Primary tabs

Joint Associations Committee on Retail Structured Products (please see uploaded file for our cover letter and word version of our response to Questions 1-59)

Anita Millar
General Comments

• We generally agree with the description of the consumer’s perspective on risk expressed in the Key Questions. However, whilst we note the intention is not that these Key Questions would actually appear in the KID itself (in contrast to the questions identified in the Level 1 Regulation which will form the basis of the headings in the KID), given the volume of Key Questions identified, it will nonetheless be challenging to provide clear answers to these Key Questions (as required by the Discussion Paper) within a KID of just 3 pages.

Specific Comments

• Whilst the circumstances in which the investor could lose their money should be explained, the performance scenarios should be used to illustrate potential gains/losses. Hence we do not think that it is appropriate to include any additional disclosure in the KID in response to the questions: “How much am I likely to win” (i.e. the probability of potential returns) and “How likely is it that I lose my money” (i.e. the probability of potential losses). See further our response to Question 6 in relation to probabilistic modelling.

• The section of the KID entitled “What are the risks and returns” and the performance scenarios should effectively illustrate both the upside and downside of the PRIIP. However, we do not think it is appropriate to include any additional disclosure specifically addressing the question: “Is risk and return balanced?” as the number of assumptions that would be required to respond effectively to this question would not be appropriate for a 3 page KID.

• The statement required by Article 8(2) of the Regulation should be sufficient to confirm the purpose of the KID:

“This document provides you with key information about this investment product. It is not marketing materials. The information is required by law to help you understand the nature, risks, costs, potential gains and losses of this product and to help you compare it with other products.”

We do not consider that further elaboration on the purpose of the KID would be appropriate for a 3 page document. Hence we do not think it would be appropriate to include any additional disclosure addressing the question: “Am I able to get this information of other products and am I able to compare this product with other products?”
Do you agree that market, credit and liquidity risk are the main risks for PRIIPs?

• We consider market risk, whilst not the only risk, is the main risk to be evaluated when assessing the risk of a PRIIP. A PRIIP is defined as (in summary) an investment where, regardless of the legal form of the investment, the amount repayable to the retail investor is subject to fluctuations because of exposure to reference values or to the performance of one or more assets which are not directly purchased by the retail investor. Market risk (being the “risk of changes in the value or the PRIIP due to movements in the value of the underlying assets or reference values”) is therefore the primary risk in the context of assessing the risk of a PRIIP and demystifying the structuring/packaging involved in the payout which makes the investment a PRIIP.

• Unlike market risk, credit risk is not specific to PRIIPs. Our preference is for this to be disclosed in narrative form by reference to the credit rating of the manufacturer (or the swap counterparty, guarantor, or any collateral assets) where available and in response to the question “What happens if the manufacturer is unable to pay out?”

• For the vast majority of investors in retail structured products, we do not consider that liquidity risk represents a comparable risk to market risk or indeed credit risk. The reason for this is that market risk and credit risk are directly linked to the risk of a retail investor not receiving the return it expected. Liquidity risk will only affect an investor’s return in certain limited circumstances, specifically if it wishes to exit the product before its final redemption date. Many investors will not wish to or need to exit the product early. The timing of when an investor exits a product early is also likely to impact on the liquidity risk (to the extent this is affected by market risk) and in some circumstances liquidity risk may therefore be within the control of the investor.

Do you agree with the definitions the ESAs propose for market, credit and liquidity risk?

• We agree with the definition of market risk (as the “risk of changes in the value of the PRIIP due to movements in the value of the underlying assets or reference values”).

• For the most part we agree with the definition of credit risk (“as the risk of loss on investment arising from the obligor’s failure to meet some/all of his contractual obligations”) which should be viewed separately from the credit risk of the underlying assets represented by the market risk. However, we consider that credit risk should refer to the risk of loss arising from the obligor’s failure to meet some/all of his contractual obligations for payment or delivery rather than failure to meet some/all of his contractual obligations. We see credit risk as distinct from credit risk in a CDS context (in the context of a credit-linked note, for example, the credit risk of the underlying Reference Entities should form part of the market risk).

• We do not agree with the definition of liquidity risk (as “(i) the absence of a sufficiently active market on which the PRIIP can be traded; or (ii) the absence of equivalent arrangements”. Many PRIIPs (i.e. retail structured products) will be buy-to-hold investments that are not intended to be traded on the secondary market and hence do not have an active secondary market. Measuring the liquidity risk of these products against concepts pertaining to tradable securities, such as the existence of an active market would result in all such PRIIPs having a high liquidity risk which we do not consider to be representative and may well distort investors’ perceptions of the possibility of exiting the product early. Whilst a manufacturer/distributor will not typically be in a position to provide any contractual assurances in relation to liquidity, in practice the manufacturer/distributor/a third party broker of such products will often provide liquidity where an investor wishes to redeem its investment early.

• In our view liquidity risk should instead be disclosed in narrative form in response to the question “How long should I hold it and can I take my money out early”. This approach is in line with current practice for UCITS and with:

(i) ESMA’s opinion of 27 March 2014 on “Structured Retail Products – Good practices for product governance arrangements” which advocates: (a) good practice in relation to disclosure where a secondary market exists and (b) good practice for firms to offer and disclose appropriate exit opportunities to investors who may need to sell a structured retail product before its end term where products are not listed on a liquid secondary market; and

(ii) “IOSCO’s Final Report on Suitability Requirements with Respect to the Distribution of Complex Financial Products (see Footnote 10 below)” which flags that when a liquid secondary market for a complex financial product does not exist, the only prices available may be from the intermediary that sold the customer the product and states that the intermediary should “know and disclose ahead of time how these prices will be computed (using models, other markets for similar products etc) and what the price represents (mid-market theoretical value, re-purchase prices etc). The customer should have access to enough information to know that the product is illiquid, including information about the means and range of timing for disinvestment.”

Footnotes

Footnote 10: Page 13 of FR01/13, January 2013.
General comments:

• The most appropriate measure(s) to evaluate each type of risk will in some cases be product specific (e.g. the most appropriate measures to evaluate each type of risk in relation to funds is likely to be different to the most appropriate measures to evaluate each type of risk in relation to insurance products).

• Access to historical data is an important factor that could put limitations on the use of some of the measures proposed.

• Market risk could in most cases be measured quantitatively. Credit and liquidity risk should be explained qualitatively.

• Members strongly support the use of historical volatility of the underlying to indicate market risk quantitatively. Credit risk should be disclosed in narrative form either in response to the question “What happens if [the name of the PRIIP manufacturer] is unable to pay out” or immediately following the indication of market risk in the section entitled “What are the risks and what could I get in return?”. A standard table should illustrate credit risk based on the credit rating of the manufacturer (or if applicable, guarantor or swap counterparty or any collateral securing the product) on a scale of available credit ratings. To the extent there are no such credit ratings, the credit risk of the PRIIP manufacturer could be explained in narrative form. Liquidity risk could be indicated using a narrative or in a table format by reference the liquidity terms of the manufacturer (i.e. single dealer contractual commitment, single dealer soft commitment, exchange traded).

• Members strongly support the use of narratives to explain liquidity risk (rather than quantitative measures). Narrative disclosure should highlight whether there will be a secondary market for a particular product and the potential impact of any early redemption on investor’s return in broad outline (i.e. that they may not receive back the amount they would have received had they held the investment until its maturity and that they may receive less than they originally invested). Potential unwind costs should also be made clear to retail investors in general terms (taking into account the fact that the exact amount of such unwind costs would not be known for future dates. See our response to Question 17 under the heading “Early redemption costs” for further details). Attempting to show other indications of liquidity risk may well be misleading rather than helpful to investors. Liquidity is very difficult to show quantitatively as it is impossible to forecast a future bid-offer and whilst the bid/ask spread of executed trades may be of some use, such information may not always be readily available. Narrative disclosure could also indicate:
 - if the product will be tradable on an exchange
 - if a market maker will be supporting liquidity
 - regardless of whether exchange trading is envisaged, whether any other buy back commitment is available from the manufacturer or from the distributor (for example, a commitment to a maximum bid/ask spread under normal market conditions).

As indicated above, this information could be included in a simple table.

Specific comments from some members:

Taking each of the measures of market, credit and liquidity risk in turn members have the following comments:

Market Risk

Market Risk – Quantitative Measures

Volatility

• JAC members have expressed a strong preference for a volatility based measure of market risk (in line with the approach used in the UCITS KII with some amendments to provide for differences in product scope). Reasons for this include the fact that: (i) members wish any quantitative measure of market risk to be based on observable data and not predicated on assumptions. A measure based on historical volatility with respect to the underlying to which the product is linked (with an indicative profile taking into account how the particular product is linked to the underlying) has the advantage that it is based on observable data; (ii) for the purposes of comparing PRIIPs and UCITS and aligning the KII to the KID (if required) it is helpful for the market risk of both products to be volatility based.

Volatility of forecast returns

Some members expressed concerns with this measure for reasons including the following:

• they consider this measure to be highly subjective and vary widely between manufacturers as it is model dependent. They therefore consider that there is a potential risk that it could produce misleading results for investors.

• VaR and volatility of forecast returns are not considered to be retail friendly concepts and would be difficult to explain especially given the assumptions that are necessary to work out VaR.


Value-at-Risk or Expected Loss for a given Value-at-Risk:

Concerns expressed with these measures include:

• as indicated above, VaR is not considered to be a retail friendly concept

• an approach based on volatility of the underlying offers more stability over time

• an approach based on volatility of the underlying is more accessible for retail investors than an approach based on understanding probabilities (for example, historical and implied probabilities require assumptions as to probable outcomes which may not be understood by retail investors and as a result could potentially be misleading)

• access to data required for these measures (namely data based on forecasts) may present barriers for smaller manufacturers

Expected Shortfall for a given Value-at-Risk (ES_VaR):

Concerns expressed with this measure include the following:

• it measures the ability of the investor to bear losses (and represents an average loss for the investor under a given scenario as opposed to a single point in the expected distribution of the returns) but disregards the upside of the product

• some JAC members have conducted a study of a representative sample of structured products. It shows that the 99% Expected Shortfall VaR (“ES_VaR”) is a “digital” indicator, meaning it is either close to 0% (meaning the expected loss at maturity in the worst 1% of the cases is 0%) for capital guaranteed protected products or close to 100% for products without a capital protection (meaning the expected loss at maturity in the worst 1% of the cases is close to the full 100% capital invested). To obtain better granularity it would be necessary to look at a number of ES_VaR in order to avoid a binary classification of products

Market Risk - Qualitative Measures

Concerns expressed in relation to the proposed qualitative measures of market risk include the following:

• A number of measures proposed may be subjective and therefore not represent a useful basis for comparison of PRIIPs (e.g. how would the “qualities of market instruments underlying the products” or the “degree or risk concentration or diversification” be assessed objectively?)
• volatility of the underlying is more relevant than the type of underlying in assessing market risk
• risk diversification and leverage can only be effectively measured quantitatively
• it would be more appropriate to include a risk warning in relation to FX risk and other product design features than use them to measure risk

Credit risk

General comment

• Some members do not agree with the statement: “Next to the product characteristics and the overall creditworthiness of the manufacturer, investors’ claims under a PRIIP may sometimes be protected by a deposit or insurance guarantee scheme. All these elements may impact the credit risk attached to a product (see Footnote 11 below) ”. As indicated, we consider that credit risk relates to the risk of loss arising from the obligor’s failure to meet some/all of his contractual obligations for payment or delivery as opposed to the product characteristics. See also our comments below in relation to deposit insurance schemes.

