on many factors which cannot be established in advance, e.g. fund volume, performance, trading activities etc. Hence, in line with the recommended use of historical data for the purpose of performance presentation we are of the view that netting of costs should be conducted on ex-post basis.
As mentioned above, we prefer a number of scenarios to be presented within one graph in line with the example on the left top of page 43 of the discussion paper. In general, we believe that three scenarios displaying positive, negative and neutral development of an investment in line with probabilistic modelling if feasible should be included.
In this context, we deem it very unfortunate that the PRIIPs Regulation does not provide for presentation of past performance. Even though not reliable in terms of future forecasts, past performance data provide useful indications for investors and enhance comparability with other products. Therefore, we would like to encourage the ESAs to assess whether past performance of a PRIIP could be shown in addition to the future performance scenarios, at least for non-structured PRIIPs following market strategies.
As explained in our reply to Q5 above, we believe that all three types of risk (market, credit and liquidity risk) should be visualised by means of separate indicators. Based on the experience with the UCITS SRRI, we definitely favour simple scales like those presented in the example at the bottom of page 39. In this regard, we would rather abandon the overall risk indicator and stick to the separate illustration of different risks on a relevant scale (please see graphic in our summary response attached).
These separate visual elements appear engaging for investors because all relevant risks are shown in an understandable manner. They also enhance the comparability of different PRIIPs because all risk types can easily be assessed when investors compare different PRIIPs. Since some risks are more relevant than others for certain PRIIPs, showing indicators for all risks ensures a balanced presentation which covers all types of PRIIPs.
Nonetheless, should the consumer tests expose a clear preference for a combined indicator, we would advocate supplementing the presentation by an illustration of the overall risk in line with the example at the bottom of page 39 whereby all risk elements (market, credit and liquidity risk) should be weighted equally (please see graphic in our summary response attached).
If that were deemed too complex, we could ultimately envisage a single risk indicator based on equal weightings of risks as presented at the top or in the middle of page 37. This way of presentation is known from the UCITS KIID and a proven and tested means to provide information on risks. Many investors should be already familiar with this presentation style. Investors and the industry have made consistently positive experiences with this indicator.
The illustration of risk should be supplemented by a visual element depicting the relationship between risk and return such as the arrows shown on the top of page 37. It is key for investors to understand that typically lower risk correlates with lower rewards and vice versa. Hence, clarification of this correlation should form part of the “risk-reward profile” description of a PRIIP as required in Article 8(3)(d) of the Level 1 Regulation. The visualisation by arrows as shown on the top of page 37 is part of the UCITS KIID and thus well-known and tested with EU investors (please see graphic in our summary response attached).
The risk indicators presented on page 38 and on the top of page 39 have the practical disadvantage that their comprehensibility will be impaired if printed in black and white. Among them, the Dutch (and also the Portuguese) risk label stigmatize risk. However, risk itself should be regarded as a neutral aspect of an investment. Since risk correlates with returns (typically lower risk correlates with lower rewards and vice versa), total avoidance of risk by consumers as a result of such illustrations could result in massive misallocations of the invested capital. Instead, the PRIIPs KID should aim at a neutral and comprehensive depiction of risk which provides an adequate information basis to assess whether the risk of a specific PRIIP is appropriate in view of the individual circumstances of an investor.
From our perspective the methodology used to prepare the presentations is decisive. If the methodology is not reliable and robust (and cannot be properly controlled by the authorities), the presentation is useless or can even be misleading. Therefore, to the extent possible performance presentations in the PRIIPs KID should be based on historical data (cf. our answer to Q6 above).
Among the examples contained in the discussion paper, we have a clear preference for the graph displayed on the top of page 43 left-hand side. This presentation appears engaging, understandable and comparable and should be very useful for investors provided that it is based on a reliable and robust methodology (please see graphic in our summary response attached).
In line with our answer to Q10 above, we advocate performance presentation net of costs in order to present credible outcomes to investors. In this regard, performance scenarios currently required for structured UCITS which are also presented net of costs should be considered a model for the further work by the ESAs (7).
In contrast, performance presentation in narrative form by using only text or a table (e.g. as shown at the bottom of page 41) are less engaging and less understandable than a visual element. The example at the top of page 42 does not provide any information on the relevant investment period nor on the likelihood for the returns that can be expected. Since this key information is missing, such presentation is useless or might even be misleading because investors could conclude that the indicated performance is to be expected in any event.
Lastly, the wording used to describe the illustrations should be selected with care in order to avoid misinterpretations by investors. For instance, the wording “pessimistic scenario” or “plausible worst case” in the examples on page 43 implies that no worse outcome is possible which should be problematic in case such outcome occurs in practice.
(7) For details, cf. CESR’s guidelines on selection and presentation of performance scenarios in the Key Investor Information document (KII) for structured UCITS (CESR/10-1318).
We have a preference for the combination option 3B as presented in table 9 on page 44. This means that risks and performance should be depicted separately, but within one section of the PRIIPs KID as foreseen by the Level 1 Regulation. Given the importance of the information on risk and reward, we recommend using the necessary space to show separate risk indicators for market, credit and liquidity risk as well as multiple performance scenarios. However, it could be appropriate to display multiple performance scenarios in one visual element like in the example at the left top of page 43.
Moreover, we would once again like to underline the relevance of proper understanding in terms of the linkage between risk and reward. Therefore, we strongly advise the ESAs to complement the summary risk indicator by illustration of such correlation which could take form of arrows used for the UCITS KIID (cf. our comments to Q12 above).
Finally, it is important that the methodologies for calculation of the summary risk indicator and performance scenarios are consistent. If there are discrepancies in the methodologies, the presentation could be misleading for investors.
From our perspective, the key questions on costs from the consumers’ viewpoint are adequately captured in table 10 as presented in the discussion paper.
We very much appreciate the depiction of the typical cost structures for insurance-based investment products, structured products, CfDs and derivatives as provided in the discussion paper. In our view, it represents a comprehensive and accurate overview of the main cost categories. In relation to insurance-based investment products, we fully agree that costs relating to the insurance cover should be included in the overall cost disclosure since they are clearly an element of the product’s cost structure (like the insurance cover forms part of the product’s benefits). As regards structured products and derivatives, we particularly value that the ESAs pay attention to transaction costs incurred in the hedging operations for a particular portfolio. These costs must be accounted for in the cost disclosure in line with the approach to be adopted for other PRIIPs.
We believe that the absence of certain transparency standards in specific sectors/with regard to particular categories of PRIIPs should not be used as an excuse to disapply the relating disclosure requirements. For instance, the EU fund industry is very transparent on costs in contrast e.g. to structured products for which no common cost disclosure standards yet exist. Still, it is the primary aim of the PRIIPs Regulation to introduce comparable information standards for all PRIIPs which should be of equal value to investors. This regulatory goal should not be compromised for practicality reasons.
In our view, the main challenges in achieving a level playing field in cost disclosures pertain to the following issues:
• Equal degree of transparency for all products: Unlike other PRIIPs, investment funds are very transparent in terms of costs. They do not only charge explicit fees which are known to investors ex-ante (such as management fee, performance fee and depositary fee), but also follow common standards in disclosing ex-post the overall costs deducted from a fund on a yearly basis (so-called ongoing charges for UCITS, previously known as “total expense ratio” or TER). Since insurance-based investment products or structured products involve implicit costs for which no generally acknowledged measure exists so far, the main challenge will be to make the costs of these PRIIPs transparent and to introduce standardised calculation methods which will ensure comparability of investor information in a way equivalent to the current UCITS rules.
• Ex-ante versus ex-post disclosure: The Level 1 Regulation requires disclosure of “direct and indirect costs to be borne by the retail investor” and hence seems to adopt an ex-ante view. However, many cost elements are known only ex-post and can be disclosed with a satisfactory accuracy only after close of an accounting period. For investment funds, this relates in particular to transaction costs and performance fees. Hence, we believe that historical data should be allowed to be used as a proxy for future costs in cases where a product’s cost history can be deemed representative. In case of expected changes and for new products, estimations should be made in line with the standards for the UCITS KIID (8). Moreover, cost disclosure should be accompanied by an explanatory statement clarifying the illustrative nature of the provided information (cf. also our comments on Q26 below).