• A number of members expressed a strong preference for disclosing credit risk in narrative form by reference to credit ratings (either short term or long term). Not all manufacturers will have a credit rating therefore this should refer to the credit rating of the manufacturer (where available) and if the manufacturer does not have a credit rating it may be appropriate to look to the credit rating of the guarantor or swap counterparty (for example, in a product where the issue proceeds are paid to the swap counterparty in return for the swap counterparty making payments due under the product) or any collateral assets securing the product. Where no such ratings are available, the credit risk of the manufacturer could be disclosed in narrative form by reference to other indicators such as prudential supervision.

Credit Risk - Quantitative Measures

Credit Spread or CDS Spread of manufacturer?

Concerns expressed in relation to using the Credit Spread or CDS Spread of the manufacturer include the following:

• Not all manufacturers will have CDS traded in the market
• Access to CDS Spread information may be problematic as this is usually not publicly available and may not be available in EUR
• Credit ratings are considered to offer less volatility than CDS Spreads

Credit Value at Risk

Concerns expressed by members in relation to Credit Value at Risk include:

• It is model dependent and therefore results from different manufacturers may be difficult to compare effectively
• It may be difficult for retail investors to understand particularly as it is not commonly used

Credit Risk - Qualitative Measures

Prudential Supervision

Whilst members see some value in this measure, concerns in relation to this include the fact that:

• it is country specific (what if, for example, the manufacturer/guarantor is based outside the EU)
• not all manufacturers (e.g. SPVs or corporates) will be subject to prudential supervision

Risk Spreading

Concerns expressed in relation to this measure are that it is not suitable for most PRIIPs which do not typically “invest” in a diversified basket of assets with different credit ratings in the same way as a UCITS fund.

The level of seniority and the secured or unsecured nature of the PRIIP

Concerns expressed in relation to these measures are that concepts of seniority and security may be difficult for retail investors to fully understand and would require explanation that may not be practicable within the 3 page KID.

Deposit insurance schemes

Concerns expressed in relation to this measure were that it appears to apply to a very limited number of PRIIPs (namely structured deposits).

Credit rating

A number of members expressed a strong preference for using the credit rating (either short term or long term) of the manufacturer (where applicable) or guarantor or swap counterparty for the following reasons:

• Objectivity (as opposed to measures based on individual manufacturer’s models)
• Ease of access to necessary information
• Ease of supervision by regulators
• Considered preferable to using CDS spreads for the reasons given above

Liquidity risk

Liquidity Risk - Quantitative Measures

The Bid-Offer Spread and average volume traded

A number of members support the use of the Bid-Offer Spread as a measure of liquidity. They note that the average volume traded can be a measure of liquidity but does not necessarily measure liquidity. Furthermore, information in relation to the Bid-Offer Spread will not be available upfront for inclusion in the KID.

The number of market makers

Members do not consider the number of market makers to be a good measure of liquidity for PRIIPs because (as explained in our answer to Question 3) PRIIPs are generally buy-to-hold investments. This does not mean that there is zero liquidity however as, whilst a manufacturer/distributor will not typically be in a position to provide any contractual assurances in relation to liquidity, in practice, the manufacturer/distributor/a third party broker of such products often provide liquidity (for exchange traded products for instance, a maximum bid/ask spread is set by exchange), where an investor wishes to redeem its investment early.

Liquidity Risk – Qualitative Measures

Characteristics of Exit Arrangements

Several members have commented that whether a product is listed but not exchange traded, listed and exchange traded, or neither listed nor exchange traded, is not necessarily indicative of its liquidity.

Footnotes

Footnote 11: Paragraph 3.4.2 on page 28 of the Discussion Paper
• We are not sure what exactly is meant by contingent costs in the context of the Discussion Paper (as this term is not used elsewhere in the Discussion Paper). We also question as to whether it is appropriate to disclose this level of detail in relation to costs in a 3 page document
• For the reasons below, we feel strongly that probability-based scenarios should not be required in the KID.

• We believe that non-probability-based scenarios are preferable as they will provide a fairer and potentially more reliable basis for comparison. This could be based either (i) on the manufacturer’s assessment of which scenarios of underlying performance best demonstrate the mechanics of a particular product or (ii) on a standardised positive, negative and neutral scenario of underlying asset performance (see further our response to Question 11 below in this regard).

• We appreciate the appeal of providing scenarios based on probabilistic modelling (in aiming to give investors some guidance as to the likelihood of the return they may receive), however, there are a number of fundamental conceptual and practical issues which mean they would be inappropriate for use in the KID.

• We believe that the key arguments against use of probability-based scenarios are as follows:

 - Requiring their inclusion is contrary to the role the manufacturer should play - The manufacturer’s role is part of a chain in the investment process. The manufacturer produces the product bearing in mind the likely needs of the selected target market and provides appropriate explanation as to the product’s structure, risks and rewards. The distributor must ensure it understands the characteristics of the product and the identified target market, and assess the suitability and appropriateness of the product for each client. The investor makes an investment decision based on his own risk tolerance, the information produced by the manufacturer and the assessment by and his discussion with the distributor. In order to give the investor the best possible outcome (both as to service provision and investment outcome), the roles of the parties in an investment transaction should be clearly delineated and the information required to be provided in the KID should reflect that delineation. The role of a manufacturer is to provide the quantitative assessment of the product by setting out the possible outcomes of the product in a set of non-probabilistic hypothetical scenarios. The role of a distributor is to provide a qualitative assessment of the product by assessing the appropriateness and suitability of the product for the investor. This way, the investor receives the appropriate information for his needs in the most appropriate form from the most appropriate constituency in the transaction.

 - A manufacturer does not know an investor’s market view – A manufacturer’s role is to provide objective information to explain the possible outcomes of a product in different market scenarios. Requiring the manufacturer to narrate on the likelihood of scenarios moves the KID away from a factual objective description of the product to one necessarily involving a market view. It is the investor’s own market view which determines whether a product is suitable for his portfolio. The manufacturer will not be, and in most cases has no way of being, aware of each prospective investor’s market view (this is not the manufacturer’s role or position in the distribution chain). The distributor should be aware of the investor’s market view and advise on the suitability of the product accordingly. Requiring the manufacturer to include probabilistic modelling in the KID would suggest to an investor that the manufacturer has taken into account the investor’s own market view in creating the product, thus misleading the investor. The risk of an investor being misled in this way creates a risk of misselling claims for manufacturers, which in turn may lead to manufacturers responding by reducing the products which they are willing to make available to investors, or even ceasing to manufacture PRIIPs altogether. A reduction in the range of investment options available to retail customers is clearly not a desirable outcome.

 - Investors are unlikely to understand the significant limitations of probabilistic modelling - Probabilistic modelling is highly technical, subjective and ultimately only indicative. We are concerned that end investors find it difficult to appreciate the limitations of probability modelling and the space restrictions on a KID would not provide sufficient space to fully explain such modelling and limitations. Furthermore, the assumptions that will need to be made to facilitate such modelling will be complex and will not be of a nature that would permit investors to readily assess their relevance to the particular PRIIPs they are considering and their particular investment circumstances. We believe the subjectivity of probabilistic modelling is not likely to indicate meaningfully to the investor the risk of a product. Such subjectivity is better assessed and explained by a distributor who can help the investor to assess the hypothetical performance scenarios provided by the manufacturer’s KID and to apply the investor’s own market view and also investment needs to enable the investor to determine whether the product is suitable for his portfolio. Experience also shows that retail investors often have difficulties in understanding probabilities or percentiles, which will increase the likelihood of them misinterpreting scenarios based on probability. There is also the possibility that, this being the case, investors may take undue reassurance from the scenarios provided and as a result elect not to take investment advice. This, in turn, is likely to result in an elevated risk of poor investment outcomes for those investors.

 - Probabilistic modelling is likely to vary significantly from manufacturer to manufacturer – As above, probabilistic modelling is highly subjective and its outcomes are likely to vary significantly depending upon different firms’ model assumptions and pricing parameters. It will be a significant challenge to formulate an approach to prospective modelling that will be appropriate and accurate for all PRIIPs and ensure consistent application across firms and products. Even if consensus could be reached on a model to compute performance scenarios, model calibration could still differ because of differing business models between manufacturers and different modelling capacities, leading to different outcomes.
 - Scenarios which are model dependant cannot be easily audited by regulators.

• It is also questionable whether the limited benefit a probability-based approach would deliver (given, inter alia, its potentially dubious accuracy and comprehensibility) could justify the likely significant cost (both in time and money) of implementing such modelling.

• If probability-based scenarios are adopted, detailed objective methodological guidance will be required to support this to ensure (i) consistency of application across firms and products (as above, we question the feasibility of this) and (ii) firms have a safe harbour from liability in the event that the modelling proves to be erroneous. It would be untenable to require firms to assume the liability for modelling probability for all products. As such:

 - the methodological guidance should not require any subjective input from firms when carrying out the modelling (i.e. it should be of a nature that firms can demonstrate objectively that they have complied with it); and
 - it should be made clear that if firms have complied with the mandated methodology they should not be liable to any person if such modelling later proves to have produced inaccurate or misleading results.
• We do not agree with using the probabilistic approach given the significant drawbacks associated with this. However, as discussed in more detail in our response to Question 6, should this approach be required, all modelling would need to be carried out in accordance with detailed objective methodological guidance prepared by the ESAs.
We believe that a flexible approach to time frames is essential to ensure that the performance scenarios section provides an accurate and fair point of comparison between PRIIPs. This is particularly the case in respect of PRIIPs with a fixed term or which are only realisable on specific dates during their term.