(8) Cf. CESR’s guidelines on the methodology for calculation of the ongoing charges figure in the Key Investor Information Document (CESR/10-674) from 1 July 2010.
In our view, the methodology adopted for calculation of implicit costs must be capable of standardisation in order to ensure that the information presented to investors provides for a fair comparability of products. Therefore, while not wishing to prejudge the methodological debate, we would like to emphasize that standardability of a calculation methodology should be considered an important factor in the upcoming discussions.
Yes, we entirely agree. Even though we are accustomed to disclosing total costs by means of total expense ratio or ongoing charges for UCITS, we appreciate that a reduction in yield (RIY) type approach might better capture the overall impact of costs and allow for combination of different cost elements. In order to provide investors with meaningful information, scenarios underpinning the RIY calculation should correspond with the performance scenarios displayed in the risk and reward section (cf. our response to Q22 below).
No, the presentation in the discussion paper seems to cover all relevant calculation methodologies.
Growth rates assumed for the purpose of estimating “total aggregate costs” should correspond with performance scenarios displayed in the risk and reward section. If including or based on probabilistic modelling, such scenarios should allow not only for a realistic assessment of possible returns, but also for quantification of possible costs. Hence, disclosure of returns and costs on the basis of probabilistic scenarios could provide added value to investors in contrast to standardised growth rates not linked to a particular product.
We deem it premature to comment on possible methods for identifying implicit transaction costs before determining the general methodology for assessing implicit costs for structured products. In any case, it is imperative that the relevant method warrants for disclosure of all transaction costs relevant to a product portfolio as envisaged for directly investing products such as investment funds.
The type of product limits the comparability – the same is true for costs which vary according to the product type. In our view, the key challenge relates to the comparability of cost information on products with different life cycles and diverging payment profiles (one-off investments vs regular contribution).
As explained above, we think that cost disclosure should be tied in with the presentation of performance scenarios. However, in the interest of simplicity, presentation covering different rates of return should not be promoted. Instead, we recommend illustration of costs based on a representative scenario from the risk and reward section.
In this regard, we have a preference for a combination of options 5 and 6 on page 68 and 69 of the discussion paper. Whereas option 5 accounts for all PRIIPs requirements since it shows summary cost indicators as well as aggregate cost figures, it has the disadvantage of being focused on a specific time horizon (in the provided example 20 years). However, costs of many PRIIPs do not apply in a linear manner (for instance, in case of insurance-based investment products, costs are often charged right at the start of an insurance policy meaning that such products would show high cost amounts and low surrender values below the level of invested capital after the first year(s) of investment). This information is valuable for investors and should be adequately visible in the KID. Therefore, we recommend presenting option 5 by reference to different timeframes as foreseen in option 6.
As regards the other presentational examples displayed in the discussion paper, options 1 to 3 appear overly simplified and not capable of presenting cost information in a discriminatory manner. Options 4 and 5 are unnecessarily focused on the differentiation of particular cost items and not linked to the information on performance. The graph presented in option 7 does not provide investors with actual figures in neither percentage nor monetary terms and thus might be difficult to understand. The downside of options 8 and 10 is that they are not based on probabilistic performance and thus of limited value for investors while taking up quite a lot of space.
Regardless of the option finally chosen by the ESAs, we see the necessity to accompany the cost disclosure by a narrative explaining the exemplary nature of the calculations and clarifying that the actual cost level depends on a range of factors which cannot be accurately established in advance.
No, we are not aware of any PRIIPs featuring such cost structures. With respect to investment funds, precise information on costs including a cost breakdown is possible on ex-post basis. Ex-ante fund managers can only estimate costs and provide an approximate breakdown of such estimates.
Visualisation of risk by way of a synthetic risk indicator must be comprehensive for consumers. Therefore, we strongly support evidencing of comprehensibility by means of consumer testing. Testing of the UCITS SRRI with consumers has already underscored the merits of a scale-based risk indicator. In addition, possible methodologies for the risk indicator should be further tested by conducting sample calculations on the basis of different PRIIPs to ensure sufficiently discriminatory outcomes.
Generally, we would prefer a multi-dimensional indicator. Summarising all aspects of risks in one single figure might be over-simplistic and could not accurately capture the differences in the relevant risks across products. If investors are not aware of the different types of risk (as distilled by the ESAs), one single indicator will not prompt them to better understand. In other words, if e.g. an investment fund is relatively high on market risk, but low on credit risk and a structured bank product features exactly the opposite, both might end up displaying the same overall risk figure without investors being able to capture the difference.
Hence, we believe that all three types of risk (market, credit and liquidity risk) should be visualised by means of specific indicators. Based on the experience with the UCITS SRRI, we definitely favour simple scales similar to the example provided at the bottom of page 39. In this regard, we would rather abandon the overall risk indicator and stick to the separate illustration of different risks on a relevant scale.
Nonetheless, should consumer tests demonstrate the need for a single integrated risk indicator, we would advocate assigning equal weightings to each market, credit and liquidity risk. Equal weighting should be appropriate because it is ultimately irrelevant from the investor’s point of view which risk materialises if in the end the result – partial or full loss of the invested capital – is the same. Hence, as long as the relevance of risks cannot be determined in the same fashion for all PRIIPs, equal weighting of the different risk aspects appears the best way forward.
Lastly, we believe that the link between risk and reward should be displayed to investors in direct combination with the synthetic indicator. In line with the current standard for UCITS (cf. example 1 on the top of page 37), the risk scale(s) should be supplemented by a visualisation depicting the typical correspondence of respectively lower risk with lower rewards and higher risk with higher rewards (cf. also our reply to Q12 below).
We agree with showing cumulative costs either under the heading ‘cumulative costs’ (option 9) or classified as ‘effect of deduction to date’ (option 6). In addition, RIY figures over standardised time horizons could be included.
With regard to table 12 listing different cost categories related to PRIIPs, we would like to observe the following:
• Portfolio Management Techniques (PMTs): We support the notion of applying cost disclosure to earnings on PMTs such as securities lending or repos as far as such earnings do not benefit the relevant portfolio of a PRIIP. Similar standards are already in place for UCITS which are bound by the ESMA Guidelines on ETFs and other UCITS issues to disclose direct and indirect operational costs of PMTs in the fund prospectus and on ex-post basis in annual reports. In order to create a level playing field, transparency of PMT costs should become incumbent on all PRIIPs.
• Dividends: Dividends should be considered costs if they are withheld from investors. This should be the case if dividends on instruments beneficially owned by investors are not paid into the relevant portfolio and thus do not increase the value of the PRIIP. Such treatment is especially possible with regard to structured products or derivatives. Dividends on instruments held by investment funds should be considered income, since they are either passed on to the fund portfolio or distributed to investors.
• Performance fees: Disclosure of performance fees as part of the overall charges figure should be problematic. Since a performance fee is conditional upon the occurrence of certain triggering events, it may or may not be charged in the years to come. Therefore, it could be misleading for investors if the overall charges disclosure purported to illustrate the possible amount of the performance fee. In our view, performance fee could play into the presentation of overall costs only if such costs were calculated on the basis of performance scenarios and accounted for performance events which trigger the performance fee.
• Look-through costs: Costs of investments in other PRIIPs should be accounted for in the cost disclosure. We believe that the UCITS standards for calculation of the ongoing charges in case of target fund investments should be taken into consideration, but a greater degree of consistency should be ensured. In case of UCITS, the understanding of a “substantial proportion” of investments which triggers the duty to account for the target funds’ costs has not yet been harmonised at EU level.
• Transaction costs: Disclosure of transaction costs raises significant issues in practice. Transaction costs are to a high degree caused by the underlying market risk of the instruments traded. In case of e.g. bonds traded with bid and ask spreads, it is not possible to determine which part of the spread is attributable to the broker and which goes down to a market momentum at the time of trading. Moreover, as regards OTC derivative contracts, transaction costs are intrinsically embedded in the instrument price. Since the PRIIPs Regulation will affect only retail derivatives, we see no prospects for quantification of costs in relation to B2B derivative transactions.
However, should the ESAs decide to further promote disclosure of transaction costs, we deem it of utmost relevance that such disclosure pertains also to transactions conducted for the account of structured products and transactions concluded by insurance undertakings for the purpose of managing their capital investments. Only such broad understanding of transaction costs will ensure a level playing field among all PRIIPs and avoid unbalanced cost disclosures to the detriment of off-balance sheet products such as investment funds.