• It would not provide a fair or accurate point of comparison if comparing the return of a fixed term PRIIP with a flexible PRIIP (i.e. a PRIIP with regular put or call dates) on an arbitrary date on which the fixed term PRIIP is either:

(i) not likely to be realisable at all (whether under its contractual terms or on the secondary market);

(ii) not realisable in accordance with its contractual terms and only realisable on the secondary market (by definition at a price that would not be knowable at the point of producing the KID and/or may reflect a significant discount to its face value); or

(iii) is subject to early redemption penalties (in the form of a reduced return, fees or otherwise).

• It would be more appropriate to highlight any such limits to liquidity and/or penalties in narrative form in the liquidity risk disclosure elsewhere in the KID. Furthermore, it will be misleading if the performance scenarios section discloses returns received other than in accordance with the intended contractual return model of the particular PRIIP (e.g. where the PRIIP is intended for buy-to-hold investment with a fixed term, the scenarios should cover only the contemplated realisation date). For a fixed term PRIIP, this ‘at expiry’ analysis is the only way truly to isolate the behaviour of the underlying asset (which is the driver of the investment and the instrument on which the investor has a view) from the other market variables which may otherwise need to be assumed. This is also consistent with the expectation that structured products (in particular) are typically intended to be buy-to-hold/hold-to-maturity products.

• As such, we would suggest the following model would permit the most reliable point of comparison for investors:

(a) fixed term PRIIPs with a single contractual realisation date: The manufacturer is only required to provide return scenarios on the realisation date. This could be accompanied by a statement clarifying that the product is not realisable in accordance with its terms prior to the stated term and a cross reference to the liquidity risk disclosure.

(b) PRIIPs with a number of set contractual realisation dates: The manufacturer should only be required to provide return scenarios on the realisation dates and where there are more than three realisation dates the manufacturer should select 3 such dates which represent a short, medium and full term holding (i.e. holding until the final realisation date). This could be presented in tabular format as follows:

Holding period Percentage return
Favourable underlying performance (X%) Neutral underlying performance (Y%) Unfavourable underlying performance (Z%)
1 year [] [] []
2 year [] [] []
5 year [] [] []

We believe that the RTS should also permit the manufacturer the discretion (but not the obligation) to include return scenarios in respect of more or less than 3 realisation dates, if in the manufacturer’s determination this is necessary to accurately reflect the product.

• The ESAs may also consider whether standards could be set to assist manufacturers in selecting the short and medium term scenarios to improve consistency of disclosure between products (for example, mandating that for products with a term of 1-5 years, if it is possible to realise a product at the 1 and 2 year marks these should always be selected as scenarios and for products with a term of 5-10 years, if it is possible to realise a product at the 2 and 5 year marks these should always be selected as scenarios).

• Given the range of PRIIPs, it may be that the above model would not be appropriate for all products. As such, the ESAs may consider specific rules/exceptions for particular products (e.g. those with a particularly long (or open-ended) term or US-style warrants).
• Percentage figures:

 - Whilst noting the concerns regarding potential behavioural bias amongst investors, for the reasons noted in the Discussion Paper it must be the case that percentage-based figures offer the most direct and meaningful point of comparison between any two products.

 - We would agree with the drawbacks noted in respect of monetary amounts and note further that if an investor is unable to compute simple percentages it is likely that they would be unable to relate the monetary example given in the KID to their actual investment amount. In the majority of cases this actual investment amount is likely to differ from the amount given in the example.

 - There is also significant risk that if different assumed monetary amounts were permitted for different products (which would necessarily be the case (see Footnote 12 below)), if percentage figures are not also disclosed, this could confuse investors leading them to conclude erroneously that the returns for those products with a higher assumed investment amount were higher, where this may not be the case in percentage terms.

• Monetary amounts:

 - Notwithstanding the foregoing, we are aware that some investors feel more comfortable thinking in monetary terms.

 - As such we believe, on balance, that the more pragmatic approach would be a hybrid approach (using both percentage figures to aid comparability but also monetary amounts as a further illustration). Such approach could either (i) mandate the use of percentage amounts and permit the manufacturer also to include monetary amounts at its discretion or (ii) mandate both percentages and monetary amounts.

 - As above, if monetary amounts are permitted/mandated it will be important that the scenarios are based on an assumed investment amount and currency that is in fact investable in the product in question (rather than a standardised investment amount for all products which may not correspond to an amount that the investor would in practice be permitted to invest).

• Absolute amounts: We see no separate benefit to using absolute amounts.

• Consumer testing: In light of the above, we agree that consumer testing will be important in determining the solution that works best for investors. We note in this respect:

(i) it will be important that the testing is structured carefully to ensure that it tests accurately the likelihood of misinterpretation under each approach (i.e. percentages only, monetary amounts only or a hybrid approach) and in which case this is most fundamental to the investor’s overall understanding of the investment proposition; and

(ii) that the sample of investors used for the testing accurately reflects likely investors in PRIIPs. Given the nature of PRIIPs products, it seems questionable that the likely investors would not understand basic percentages.

Footnotes

Footnote 12: On the basis that it would also be potentially misleading to require a manufacturer to assume an investment amount for a product for the purposes of the performance scenarios if that amount was not in fact investable.
• Only those costs known to the manufacturer at the time of preparing the KID should be subject to the inclusion in the performance scenarios, with the exception of costs that may vary as between any wrapper, distribution channel, holding model or investor. Such variable costs will vary from investor to investor and as such cannot be disclosed in the KID. In addition to any fees charged by distributors, these variable costs may include costs paid directly by the investor to third parties such as custody costs or brokerage fees or in tax.
• The Discussion Paper notes the balance to be struck between providing investors with useful insight into performance whilst at the same time being realistic about the space available in the KID. It will also be important to ensure that investors are not confused or disengaged by the volume of performance information provided.

• We believe that the number of scenarios provided (which we take to mean the number of assumed scenarios of performance for the underlying in respect of which corresponding product performance disclosure is given) must be considered in conjunction with Question 8 in respect of timeframes. For example, assuming that a number of timeframes may be mandated (likely 3 (short, medium and long)) (see Footnote 13 below0 , mandating 5 performances scenarios to be disclosed in respect of each of these 3 timeframes would result in 15 outputs for an investor to analyse which would seem too many.

• Further, the challenges of space cannot be underestimated and given the volume of disclosure already contemplated in the Discussion Paper it would seem that, even with a limited number of scenarios, it will be very difficult to include all the information currently contemplated within the 3 A4 pages.

• We believe that a sensible compromise would be to include 3 scenarios. We believe that the most informative approach for investors would be to permit the manufacturer to select these scenarios based on what, in the manufacturer’s assessment, best demonstrates the mechanics of the particular product to the investor (e.g. if the product contains a barrier, scenarios showing the return where the performance of the underlying is below, at and above the barrier). If the ESAs are of the view that a greater degree of standardisation is required between products, an alternative could be to require scenarios based on a positive, negative and neutral outlook for the underlying. This would allow for a simple and standardised (if somewhat arbitrary) point of comparison between products.

• Where the product relates to a basket of underlyings it would need to be assumed that all underlyings had this same performance unless, in the determination of the manufacturer, this would not be appropriate in the context of the particular product (e.g. a “worst of” product).

• We would agree that anything less than 3 scenarios would not give an appropriate range of outcomes, however, we believe that 5 scenarios is both potentially confusing for investors (in providing too much information) and challenging to display clearly and intuitively in the space that will be available in the KID.

• The only caveat to this would be where an additional scenario(s) may be required to effectively disclose the consequence of a particular product feature (e.g. an autocall or a conversion right). We would suggest that in these circumstances a manufacturer should have the flexibility to include a further scenario(s) if the manufacturer considers that beneficial to investors. This could be achieved by adding additional rows and/or columns in the table we suggest in our response to Question 8 or providing a separate table addressing the relevant feature.

• It will be important to allow the manufacturer to make assumptions regarding other variables (i.e. other than the performance of the underlying assets) which may impact performance (as these will necessarily need to be fixed for the purpose of generating scenarios) and to disclose these assumptions in narrative alongside the performance scenarios. Indeed, for additional variables that occur in the context of common products it may be most effective for the RTS to include guidance as to what assumptions should be made.

Footnotes

Footnote 13: As per our response to Question 8, we would argue multiple timeframes should only apply to products with multiple realisation dates. If a product has a single contractual realisation date, the performance scenarios should only relate to redemption on that single date.
In relation to the risk indicators illustrated in the Discussion Paper, JAC members expressed the following views:

• There was some preference for the sliding visual scale from Italian research (perhaps utilising colour and letters instead of numbers). Others had a preference for the risk indicator from the Dutch financial leaflet. The Belgian risk indicator was considered to look more like a label used on consumer goods rather than financial products. The risk indicator from the Netherlands was considered overly simplistic.

• The table tested in consumer research in Italy was considered to use concepts which were not considered to be retail friendly and could be confusing (VAR, Turn-over-ratio, Annual expected default frequency). Further the concept of comparing with “similar products” was considered impossible to achieve in practice as structured products have a pre-defined payment profile that is not designed to outperform a specified benchmark/similar product and hence cannot be readily correlated with any benchmark/similar product. The suggestion for comparisons to be made with “similar products” also does not take into account the fact that the retail market is in part a privately placed market, where not all products are publicly available. There would therefore not be a single source where dealers could be sure to find data relating to “similar products”.

• For the most part, however, JAC members are strongly in favour of a single indicator of market risk (determined by reference to the historical volatility of the underlying). We consider that credit risk should be illustrated in narrative form in response to the question “What happens if the manufacturer is unable to pay out” (although if it is considered necessary to include this in the section entitled “What are the risks and what could I get in return” then it could be included immediately following the risk indicator). This could be illustrated using a simple table to illustrate credit risk based on the credit rating of the manufacturer (or guarantor or swap counterparty or collateral assets where applicable) on a scale of all available credit ratings. We consider that liquidity risk should be disclosed in response to the question “How long should I hold it and can I take my money out early” (with only a cross-reference to the section entitled: “What are the risks and what could I get in return” if this is deemed necessary). This should be disclosed in a narrative form or in a table format by reference to the liquidity terms of the manufacturer (i.e. single dealer contractual commitment, single dealer soft commitment, exchange traded).

• As indicated in our response to Questions 3, 4 and 5, all JAC members emphasise that the different types of risk (market, credit and liquidity) must not be combined as to do so would potentially misrepresent such risks to investors. A single visual indicator (aggregating credit, market and liquidity risk) may offer apparent simplicity but it is not informative enough. In mixing different notions of risk it would not give investors the opportunity to discriminate between them even though they may attribute different weightings to these different types of risk. As such, combining all risks in a single indicator may prevent investors making an informed investment decision and potentially lead to inappropriate investment choices contrary to the purpose of the KID.
• We agree with the ESAs’ assessment (in Table 8) of the relative merits of the different approaches to presentation of performance scenarios.

• Simplicity will be paramount to ensuring investor engagement and comprehension.

• A single visual element would appear preferable both on the grounds it is likely to be more engaging and intuitive for investors and represents the most efficient use of the space available in the KID.

• As indicated in our response to Question 1, we note that it is also important that performance scenarios required by PRIIPs will, to the extent relevant, satisfy and/or be aligned with the requirements for performance scenarios in MiFID II to ensure that there is no duplication of work or cost.

• We believe that a tabular approach is favourable. Please see our response to Question 8 which sets out our rationale for this and suggested presentation. This table would need to be accompanied by clear narrative disclosure noting that the scenarios shown are examples only and summarising any key assumptions used in compiling the scenarios.

• We do not believe that graphs are an appropriate format on the basis that:

 - They may imply to investors that a product could be realised at all points (i.e. anywhere along the line plotted) and at the value plotted in each case. In the case of many products this will be incorrect (as realisation will only be possible on one or more set dates) and in the case of products with a single contractual realisation date would be entirely misleading.
 - Necessarily, any points appearing in the graph that relate to dates other than the contractual realisation dates would either be misleading (if plotted contractual realisation dates were simply joined up) or would need to be plotted on the basis of extra-contractual realisation (e.g. secondary market sale) the return for which is like to be unknowable. As discussed in our response to Question 8 this would not appear to offer a fair or valuable point of comparison.
 - Even for products that are realisable on a more frequent (e.g. daily) basis it may not be realistic to expect a manufacturer to produce a continuous model for return scenarios for the purposes of plotting such a graph.

• In respect of the other examples provided in the Discussion Paper:

 - For the reasons discussed in our response to Question 6, we believe that an approach based on probability would be unreliable, with significant potential to mislead. In any case the dot-based form on page 42 does not appear particularly engaging or intuitive and may require detailed explanation.

 - We believe that the “benchmarking” elements included in the Italian example on page 42 and Dutch example on page 43 would be entirely inappropriate for PRIIPs that are structured products. Unlike some funds which may have a stated investment objective to outperform a specified benchmark, structured products by contrast have a pre-defined payment profile that is not designed to outperform a specified benchmark and cannot be readily correlated with any benchmark. As a result we do not consider that it would be possible to identify benchmarks that would provide a meaningful point of reference for structured products that are PRIIPs. As is the case for probability modelling, given the breadth of products contemplated under the PRIIPs regime, if a benchmark were to be artificially selected for a PRIIP that is a structured product, it would be necessary to include a significant number of assumptions with the result that the scope to mislead or create unfair bias in favour of certain products would be significant.
 - the concept of “plausible best case” in the Dutch example at the top of page 43 would necessitate the use of forecast models that dealers do not have and would likely get wrong.

 - The Italian example on page 41 is not visually engaging.

 - The Spanish example on page 41 seems cluttered and we question the value of including the second row (“Return =”) which would appear to add to the volume of information without necessarily improving comparability.
• We believe that a single risk indicator reflecting market risk only is preferable. This should be combined with a single “visual” performance scenario in tabular form. A single visual performance indicator is important to ensure that investors do not feel overloaded with information and hence prone to disengage.

• From a risk perspective, our preferred approach would be:

• for credit risk to be disclosed in narrative form in response to the question “What happens if [the name of the PRIIP manufacturer] is unable to pay out?” and by reference to credit ratings of the issuer (or swap counterparty, or guarantor, or any collateral supporting the issue) where available.

• For liquidity risk to be disclosed in response to the question “How long can I hold it and can I take my money out early” with no more than a cross-reference to the section entitled “What are the risks and what could I get in return” if this is deemed necessary.

• If it is determined that the risk indicator should reflect both market and credit risk (as discussed above, we believe that it should not) we believe that it is important that these risks are not aggregated for the reasons set out in response to Questions 3, 4, 5 and 12.

• For the reasons discussed in our responses to Questions 3, 4, 5 and 12, we believe it is important for investors to be able to distinguish between the different elements of risk.

• It seems most logical for liquidity risk to be disclosed in the sections that discuss realisation and as such this should assist with investor comprehension.

N.b. We have assumed that in referring to “combinations” you do not intend that the risk indicator and performance scenarios would be visually combined into a single indicator. We would not support this as it would be unduly complex and is unlikely to be something that would be intuitive for investors.
General Comments on Key Questions:

• The description of the consumer’s perspective on costs expressed in the Key Questions appears overly complex for a 3 page disclosure document focusing on key information only

Specific comments on Key Questions:

(1) How much will this investment cost?

• JAC members emphasise that the KID can only be used to show costs known by the manufacturer. There may, however, be other costs an investor incurs (e.g. from a financial advisor, broker or custodian or in terms of applicable taxes) which cannot be disclosed in the KID as they will not be known by the manufacturer and/or may be investor-specific. As indicated in our response to Question 1, advisors, distributors or any other person advising on, or selling, the PRIIP should provide cost disclosure in accordance with MiFID II. Such costs disclosure should not be included in the KID itself as this is a manufacturer document and the manufacturer will typically have no knowledge of a distributors costs etc). This approach is supported by Article 8(3)(f) of the Regulation which states:

“The key information document shall include a clear indication that advisors, distributors or any other person advising on, or selling, the PRIIP will provide information detailing any cost of distribution that is not already included in the costs specified above, so as to enable the retail investor to understand the cumulative effect that these aggregate costs have on the return of the investment”

• JAC members also question the need to include a breakdown of all costs in the KID. It seems doubtful that retail investors will be interested in such a breakdown which would also appear to add unnecessary complexity and detail to a 3 page disclosure document. The Level 1 text of the Regulation also does not require any breakdown of costs (Article 8(3)(f)) states that the costs associated with an investment in the PRIIP “comprising both direct and indirect costs…including one-off and recurring costs” should be “presented by means of summary indicators of these costs, and, to ensure comparability, total aggregate costs expressed in monetary and percentage terms, to show the compound effects of the total costs on the investment.”

(2) How accurate is the current estimation of the overall costs?

• JAC members note that for some PRIIPs costs will not be known up-front. Where this is the case, it will be difficult to specify the accuracy of cost estimations. In such circumstances members have suggested that it may be appropriate to disclose maximum costs.

(3) What are the possible circumstances that trigger additional charges?

• JAC members consider that the questions relating to early exit should not be addressed in the costs section of the KID but in the section entitled: “How long should I hold it and can I take my money out early?”

• JAC members consider that answering the following two questions would require a level of detail not appropriate to a 3 page KID and question whether investors would want to know this level of detail:

- How does the manufacturer calculate what I must pay?
- When I am being charged costs on my investment, how or why are these costs being generated?

(4) How do the costs impact my returns?

• JAC members have expressed concerns that the following questions: “How much of my initial investment remains after cost deduction?” e.g. “How much of my investment is really invested?” may be misleading as retail investors may assume that the initial investment net of costs represents the value at which they may exit the product or that it represents the amount invested on which its return will be calculated. In fact, in the case of structured products with embedded costs, the amount invested will typically be the full notional amount and returns will also be calculated on this amount and not the notional net of costs.

(5) What part of the costs will contribute to the level of capital protection provided by the PRIIP?

• JAC members do not consider that the costs of capital protection in the KID can be accurately and consistently shown across the market. Providing a breakdown of the zero coupon bond price and other components would appear to require excessive detail for a 3 page KID and be of questionable interest to an investor primarily concerned with the overall impact of fees on his investment rather than any breakdown of such fees.

(6) Will I receive updates on my costs?

• Any updates on costs should be provided in accordance with the Article 10. JAC members note that for PRIIPs for which all costs can be accurately disclosed upfront it will not be necessary to provide an update on costs.

(7) Comparing Costs

• JAC members note that they do not consider it to be feasible or practicable to include costs for other PRIIPs in a 3 page KID. It should be for the retail investor to compare the costs by comparing different KIDs for different products.

• JAC members consider that the following question would be difficult to answer in the KID as it will depend on the performance of the product (and often may not be ascertainable until after its final redemption): “If I pay more in costs for this product will I receive a better return on my investment?”
Implicit costs and Embedded Costs

• The PRIIPs Regulation requires disclosure of: “both direct and indirect costs to be borne by the retail investor, including one-off and recurring costs, presented by means of summary indicators of these costs and, to ensure comparability, total aggregate costs expressed in monetary and percentage terms, to show the compound effects of the total costs on the investment” (Article 8) – we do not agree with the assumption in the Discussion Paper that the reference to “indirect costs” in the Level 1 text of the Regulation necessarily equates to disclosing “costs embedded in the purchase price”. “Indirect” costs could instead refer to ancillary services or costs incurred in relation to the underlying portfolio (rather than the PRIIP itself) rather than embedded costs. Further, there is no suggestion in the Level 1 text that to the extent embedded costs must be disclosed they should be broken down in the KID.

• Members agree in general with the outline of the main features of structured products and derivatives with the exception of the references to: (i) “implicit costs” and (ii) “costs embedded into pricing”. Members do not consider that these are meaningful indicators as investors should focus on the participation they are getting rather than what it took to get there. For instance, the same product may be offered by 2 different firms as follows:

 - One firm may offer a structured note at £100 with a 90% participation into the FTSE Index with a fair value of 98.5 - in other words, the embedded costs for the firm is 1.5%
 - Another firm may offer the same structured note at £100 with an 85% participation into the FTSE Index with a fair value of 99 - in other words, a lower cost base and charges 1% only.

Assuming the credit rating of the manufacturer (or where applicable, guarantor or swap counterparty) is the same and that there are no distribution fees, the option offering the investor 95% participation into the index, is preferable for the investor notwithstanding the higher embedded costs. The fact that some firms have a greater funding appetite or a more efficient trading platform which means that they can extract potentially more revenue is irrelevant. The pertinent information for the investor to compare is (i) the participation level in the index, (ii) the credit rating of the manufacturer and (iii) the liquidity terms of its secondary market (if they may wish to exit the product early).

• Other reasons why JAC members do not consider it beneficial to disclose embedded costs in KID include the following:

 - There is no consistent approach across firms for calculating embedded costs. Different firms will have a different definition of how the component parts would be calculated for a given underlying and product. The more complex the product/the underlying the greater such differences are likely to be. The ability to compare embedded costs as between different manufacturers is likely to be severely compromised by such inconsistencies which has the potential to confuse or mislead investors. This is noted in the Discussion Paper where it states: “One limitation that applies to both approaches is that there is no guarantee that two manufacturers would agree on the costs of such products.”

 - Firms typically take a “portfolio” approach to risk, hedging and cost – i.e. you may be able to offset some risk of a given product versus other products which may mean the “cost” you charge to the end consumer may be lower or higher.