Moreover, it should be noted that “market impact” must not be considered costs since it simply accounts for the relevant market movement at the time of execution. MiFID II also recognises that any expenses “caused by the occurrence of underlying market risk” should not be accounted for in the overall cost disclosure (9). As the PRIIPs KID will be relied upon by MiFID distributors for providing appropriate information on product costs, the understanding of the relevant cost elements should be congruent under both EU frameworks.
(9) Cf. Article 24(4) second subparagraph of Directive 2014/65/EU (MiFID Level 1).
We agree with the suggestion to use a minimum standardised format. This should in our view include the manufacturer’s name, the link to the manufacturer’s website as well as a telephone number for consumers with no internet access. Given the limited space of the PRIIPs KID, the decision to include further information should remain with the manufacturer. This applies in particular to information such as a postal address since this requires much more space than a telephone number or a website link. We think the manufacturer should be allowed but not required to include an ISIN where available. It is an easy way for the consumer to identify the PRIIPs and to find information on the internet. Further, the ESAs should consider including the Legal Entity Identifier of the manufacturer.
We see merit in clarifying the criteria set out in recital 18. We agree that the assets in which a UCITS invests have to be considered as the assets in which a consumer commonly invests. As correctly pointed out by the ESAs, a UCITS is a product tailored for retail customers, hence the underlying assets should be considered as some in which a consumer invests. We think, however, the list should not be limited to the UCITS assets but should also include at least real estate and precious metals. These should be added to the list, too.
For the clarification, it would generally seem sensible to align the inclusion of the comprehension alert with the definition of complex products under MiFID II and IMD II. However, ESMA suggests in its technical advice to the Commission on MiFID II and MiFIR that all non-UCITS collective investment undertakings should per se be considered complex, regardless of whether they take the legal form of units or shares. As a consequence, all non-UCITS would qualify as complex products without any possibility to change that qualification by undergoing a complexity test. AIFs in general are often wrongly considered as risky products due to the AIFMD’s original approach to regulate hedge funds and private equity funds. Nevertheless, AIFs are all non-UCITS collective investment undertakings and therefore cover a broad range of products including highly regulated retail funds investing in the same assets as UCITS or in e.g. real estate, an asset class which is very popular for direct investments by consumers.
MiFID II, IMD II and PRIIPs Regulation together form the outcome of the overall PRIIPs initiative and hence should provide for a coherent set of rules. MiFID II intends to harmonise investor protection also with respect to complex products offered (see recital 70). On the one hand, MiFID II also recognises that a client is likely to require more information on complex products (see recital 83). On the other hand, MiFID non-complex products may be sold without the need to obtain information from the client and are understood as products that do not incorporate a structure which makes it difficult for the client to understand the risks involved. The PRIIPs comprehension alert aims to clarify for the retail investor that it might be difficult for him to understand the product. Consequently, such alert is intended to raise awareness that he might need more information. It therefore seems logical to require all complex products under MiFID II and IMD II to include a comprehension alert and all other products not.
Given ESMA’s suggestions regarding treatment of AIFs under MiFID II, however, an alignment of complex products under MiFID II and PRIIPs would hence exclude certain types of PRIIPs which invest in assets also commonly invested in directly by consumers and would therefore contradict the understanding of recital 18 of the PRIIPs Regulation. Nevertheless, an alignment would be preferable for investors and distributors: In light of the PRIIPs Regulation, ESMA’s interpretation under MiFID II should therefore be seriously called into question. In any case, it would not be in line with the PRIIPs Regulation if a comprehension alert would be required for products because they invest in assets in which consumer commonly invest (such as real estate).
In our view the legal form of a PRIIP should be sufficient to give the consumer a clear view on what kind of product he or she is dealing with. A very general description of the legal type of the PRIIP would give the consumer a general understanding of the context in which he has to put the product. The classification of the PRIIP could be refined further in order to clarify for the consumer what type of sub-product the PRIIP is. For example:
• Insurance (with-profits life insurance)
• Investment fund (open-ended real estate fund)
• Derivative (warrant)
In terms of space in the PRIIPs KID and given that other PRIIPs sections broadly describe the product features, the type of the PRIIP should in our view not include further information.