 - Asking firms to be transparent about (each individual) cost in a consistent way may force firms to increase the disclosed cost across the board to the detriment of investors. An analogy would be to a builder who may quote an overall price which will be lower than if he is asked to quote separately for each individual piece of work.

 - Firms have different funding needs and may treat any funding that they can price into a product in a different manner to one another. Some may offset low funding needs (i.e. lower than average funding priced into the product) by reducing their profitability and achieve the same participation levels or coupon levels as the competition. Investors would therefore see the same participation level with widely different “hedging costs” and “manufacturer profitability”. This information is more likely to be misleading for investors than helpful disclosure.

 - As indicated in our response to Question 15, disclosure of embedded costs in a KID may also be misleading for retail investors because such embedded costs are not deducted from the returns an investor receives and they may vary depending on market conditions. For example, a retail investor may well assume that their initial investment net of costs represents the value at which they may exit the product or that it represents the amount invested on which the product’s return will be calculated (whereas in fact, in the case of structured products with embedded costs, the amount invested may well be the full notional amount and returns will also be calculated on this amount and not the notional net of costs). This fact is also misrepresented in the following example given in the Discussion Paper which states:

“if a manufacturer sells a structured Euro Medium Term Note (EMTN) at 1,000€, he should disclose in the KID that 3% (30€) of the purchase price is a sales commission and 2% (20€) of the acquisition price will be absorbed upfront to recompense the manufacturer for the costs the manufacturer incurs when structuring the note. The result is that 95% (950€) of the acquisition price will be invested in the note: there are 5% costs”

 - It is not correct to conclude that because costs amount to 5% of the issue, only 95% of the amount paid by investors will be invested in the product. This is because, as indicated above, a manufacturer will often use portfolio hedging and also because the cost of the product is not deducted upfront but will be taken during the tenor of the product and may be absorbed by any profit higher than expected resulting from hedging activities.

In this regard it is worth noting that MiFID II states that “Loss to the investment as a result of underlying market risk is not to be viewed as a cost”. A manufacturer’s fees and costs may be reflected in the pay-out formula and the profit or loss made by the manufacturer may vary depending on market conditions. As profit or loss made on a retail structured product is therefore related to underlying market risk arguably it should fall within the exemption for non-disclosure of losses related to underlying market risk in MiFID II and should similarly not be disclosed in the KID.

 - We assume regulators may wish embedded costs to be disclosed on the misguided assumption that these are comparable to fees deducted from an investor’s returns such as management fees charged by funds. However, the costs associated with non-fund PRIIPs are not equivalent or comparable to fees such as management fees charged by funds:

 - In the case of management fees charged by funds there is a contractual provision for fees to be deducted from the performance of the underlying assets and charged directly to the customer. In an unwind scenario, there will be a contractual provision for accrued management fees to be deducted from an investor’s accrued entitlement. Without full disclosure of the management fees, the investor cannot calculate his expected returns.

 - In the context of retail structured products by contrast, (as mentioned) embedded fees are not deducted from the returns an investor receives, nor does an investor’s initial investment net of costs represent the amount invested or on which their return will be calculated (which will generally be the full denomination of the security). Therefore the investor can calculate his expected return without the need for disclosure of the embedded costs.

 - In contrast to funds, retail structured products are buy to hold products and investors will not generally exit the investment early at all and to the extent they do, the price at which the manufacturer/distributor/another entity may offer to buy-back the product will generally reflect unwind costs. However, such unwind costs are not comparable to accrued management fees deducted from an investor’s return where an investor exits a fund as they do not represent a contractual entitlement to accrued fees deducted from an investor’s return, rather they represent the market risk of the product at the relevant time which cannot be accurately estimated in advance.

 - For the reasons set out above, it is incorrect to assume that embedded fees are comparable to fees deducted from an investor’s returns such as management fees charged by funds. However, we are cognisant of the fact that in the event an investor were to exit their investment early, potential unwind costs should be disclosed in general terms the KID (taking into account the fact that the exact amount of such unwind costs would not be known for future dates). See the section of the response to this Question 17 headed “Early Redemption Costs” below for further details of our proposals in relation to this.

 - Retail investors should be able to compare the overall economics and competitiveness of products between manufacturers without the need for disclosure of these embedded costs by looking at KIDs from different manufacturers and seeing how the economic pay off of a particular product might differ between manufacturers.

• To the extent regulators conclude that embedded costs must be disclosed in the KID, this should cover the manufacturing costs (upfront and on-going if any). Such costs should cover operational and regulatory costs related to issuance (e.g. manufacturer fees, hedging costs, listing costs).

It should not be necessary for such costs disclosure to be broken down further as it is not clear what benefit this would have to the retail investor or how this could be achieved within a 3 page KID.

If embedded costs must be disclosed, some JAC members have suggested these be disclosed as follows:

The template breakdown below assumes a product with a 3 year term, 3% structuring fee and no running manufacturing costs. Distribution costs are not reflected as these will often not be known by the manufacturer and even if they are known, disclosure of distribution fees for some products but not others would distort comparability across products.


Cost Type Upfront Running (per annum) Performance Fee Total costs over the product tenor Total costs
per annum
Manufacturing Cost Manufacturing Cost

Percentage (%) 3% 0% 0% 3% 1% p.a.
Absolute
(based on
an investment
of EUR 1000 ) EUR 30 0 EUR 0 EUR 30 EUR 10 p.a.




Fair Value Approach

JAC members have expressed strong views against the fair value or intrinsic value approach for reasons including:

• Fair value is based on internal pricing models and assumptions which differ as between manufacturers thus undermining its consistency and comparative value

• Even if a common, binding methodology is defined, JAC members express doubt as to its usefulness for investors. It represents an indication of potential margins for the manufacturer, but:

 - fair value is not static and it changes daily (so as a minimum, it should be clear that only fair value at issue date is required, and no further update is needed);

 - until maturity, fair value is a good faith representation of the expected margin, although the occurrence of profits or losses will be entirely dependent on the movement of the underlying;

 - other costs are more meaningful for investors, in particular: (i) costs/fees that directly reduce potential returns for an investor; and (ii) manufacturing costs.

 - retail investors may struggle to understand the concept of fair value and the inputs used to compute this

 - differences in fair value do not necessarily equate to differences in costs for investors (a difference in fair value may for example represent a difference in the hedging strategy of two manufacturers)

 - not all costs may be included in the fair value (e.g. running costs booked on an accrual basis would not be reflected in the fair value of a product on day 1)

Cost structure of derivatives

• The cost structure of a derivative is described as follows in the Discussion Paper:

 - the intrinsic value of the derivative;
 - a front-end load fee;
 - a sales commission; and
 - the manufacturer margin (covering the operational costs incurred by the manufacturer for structuring the derivative, market-making costs (spread), settlement costs and the profit of the manufacturer)

However, the intrinsic value of a derivative is not necessarily the intrinsic value of the option in financial terms, but the cost of acquiring such derivative in the market. This represents the product value but it is not a “cost” charged by the manufacturer to the client.

• In relation to “dividend costs”, JAC members note that it is true that for some structured products, investors may “give up” some dividend return which is effectively used to increase their participation into an index or their barrier / coupon level. However, this is not something that can be easily quantified in a meaningful way. For example, a structured product may include a call option on an index referencing a price return index. The dealer has to make an assumption as to how much dividend yield he will receive if the physical hedge is held on its balance sheet. However, each dealer will be subject to a different tax treatment for these dividends, or may even decide to hedge the product in a synthetic way and not benefit from the dividend yield. Finally, the dealer may not hold 100% of the option notional on day one but only the “delta” of the option as recommended by its model. So any dividend yield disclosed would be unlikely to be of assistance when comparing different PRIIPs and may have no resemblance to the yield published for the relevant index.

Early redemption costs

• For the reasons set out above, it is incorrect to assume that embedded fees are comparable to fees deducted from an investor’s returns such as management fees charged by funds. However, we are cognisant of the fact that in the event an investor were to exit their investment early, there would be potential unwind costs which should be disclosed in the KID. These should be made clear to retail investors in general terms in the KID (taking into account the fact that the exact amount of such unwind costs would not be known for future dates). The manufacturer could be required to disclose (i) the existence of the potential costs and (ii) the factors that would be used to determine/impact the amount of such cost. Such disclosure should make clear that unwind costs will be deducted from an investor’s returns in the event of early redemption and that in the event of early redemption they may receive back less than they invested and less than the amount they would have received had they held the investment to maturity. Further details on early redemption costs could then be provided by the manufacturer/distributor/whoever agrees to buy back the product from the retail investor at the point in time (if any) a retail investor requests to redeem the investment early in line with IOSCO’s Final Report on Suitability Requirements with Respect to the Distribution of Complex Financial Products (see Footnote 14 below).

• Furthermore, to the extent that exit fees/commissions may be charged by the distributor/broker, the manufacturer may not know (a) the amount of such fees, (b) the factors that may determine whether they apply and indeed (c) whether or not they are contemplated at all. The conditions may also vary between distributors/brokers. It would be helpful for the RTS to clarify that in such circumstances the manufacturer would only be required to include a generic statement that exit fees/commissions might be chargeable by the distributor/broker and that an investor should contact the distributor/broker to establish whether such fees may apply and if so on what basis.

• JAC members consider that details in relation to early redemption should be disclosed in the Section of the KID entitled “How long should I hold it and can I take my money out early?”

• Please also see our response in relation to Question 15 and Question 40 (see Footnote 15 below).

Costs not known by the manufacturer

• As previously indicated, as the KID is a manufacturer document it is key that manufacturers are only required to disclose costs within their knowledge and under their control (as the manufacturer is responsible for the accuracy of the KID). Costs not known by the manufacturer cannot be disclosed in the KID including, for example, advice fees and certain brokerage/distribution costs.

• Licence fees and brokerage costs paid by the manufacturer should not be disclosable as it is difficult to assess the cost per product and these are usually part of an overall pricing package.

• Costs that may vary according to the identity of the investor cannot be disclosed. For example, tax duties are usually the responsibility of an investor and the manufacturer should not be required to make an assessment of the likelihood or the amount of stamp duty or FTT for an individual investor, aside from the generic tax disclosure and warnings in the prospectus or equivalent contractual document.

Footnotes

Footnote 14: See page 13 of FR01/13, January 2013. This flags that when a liquid secondary market for a complex financial product does not exist, the only prices available may be from the intermediary that sold the customer the product and states that the intermediary should “know and disclose ahead of time how these prices will be computed (using models, other markets for similar products etc) and what the price represents (mid-market theoretical value, re-purchase prices etc). The customer should have access to enough information to know that the product is illiquid, including information about the means and range of timing for disinvestment”.

Footnote 15: In our response to Question 40, we state: “Exit fees/costs: If RTS is to be developed in relation to early exit fees/costs the ESAs should bear in mind that such fees may be contingent or vary in accordance with parameters whose values would only be known at the date of exit (or otherwise following the date of the KID) and could be particular to a specific investor. It would not be possible to give a monetary value of such fees/costs in the KID and as such it is important to ensure that the RTS does not require this. In such circumstances we assume that the manufacturer would be required to disclose (i) the existence of the potential fee/cost and (ii) the factors that would be used to determine/impact the amount of such fee/cost. Furthermore, to the extent that exit fees/commissions may be charged by the distributor/broker, the manufacturer may not know (a) the amount of such fees, (b) the factors that may determine whether they apply and indeed (c) whether or not they are contemplated at all. The conditions may also vary between distributors/brokers. It would be helpful for the RTS to clarify that in such circumstances the manufacturer would only be required to include a generic statement that exit fees/commissions might be chargeable by the distributor/broker and that an investor should contact the distributor/broker to establish whether such fees may apply and if so on what basis.”
Pure manufacturer versus retail manufacturer/product provider

• One area where costs could be disproportionate is in the situation where there is uncertainty as to the identity of the “PRIIP manufacturer” and the entity issuing the product has to take on the responsibility for drawing up, reviewing and updating the KID where arguably the true “manufacturer” of the product is another entity. An example of this would be where a distributor approaches a financial institution on a reverse enquiry basis with a bespoke structure for its clients which it wishes to issue off one of the financial institution’s note programmes. As the distributor is the person who undertakes the product design here and not the financial institution, in this scenario we consider that there should be flexibility for the distributor to be classified as the “PRIIP manufacturer” for the purposes of the Regulation as they are in the best position to know the product and the financial institution in this scenario would find it very difficult to prepare a KID and keep this updated.

• We consider that it is important that the obligations of a “PRIIP manufacturer” are triggered by participation in the design of the product. This distinction is one that is made by the UK FCA in the Financial Conduct Authority’s Handbook on “The Responsibilities of Providers and Distributors for the Fair Treatment of Customers” (the RPPD) which distinguishes between product providers and pure manufacturers, with the majority of responsibilities falling on the product provider and not the pure manufacturer.

• Importantly, the RPPD recognises that the respective responsibilities of product providers and distributors will flow from the actual roles or functions undertaken in each transaction and not merely the label given to a firm in respect of a particular transaction. Therefore, for example, if a bank has a limited role in the structuring of a product, the RPPD obligations may fall on the distributor as opposed to the bank. Typically it will be a question of fact whether the bank is a product provider or a pure manufacturer. This will depend in broad terms on who is responsible for the origination of the product in question – in particular the creation, development, design and determination of economic terms. It is increasingly common for distributors to be responsible for this aspect, namely in the reverse-enquiry context where a distributor has already identified a target-market in its client base, and decided upon the exact exposures it wishes to deliver to those clients and is using the bank merely to implement what it has designed.

• Critically, the RPPD also permits product providers and distributors to agree between themselves how to apportion the various responsibilities under the RPPD in many circumstances. In the same way we consider that it is important that each party should be able to agree (contractually or otherwise) its responsibilities consistent with MiFID II/MiFIR such that responsibilities flow from the actual role or functions being undertaken and that firms should be able to decide between themselves which party or parties will take on the “manufacturer” responsibility. For example, where a commissioning distributor is structuring the product for its clients then approaching an issuer with the terms of the product on a “reverse enquiry” basis as described above we think it should be made clear that the parties can agree that the distributor will take on both the distributor and manufacturer responsibilities (as it is the product provider rather than the issuer who is the pure manufacturer).

• In this context it is worth noting uncertainties in the Regulation as to the identity of the PRIIP manufacturer. Article 4(4) of the Regulation defines the “PRIIP manufacturer” as:

“(a) any entity that manufactures PRIIPs;
(b) any entity that makes changes to an existing PRIIP including, but not limited to, altering its risk and reward profile or the costs associated with an investment in a PRIIP”

However, no guidance is provided as to what “manufacture” actually means in the context of the Regulation. Is it the person(s) who undertake(s) product design (the product provider) or the person who creates the product to meet the criteria or designs specified by the distributor/others (the pure manufacturer)?

• It is also not clear how the definition of “PRIIP manufacturer” fits with the references to the “manufacturer” in MiFID II?

• Recital 12 to the Regulation provides some high level guidance in that it states:

“PRIIP manufacturers — such as fund managers, insurance undertakings, credit institutions or investment firms — should draw up the key information document for the PRIIPs that they manufacture, as they are in the best position to know the product”

One would assume that the PRIIP manufacturer should be the issuer (in particular in light of Article 8 which requires the inclusion of a section entitled “What happens if [the name of the PRIIP manufacturer] is unable to pay out?”). However, as indicated above, the absence of a definition of “manufacture” in the Regulation means that uncertainties remain as to the identity of the PRIIP manufacturer for some products (e.g. SPV issuances, derivatives where there is no issuer, or reverse enquiry scenarios where a distributor comes to an entity such as a financial institution with a bespoke structure for its clients). Further clarification in this regard will be important to enable appropriate allocation of responsibilities and an audit trail. Indeed, to the extent certain products issued by SPVs fall within scope of the Regulation, a number of compliance considerations also arise as certain concepts/definitions in the Regulation are difficult to apply sensibly in the context of an SPV issuance and further engagement will be required with regulatory authorities to find a workable way forward.

• It is also not clear whether there is flexibility in the Regulation for parties to agree as between themselves who will take on the role of “PRIIP manufacturer”. Recital 12 of the Regulation states: “where it is impractical for the PRIIP manufacturer to draw up the key information document, it should remain possible for this task to be delegated to others” but this would appear to require a PRIIP manufacturer to actively delegate the task of drawing up the KID to others (and ultimately the PRIIP manufacturer would still appear to remain responsible for the content of the KID and for reviewing and revising it). What if the PRIIP manufacturer has not prepared a KID as it did not intend the product to be distributed to retail investors or because it is a third country entity or even because it considers that another entity is the PRIIP manufacturer for the purposes of the Regulation?

• We note that the 2012 Impact Assessment states (at page 51):

“In regards third countries, the application of requirements relating to who may or may not produce disclosures is of particular significance, given that in some cases a product that would be a PRIP in the EU, but produced in a third country, is sold in the EU. The proposed option relating to the preparation of information seeks to address this circumstance; distributors in these cases might prepare the KIID.”

Page 87 of 2012 Impact Assessment goes on to state:

“An additional complexity would be where there is a product that would be a PRIP in the EU that has originated in a third country not subject to EU rules; in these cases, in principle it would seem reasonable that a distributor might take responsibility for preparing a KIID where the product originator – not directly subject to EU rules – has failed to do so. In addition, distributors might be permitted to prepare a KIID (to bring a product to the retail market) where the product producer has not done so (not intending the product for retail distribution themselves), so long as they take full business responsibility for this (see Footnote 20 below)”.

Our view however is that if the manufacturer (and as noted clarification would be needed as to who the manufacturer is particularly in reverse enquiry situations where we consider that a ) has not prepared a KID, or effectively delegated responsibility for the task of preparing a KID to others in accordance with the Regulation, the distributor should simply be prevented from making the product available to EEA retail investors.

Application of the PRIIPs Regulation to products manufactured prior to 31 December 2016 and need for a transitional period for all products

• The Regulation states that “before a PRIIP is made available to retail investors, the PRIIP manufacturer shall draw up for that product a key information document…” (Article 5), but does not explicitly state whether this obligation only applies to products issued after the date of application of the Regulation (31 December 2016) or whether it will also apply to products manufactured prior to 31 December 2016 (Pre-Existing Products) if made available to retail investors via secondary market trading after that date. It is important that the market receives clarification on this as soon as possible as to the extent the Regulation is intended to apply to Pre-Existing Products, market participants will need time to implement appropriate operational and documentation controls to ensure they do not fall foul of the Regulation in relation to such products.

• There would also be significant costs involved for manufacturers in complying with the Regulation in relation to Pre-Existing Products. In general such products would have been manufactured without knowledge of the requirements of the PRIIPs Regulation and the costs of complying with this Regulation would not have been taken into account when pricing the product. The costs of complying with the Regulation also seem disproportionate particularly where the product is soon to mature or where there is no active secondary market (or a low volume of secondary market sales). In order to avoid any inadvertent breach of the Regulation, some manufacturers may simply redeem issues early which would not be in the interests of retail investors. As a result we consider that it should be made clear that Pre-Existing Products, should not be subject to the PRIIPs Regulation.

• If this approach is unacceptable, we consider that at a minimum Pre-Existing Products:

 - should not be subject to the Regulation to the extent they are not actively sold anymore (Article 13 of the Regulation);
 - with regard to Pre-Existing Products that are actively traded on secondary markets, such Pre-Existing Products should only be subject to the Regulation to the extent they have a maturity date falling after 31 December 2018; and
 - with regard to Pre-Existing Products that are actively traded and have a maturity date falling after 31 December 2018, a sufficient transitional period should be allowed for manufacturers to comply with the Regulation in relation to future sales of such Pre-Existing Products.

• We also consider that a transitional period will be needed for products that are not Pre-Existing Products given the fact that the draft RTS relating to the content of the KID (Article 8 of the Regulation) are only expected to be submitted to the Commission by February 2016 who can then take up to three months to adopt them (including with amendments). This will then be followed by another delay for the Parliament and the Council to raise any objection, which can take up to three months. This means that the definitive rules will most likely not be published until April-September 2016 and manufacturers would have between 3-8 months for IT developments. Market participants are concerned that this is not sufficient time given that much information which must be included in the KID requires prior data processing and modelling and given the sheer volume of KIDs that manufacturers may be tasked with producing.

Product manufacturers subject to both the PRIIPs Regulation and the Prospectus Directive and/or national short-form disclosure regimes

• As indicated in our response to Question 1 there is considerable overlap between the prospectus summary and the KID. It would be helpful for the European authorities to be cognisant of this considerable overlap when making recommendations for regulatory reform in the future as currently a product manufacturer subject to both the Prospectus Directive and PRIIPs will arguably incur disproportionate costs as a result of the need to prepare both the prospectus summary and the KID. A PRIIPs product manufacturer not subject to the Prospectus Directive regime would therefore have an advantage over manufacturers subject to both regimes.