Further, we do not see any merit in using the classification of AIFs related to the reporting requirements under AIFMD. The AIFMD classifies funds into Hedge Funds, Private Equity Funds, Real Estate Funds, Funds of Funds and Other Funds (including equity AIFs). Hedge Funds and Private Equity Funds are mainly targeted at professional investors, i.e. the PRIIPs regulation does not apply. In addition, typical retail AIFs investing in securities would be classified as “other funds” which would not provide the consumer with a better understanding regarding the type of the product.
We generally agree that common principles could be helpful. These should be based on the principles laid down in Art. 5 and 7 of the UCITS KID Regulation (10) and should consider the CESR Guidelines regarding clear language and layout for the UCITS KID (11). We would support clarification regarding the main instruments that should be considered as direct or indirect investments.
Regarding the summary information on the pay-off structure, the ESAs already suggest a cross-reference to other sections. We strongly support this view for the following reasons: The PRIIPs regulation requires the manufacturer under the risk and reward section to include (i) performance scenarios and assumptions made to produce them, (ii) information on conditions for returns to retail investors or built-in performance caps, (iii) a statement that tax legislation may have an impact on the actual payout. This information is closely linked or even overlaps with the description of how the return is determined. Moreover, we believe that the description on how return is determined can only be understood within the context of this additional information. Manufacturers should therefore be allowed to include a description on how the return is determined within the risk and reward section and simply refer to this in the product section.
Furthermore, we would like to note that the phrase “where applicable” in article 24(3)(b)(ii) of the PRIIPs Regulation requires such description only for those investment products that are promoted as so called Responsible Investment products. Should the ESAs consider common principles or further guidance for those products, it is in our view crucial to use and refer to existing standards on transparency on the investment processes and selection methods but no rules which would affect the choice of content. In this regard we refer to the EFAMA’s note sent to EIOPA on November 26, 2014.
(10) Regulation (EU) 583/2010 of the Commission dated July 1, 2010.
We would like to remark that the PRIIPs text does unfortunately not refer to the known triad of ESG: Environmental, Social and Governance. The G for Governance is essential to ESG strategy, because if a shareholder has no relevant rights vis-à-vis the management of a corporation, it cannot effectively engage in a dialogue with the management and hence may not be in the position to improve social or environmental issues the corporation might have. We believe that it should be possible for PRIIPs manufactures to include in the description also governance criteria where applicable. This would also improve the quality of the information for the retail investor who would then be able to evaluate the product with respect to the complete triad.
Yes. As the ESAs already indicate, the information PRIIPs manufactures should provide on consumer types should be closely linked to the information on the target market under MiFID and prospectively also under IMD. It is in our view extremely important to follow a coherent concept regarding the target market and the consumer type for the following reasons:
• Target market in MiFID II and PRIIPs consumer type overlap significantly: The manufacturer under the PRIIPs Regulation has to identify the consumer type. For products which qualify as PRIIPs and financial instruments under MiFID II, in addition, a target market has to be defined. The obligation applies to the manufacturer as well as to the distributor, even if the manufacturer is outside the scope of MiFID II and does not identify a target market. The target market has to be identified in order to ensure that products are sold in compatibility with the needs, objectives and characteristics of the target market. For this, the product’s features and the risk and reward profile have to be taken into account. ESMA has so far refrained from providing details on the level of granularity other than abstract measures which have to be taken to define the target market according to MiFID II.
Both concepts shall take into account the risk and reward profile of a product. In addition, both features aim to give information on the investor/consumer type for whose needs, objectives and characteristics the product is intended. Both MiFID II and PRIIPs focus on abstract information from the manufacturer of the product. Both the target market and the consumer type are different from the suitability/appropriateness requirements applicable at the point of sale.
• Implementation of MiFID II target market and PRIIPs consumer type requirements will be complex: Manufacturer and distributor have to set up an appropriate level of information exchange in order to comply with the MiFID II requirements. PRIIPs manufactures will have to set up a system to evaluate their consumer types.