• Similarly, as indicated in our response to Question 1, it is important that the KID supersedes and replaces any national short form disclosure regimes otherwise the purpose of the Regulation to harmonise regulatory regimes across the European Union would be defeated, resulting in regulatory arbitrage, divergences throughout Europe and introducing barriers to the main purpose of the Regulation: to create a level playing field for manufacturers, distributors and retail investors. There would also be significant cost implications to the extent a manufacturer has to prepare a prospectus, prospectus summary, KID and national short form disclosure document which we consider would be disproportionate given the purpose of the Regulation to harmonise regulatory regimes across the European Union.

Footnotes

Footnote 20: This reflects the extent to which the allocation of responsibilities can be a tool which controls what PRIPs can in fact be sold to retail customers.
The main challenges to achieving a level-playing field in cost disclosures from the perspective of JAC members include the following:

• There are different costs associated with different products (for example, some products are open ended and charge investors an annual management fee whilst others have no on-going fees and the investor pays an implicit fee embedded in the price of the product). It will be challenging to provide consistency and comparability across such different products.

• If it is necessary for embedded product costs to be disclosed it will be challenging for retail investors to understand the difference between costs the investor might bear on investment/during the life of the PRIIP, compared with embedded product “costs” which will not be deducted from the returns investors receive. Other challenges in relation to disclosing embedded costs include the fact that such costs are inherently subjective and involve costs which relate to a portfolio of trades rather than a single product. See our response to Question 17 for reasons as to why we do not think it is helpful to include details of such embedded costs in the KID.

To assist in achieving a level playing field on cost disclosure JAC members note:

• It will be important to ensure that all manufacturers disclose the same types of costs using the same methodology (for example, currently methodologies used in relation to funds vary considerably to those used for structured products and methodologies also vary as between manufacturers).

• It will be important to ensure that manufacturers are not required to disclose costs information in the KID which relates to third party costs or other costs to which they do not have access (e.g. broker commissions, stamp duties, transaction taxes, foreign exchange costs, advice fees, distribution fees from third parties among others). Typically manufacturers will not have access to such costs and to the extent they do such information will be aggregated across a number of trades/products rather than quantifiable with reference to an individual product. It would also contravene the overarching purpose of the KID to provide comparability across products if distribution fees are included for some products but not others. Accordingly it will be important for the KID to include generic language that discloses the possibility of additional entry/exit/management/advisor costs being charged by the distributor/third parties.

• Where costs are not known in full up-front, the methodology by which costs are estimated by manufacturers should be as consistent as possible (e.g. maximum costs could be included).
• Please see our response to Question 17 in relation to implicit costs and embedded costs explaining why JAC members do not consider that it should be necessary or beneficial to retail investors to be provided with disclosure regarding these costs.

• To the extent regulators conclude that embedded costs must be disclosed in the KID, please see our response to Question 17 for our suggestion as to how this may best be achieved. As previously indicated, a KID should only disclose costs within the manufacturer’s knowledge and control and should not require any further break-down of embedded costs beyond manufacturing costs.
JAC members do not agree that RIY or a similar calculation method could be used to prepare a “total aggregate cost” figures for a PRIIP for reasons including the following:

• Such a calculation method may work for products where principal is not at risk but not for structured products.

• Trying to compare costs to a potential yield is not helpful as fees do not necessarily map to yield.

• RIY is difficult to compute and is likely to be difficult for retail investors to understand.

• For defined return products, the key is the defined return. It is unclear why embedded costs built into the pay-off are relevant for the purposes of key information as these will not reduce the amount received by investors.

• Some JAC members have expressed a preference for the total expense ratio instead of the RIY for reasons including the following:

 - It is already used for UCITS, hence known by a portion of investors and allowing comparability between UCITS and other types of PRIIPs.

 - It is simple to compute and understand.

 - It includes running costs, which is not the case for the fair value approach.

 - It is not dependent on a model and assumptions peculiar to each manufacturer.
• Please see our response to Question 20
• JAC members do not agree that implicit or explicit growth rates should be assumed for the purpose of estimating ‘total aggregate costs’ for reasons including the following:

 - in the context of structured products, in general, costs are not dependent on growth rates, unlike, for example, open ended products with variable fees which are affected by the length of time the product is held.

 - it may be misleading to show upfront costs in such a way. If, for example, the market rises by 10% per year, the fee will look low.

 - it would be necessary to include various assumptions which would be inappropriate for a 3 page KID.
• JAC members are of the view that implicit portfolio transaction costs cannot be effectively incorporated as such costs will typically be on an aggregate basis across a number of different products (as opposed to an individual PRIIP basis) and certain techniques may have changing fees/amounts over time which it would not be reasonable/practicable to take into account.

• It is also important to note that for defined return products, portfolio transaction costs will not affect the return an investor receives. This is a key difference between UCITS (where portfolio transaction costs do negatively impact an investor’s return) and structured products.
• Given the range of different PRIIPs, it will challenging to ensure costs disclosure follows a standard format without being confusing/difficult for retail investors to understand (e.g. for some products there may be multiple “zero” entries as certain costs will not apply)

• As indicated, a key challenge will be to distinguish between costs which are deducted from an investor’s returns and embedded costs, already included in the price of the product in a manner that is clear and unambiguous for retail investors

• For a fixed maturity PRIIP, presenting costs information relating to a standard holding period, which does not correspond to the tenor of the PRIIP, would be misleading for retail investors
General Comments:

• Some JAC members consider that for products which have a defined maturity (and most retail structured products do), the only sensible scenario is where the product is held to maturity and the fees come out upfront. If the product is open-ended it may be helpful to look at the impact of costs on different years of the investment and show the compounding effect.

Members have expressed the following views on the presentational examples:

• Option 1: this does not provide sufficient information (e.g. what is the exact amount of costs in monetary/percentage terms among other things)

• Option 2: this has the benefit of being simple to understand and concise, however, the majority of JAC members were not in favour of indicating costs on an annualised basis on the grounds that this form of presentation is artificial for structured retail products with upfront/embedded costs and no annual/ongoing fees and may hence be detrimental to an investor’s understanding of the cost structure of those products. The cost structure can be an important factor in that investment decision and hence it is important that the KID discloses costs in a way that will not confuse or mislead investors. Notwithstanding this, certain members felt that for certain products (in light of the fee structure for those, or directly comparable, products) it may be helpful to investors also to include annualised costs figures (see Footnote 16 below) and that, consequently, it will be important for manufacturers to have the discretion to do so if they consider that this form of presentation would assist investor comprehension.

• Option 3: APR is a concept applicable to loans. This does not appear suitable for other products (e.g. equity linked products)

• Option 4: this enables investors to see initial costs, ongoing costs and exit costs as both percentages and monetary figures. We do not consider it helpful to include the information on fair value (see our response to Question 17 for an explanation as to why we do not consider this information helpful). It would also be helpful to include total costs in percentage and monetary terms for the full product tenor

• Option 5: this appears both concise and seemingly easy to understand. However, in relation to presenting costs on an annualized basis – whilst manufacturers should have the discretion to determine whether or not costs should be presented on an annualized basis, the majority of JAC members were not in favour of indicating costs on an annualized basis on the grounds that this form of presentation is artificial for structured retail products with upfront/embedded costs and no annual/ongoing fees. We therefore submit that the annual percentage rate should be omitted for such products.

• Option 6: the number of assumptions required to produce this information would also need to be disclosed which may not be feasible in a 3 page KID

• Option 7: this does not show aggregate costs and the figures in percentage terms. A graph also assumes a specific performance scenario and we think costs disclosures and performance disclosure should be kept separate. The depiction of gross costs as well as net costs on the same graph may also be complex for retail investors to understand

• Options 8 and 9: these combine costs with performance – as indicated we consider it preferable for details of performance to be contained in the section of the KID entitled “What are the risks and what could I get in return”

• Option 10: this may be difficult for retail investors to understand

Option 5 therefore seems the most appropriate approach subject to the manufacturer having the discretion to omit the annual percentage rate for products which do not incur annual/ongoing fees.

Footnotes

Footnote 16: For products with no annual/ongoing fees, annualised costs could be computed by reference to the total costs divided by the tenor of the product.
• No, although for some products a number of these cost types would not be applicable
• JAC members are unanimous in their view that these risks should not be aggregated in any circumstances. Aggregation of these fundamentally different types of risk without explanation of the different contributions of each type of risk is likely to have little meaning/be misleading to investors. To the extent an explanation of the different contributions of each type of risk is included, this may not be easily understood by retail investors who may well simply focus on the overall risk indicator rather than the explanation.

• Market, credit and liquidity risks are independent and should be looked at separately in order for an investor to understand the product. Different types of risk may also vary in significance for different investors (e.g. if any investor has no intention/need to exit the product early liquidity risk will be of little relevance) and where risks are aggregated this would prevent them from isolating less relevant risks. Investors might also sensibly want to include assets with different levels and types of risk and an integrated credit/market/liquidity risk indicator may be misleading to them. For retail structured products that are typically buy-to-hold products, grouping liquidity risk with credit risk and market risk may distort an investor’s impression as to the overall risk of a product (for example, a product with an otherwise low risk profile but which has a fixed term and limited secondary market would appear far more “risky” than is in fact the case).

• Please see our response to Question 12 for further details on what should be shown in relation to each of credit risk, market risk and liquidity risk in the KID.

• As mentioned in our response to Questions 3 and 4, we consider that liquidity risk should be explained as a narrative or in a simple table format. Our preferred approach would be for liquidity risk to be disclosed in response to the Question: “How long should I hold it and can I take money out early” with only a cross-reference to the section entitled: “What are the risks and what could I get in return” if this is deemed necessary.

• As mentioned in our response to Question 4, we consider that credit risk should also be disclosed in narrative form in response to the question “What happens if [the name of the PRIIP manufacturer] is unable to pay out” or (if it is considered important that this is included in the section entitled “What are the risks and what could I get in return?”) immediately following the indication of market risk. A standard table should illustrate credit risk based on the credit rating of the manufacturer (or if applicable, guarantor or swap counterparty or any collateral securing the product) on a scale of available credit ratings. To the extent there are no such credit ratings, the credit risk of the PRIIP manufacturer could be explained in narrative form.
• Cumulative costs should be shown as a percentage of the nominal amount invested
• JAC members consider that it is imperative that information about costs contained in the KID is limited to key information and is simple, engaging and understandable. The costs listed in Table 12 are, according to the Discussion Paper, not a “complete list of costs” rather just those costs that may be difficult to assess. However, JAC members consider that this incomplete list of costs of itself would almost certainly be overly detailed and complex for retail investors. Retail investors will be primarily concerned with the total amount of costs they will be charged upfront and during the life of the investment. The breakdown of such costs is not likely to be of interest or relevance to the investor.