Given the coherent aim of both requirements and the complexity both MiFID II and PRIIPs provide for, additional complexity should be avoided and the consumer type should be defined as the consumer within the target market. The consumer type within the PRIIPs KID should hence describe the type of investor within the target market. This would avoid confusion and additional complexity as well as potential liability risks that could evolve if the target market and the consumer type are to be defined differently.
We agree with the ESAs that the key information on the insurance benefits should in case of life insurances cover what happens if the insured person dies as well as any other benefits of the insurance cover, e.g. insurance of payments in case of occupational disability.
Not generally. We would, however, like to point out that besides a fixed length or open-ended term, some products have a fixed term that might be extended. This applies for example to ELTIFs: The manager is required to define a term but may also define conditions which allow him to extend the term temporarily (12). Any rules regarding description of a term should take this into account.
(12) See Regulation on European Long-term Investment Funds.
No. We would like to note that some types of PRIIPs are protected against the manufacturer’s insolvency due to the product structure. This applies for example to UCITS and many retail AIFs. The manufacturer should be allowed to describe this within this section.
For some products, recommendation of a holding period might be difficult. Generally, the manufacturer should be allowed to explain the reasons for recommending such period, e.g. because the PRIIPs is typically subject to cyclical factors.
In general, the information requirements within this section should be coherent with the ESAs guidelines’ for complaints-handling, where covered firms are required to publish details of their complaints-handling process in an easily accessible manner. The manufacturer should be allowed to choose the link to a publication of the complaints-handling process on a website.
Yes, we generally agree with the requirement to publish additional information on the manufacturer’s website.
We agree that Article 6(3) should only apply if the choice of investment options is inherent in a product meaning that the legal form of a product should remain the same regardless of the final choice being made. Having said that, we are not aware of the existence of such products in the investment fund sector.
In Germany, the major category of PRIIPs falling under Article 6(3) should be unit-linked life insurance contracts. In the investment fund sector, we are not aware of products offering different investment options within one wrapper. Different investment strategies are usually being realized in individual funds set up either separately or as different compartments of an umbrella structure. Fund compartments are considered separate funds, each providing separate KIIDs to investors.
We would not recommend using examples or representative investment options for illustrating possible risks and rewards as well as costs of an investment. If a product offers investors a free choice between different investment strategies, stipulation of a representative investment is not possible and could mislead investors into thinking that the presented figures bear some relevance in terms of their final choice. Instead, the use of ranges for the risk/reward information and the cost data could be envisaged.
Description of the different investment options should at least correspond with the information required in the section “What is this product” under Article 8(3)(c) of the Level 1 Regulation for individual PRIIPs. Hence, the main features of the relevant investment options should be explained in particular to depict potential differences in the investment strategies and the relevant consumer types. This depiction should be accompanied by a clear reference to other pre-contractual information on the underlying investment options.
In case of products offering many options, it could be difficult for investors to relate the information obtained on the product wrapper to the information on specific investment options. This pertains in particular to the overall risk profile and overall costs as well as performance scenarios presented net of costs. Whereas costs incurring at the level of the product could possibly be added to the costs of the underlying investment option, obtaining an aggregated picture of risks could be more complex, especially if the product wrapper provides for some elements of insurance/capital protection not present in the chosen investment.
No, we do not agree with all of the measures outlined. We agree with the suggestion to operate on the UCITS KIID rules regarding review, revision and republication as a starting point. Any requirement, however, for the PRIIPs manufacturer to inform the investor of changes beyond a publication of the revised KID e.g. on a website would lead to significant practical issues and a regulatory overlap. For instance, a manufacturer might not have direct access to the consumer and would hence have to distribute the revised KID through a chain of intermediaries which would be very costly. If interested, the investor may easily download the KID from the manufacturer’s website. In addition, the PRIIPs KID is structured as a point of sale document to inform the investor prior to any decision on his investment. After-sale information is covered by MiFID II requirements as well as reporting requirements for the PRIIPs manufacturer which in case of funds include periodic reports to the investor.
Moreover, we support the notion to specify circumstances which amount to material changes triggering the duty to revise the KID. It is important to reduce the administrative and cost burden for the PRIIPs provider and investors by requiring the KID revision only if materially relevant changes occur. This approach is also in line with the current practice pertaining to the UCITS KIID.
For open-ended funds, we do not see any particular challenges. For closed-ended funds, however, it seems very burdensome to keep a KID up-to-date in case these funds are traded on a secondary market if such trading is based on the decision of a third party and not the manufacturer. Closed-ended funds have signing periods during which the product is offered. If a product is no longer sold by the PRIIPs manufacturer, there should be no obligation to keep the KID up-to-date.
No, at least not if the information in terms of risk/reward and costs of the underlying investment options is provided by displaying ranges in line with our suggestions made above.
We refer to our answer regarding Q49. Manufacturers should not be required to inform the investor actively on any KID changes.
Given the overlap between MiFID II products and PRIIPs, we agree with the ESAs’ proposal. The definition “in good time” should be identical with the definition in MiFID II. Regarding the question how the PRIIPs KID may be delivered, we would suggest to refer to Art. 38 of the UCITS KIID Regulation.
Yes. From our experience with the UCITS KIID, CESR’s template (13) provided a good starting point. It would, however, be helpful both for the ESAs as well as for the industry to develop sample templates for typical products such as a typical with-profits life insurance or an equity fund. The drafting of such sample templates might show some difficulties which could then be removed. Sample templates could further enhance standardization throughout Europe and would hence also be in the interest of consumers. Manufacturers, however, should be allowed to deviate from such templates provided they are compliant with the PRIIPs Regulation and the related Regulatory Technical Standards.
If the payment option is not embedded in the PRIIP, we do not think the payment options should be reflected. Hence the information in the KID should be based on the assumption of single payments in and single payments out. An example would be the option to use saving plans for fund investments, which is often offered by the distributor. In this respect, the manufacturer should be allowed to inform about the so-called cost-average effect. If, however, the PRIIP is structured for regular payments, such as life insurances, this should be reflected.
For fund managers, significant one-off costs will be caused by replacement of the UCITS or UCITS-like KIID by the KID following the PRIIPs requirements. This should be borne in mind in the review of the UCITS treatment as foreseen in Article 33(1) of the Level 1 Regulation. As regards the PRIIPs KID, material ongoing costs should be expected for keeping the information up to date, since performance scenarios and other information items will require regular reviews in accordance with Article 10. In these terms, the administrative and cost burden for the PRIIPs provider/investors could be reduced if the duty to revise the KID were only prompted by material changes of the relevant circumstances (cf. our reply to Q49 above).
As regards possible assumptions for the calculation of costs, we would like to make the following remarks:
• Ex-post vs ex-ante: As explained with reference to Q16 above, we have a preference for the use of historical data for the purpose of performance and cost calculation as far as such data can be deemed representative for future developments. Hence, historical figures should be used as a proxy in order to illustrate the impact of charges. In the absence of reliable historical data, estimates should be used until meaningful ex-post figures become available.
• Holding period: In line with our comments on Q8 above, we believe that cost disclosure should refer to a number of standardised time horizons in order to provide comparable information on costs relating to short-, medium- and long-term investments. In this regard, investment periods of 1 year, 3 years, 5 years, 10 years and the PRIIP’s lifetime or recommended holding period if longer should be covered. If such differentiated information is deemed too complex, we think that cost disclosure could focus on the timespans of 1 year, 5 years and 10 years (or the product lifetime/recommended holding period).
• Rates of return/growth rates: The rates of return assumed for the purpose of cost calculation should match with the return rates underpinning the performance scenarios in the risk and reward section (cf. our response to Q22 above). Since we recommend the depiction of probabilistic scenarios illustrating performance outcomes which can be reasonably expected in a PRIIP, it should be consequent to refer to the same scenarios for the purpose of cost disclosure in order to show probable results also in terms of costs. We believe that such approach would provide added value to investors in contrast to standardised growth rates not linked to a particular product.
• Assumed amount invested: Again, the assumption of the invested amount should be congruent with the investment assumed in the context of performance scenarios. Generally, we think that a one-off investment of €5,000 or €10,000 should provide an adequate basis to illustrate the expected performance and costs of most PRIIPs. Modifications could be allowed for products relying on regular contributions. However, it should be borne in mind that a modified basis of cost disclosure would impede comparability across different PRIIPs.