• Please also see our comments in response to Questions 16-18 in relation to costs such as implicit costs, dividends, early redemption costs and taxes and the fact that a KID should only be used to disclose costs within the manufacturer’s knowledge and control.

• It is also important to note that it would be unrealistic to require the manufacturer to be able to quantify costs on an individual PRIIP basis where such costs have been incurred on a portfolio basis.
PRIIPs Manufacturer contact details

We agree that standardisation would be helpful. In respect of the PRIIPs Manufacturer we believe that a website link would be the most efficient presentation method and should be sufficient (whilst noting that if the manufacturer does not have a website (for example, if it is an SPV) it would be important for flexibility to be preserved for the website of another appropriate transaction party to be used or other information delivery routes). We would suggest that consumers who do not themselves use the internet would almost always have contact with someone who would be able to use the internet to access the information on their behalf. If the ESAs ultimately conclude that a non-internet option should also be included, we would suggest that the RTS requires either a postal address or telephone number but not both given the space available. Of the two, we believe a postal address would be preferable as there may not always be a single telephone number that could be specified. We would also suggest that manufacturers are permitted (but not obliged) to provide a contact email address.

Competent authority details

We agree that the name of the authority should be provided. Given the limited space available we would query the value of also requiring a web link. This is on the basis that with the name of the authority, its website could be easily identified through a search engine. The approach to be taken for a non-regulated entity (e.g. a corporate issuer who would not have a competent authority in this sense) should also be made clear.

ISINs and similar identifiers

We agree that manufacturers should have the option to include ISINs or similar identifiers where these are available at the date of publication of the KID. We do not, however, believe that their inclusion should be mandatory as this may cause unnecessary complication if for any reason an identifier is not available or the manufacturer believes that its inclusion is not appropriate.
Further specific guidance is needed. This guidance should be sufficiently specific to allow manufacturers to have confidence when omitting the warning that they fall within this guidance. Clarity is particularly important as inconsistent approaches between manufacturers or competent authorities have the potential to distort the market (with different approaches being taken by different manufacturers/competent authorities to similar products). Further, if manufacturers cannot omit the alert with confidence this is likely to result in the alert being used too frequently which will diminish its impact. We presume that the intention is that the alert should only apply to products which are particularly complex.

Assets not commonly invested in

We agree that Article 50 of the UCITS Directive is a sensible starting point for compiling a list of assets which are commonly invested in by retail investors. However, this list will need to be adapted to reflect the fact that neither investors nor the manufacturer will necessarily have to hold the underlying. A wider definition of indices will also be essential to capture normal market indices.

Calculation mechanisms for final return

As above, we would welcome a far more detailed analysis as to exactly which products would and would not fall within this category. This will ensure conformity in the market and that the alert is only used for those products for which it is specifically intended. The approach currently suggested in Recital 18 is very broad and has the potential to catch many structured products which, even in their most simple form, use a number of different mechanisms to calculate the final return of the investment. We disagree with the view that using a number of different mechanisms necessarily increases the risk of misunderstanding on the part of a retail investor. We would argue that the fundamental consideration should be an investor’s capacity to understand the risk and return proposition of the investment (and how this relates to the investor’s market view) and not necessarily the detail of how this may be delivered. If an investor is readily able to do this, we do not think they should be dissuaded from investing in the product and a product should not be considered sufficiently complex to warrant a risk warning simply because it uses derivatives or sophisticated mechanisms to deliver its performance.

Behavioural biases

As above, we would welcome a far more detailed analysis as to exactly which products would and would not fall within this category. We would agree with the basis suggested for this in the Discussion Paper.
Subject to clarifying the product scope of the Regulation, we agree that the fundamental starting point for determining and disclosing type should be legal form. In the context of securities, differentiation should be permitted between different types of security (e.g. notes, certificates and warrants). This should in each case be clarified in the RTS. Guidance will be important to ensure consistency of classification between manufacturers.

• We do not believe that attempting to introduce a more detailed labelling system based on product features or economic exposures would add value for consumers. This is on the basis that:

 - Given the broad range of products covered by the regime it is likely that there would need to be an extensive list of labels. It is questionable whether investors would be willing to engage with such a complex system. Adopting a simplified system could equally be counterproductive as it would result in products that may be fundamentally different in certain ways being classified under the same broad label. This is unlikely to assist investor comprehension.

 - Some concepts or payout types can be delivered in a number of different ways, the economic effect varying depending on which methodology is used. This would mean that, if even a more granular system of labelling is adopted, some labels would likely cover a range of ultimate economic outcomes which has the potential to confuse investors.

 - Existing labelling systems use some technical terms which may not always be intuitive to less sophisticated investors. The EUSIPA model is only one of a number of current labelling systems.

 - Different market participants have different interpretations of/approaches to commonly recognised “payouts”. We believe that this variety is good for the market (as it offers consumers choice) but it is likely to mean that it may be a challenge to agree a common definition for many payouts.

 - Products may combine a number of features and giving a single product a number of labels may be confusing for investors.

 - The system would need to be subject to ongoing ESA review to ensure that it is updated to reflect product innovation.

• If a labelling system based on product features or economic exposures is introduced:

 - The system would need to be compiled in full consultation with the industry (and other interested parties) and should not be introduced unless consensus on the typology can be reached.

 - Manufacturers should not be obliged to label a particular product if they do not believe that it fits into any of the labels/types provided by the ESAs. Where a manufacturer does not believe that a product fits into any of the labels provided it would not be reasonable to require the manufacturer to assume the potential misselling liability associated with applying a label to a product which it does not think accurately reflects that product’s features. Where no appropriate label exists, this should not prevent a KID being produced rather the manufacturer should simply be permitted not to include a label.

 - As above, the ESAs would need to ensure that the typology is subject to their ongoing review to ensure that it can be updated to reflect product innovation.
As part of the process of determining a typology for PRIIPs it will also be essential to provide further clarity as to the products for which a KID must be produced (being products which are within the PRIIPs definition and not excluded under the Regulation). In addition, we consider that the required clarifications must be fully consistent with the Level 1 text of the Regulation and not seek to modify this (directly or indirectly).

The Discussion Paper includes a table (Table 2 – Summary of scope) containing a non-exhaustive assessment of which products will be in/out of scope of the Regulation. Arguably however this table raises more questions than it gives answers, and sections of it do not fully reflect the Level 1 text. For example:

• the table states “derivatives of all types” are within scope if sold to retail investors. We have the following comments in relation to this statement:

 - there is no indication in the Level 1 text that derivatives are within scope of the Regulation. Indeed, extending the scope of the Regulation to all derivatives is contradictory to the Level 1 text. The definition of PRIIP contained in the Level 1 text requires an “investment” where the “amount repayable to the retail investor is subject to fluctuations…”. To the extent an instrument is not an “investment” or does not involve an “amount repayable” which fluctuates as described in the definition, the relevant instrument is not within scope. Unlike deposits, or structured securities a typical derivative transaction does not involve any repayment of principal to the retail investor. Derivatives are bilateral agreements involving two-way payments between the parties and generally there is no principal investment at all.

 - in the context of derivatives used for commercial hedging, there is no “investment” as required by the Level 1 text. The intended use/purpose of such a derivative used for is not “investment” but commercial hedging. The distinction between derivatives used for speculative/investment purposes and those used for commercial hedging is one that was recognised in EMIR and we submit that it should also be recognised in the context of the PRIIPs Regulation as the Level 1 text mandates that a product will only be within scope of the Regulation to the extent that it is an “investment”.

• the table states “structured securities (convertible bonds and other securities that embed a derivative within this category) and other ‘structured products’” are in scope. Again this purported extension of the Regulation to convertible bonds appears to go beyond the Level 1 text. As indicated, the PRIP definition in the Regulation indicates that investments are only in scope where “the amount repayable to the retail investor is subject to fluctuations because of exposure to reference values or to the performance of one or more assets which are not directly purchased by the retail investor”. In the context of a convertible bond, the amount repayable to the investor (i.e. the principal return) will not fluctuate as described in the definition. The retail investor will receive a stated principle return if conversion rights are not exercised and if they are they will receive shares held directly. Shares held directly are stated to be outside the scope of the Regulation according to Recital 7 of the Level 1 text.

• the table states “SPVs (instruments issued by)” are in scope. However, as indicated, the PRIP definition in the Regulation clearly indicates that, in all cases, investments are only within scope where “the amount repayable to the retail investor is subject to fluctuations because of exposure to reference values or to the performance of one or more assets which are not directly purchased by the retail investor”. The Level 1 text makes it clear that this defining feature applies regardless of the nature of the issuing entity (including whether or not the issuer is a securitisation special purpose entity). Accordingly, we respectfully submit that it is not correct to state that a KID must be produced in respect of all instruments issued by SPVs and/or to suggest that this is the effect of the reference to instruments issued by certain special purpose vehicles and entities in the definition. To the extent that an instrument issued by an SPV issuer does not involve an amount repayable which fluctuates as described in the definition, the relevant instrument is not within scope. We note that the distinction drawn here is not a theoretical one. Instruments issued by certain SPV issuers do not involve a relevant repayable amount. For example, SPVs issuing debt instruments with a stated principal return do not involve a fluctuating repayable amount as contemplated by the definition and it is arguable that any extinguishment of rights through the operation of any applicable limited recourse provisions is a separate matter.

• Some comments in a European Commission press release last year (see Footnote 17 below), suggested that subordinated or perpetual notes may have been intended to fall within scope of the Regulation. However, neither perpetual nor subordinated notes fall within the definition of “PRIP” in the Level 1 text and therefore must fall outside the scope of the Regulation. The reason for this is that again they do not involve an amount repayable that “is subject to fluctuations because of exposure to reference values or to the performance of one or more assets which are not directly purchased by the retail investor”.

• The table is non-exhaustive and does not consider all products where guidance may be required in relation to product scope.

Footnotes

Footnote 17: A press release (http://europa.eu/rapid/press-release_STATEMENT-14-122_en.htm) published by the European Commission on 15 April 2014 states that: “The new rules are a response to a myriad of problems that retail investors faced in the past. For example, a consumer Ombudsman in one Member State recently found 12-year subordinated notes, carrying significant losses in case the issuer defaults, being sold to the very elderly. Risks of exiting these investments were not fully explained. In another example, a study for a consumer affairs ministry in one Member State suggested that up to 50-80% of consumers wanted to get out of long-term investments prematurely, indicating investments made were not suited to their needs.”
See our response to Question 32. Given the space available in the document we believe that the classifications should be limited to legal form. For example, we do not believe that the class of the underlying asset(s) (e.g. equity, index, commodity etc) should be included as a classification under “Type” as this will be disclosed as part of the “Description of the Underlying”. We do not believe that replicating such information when disclosing the product “Type” would be beneficial to consumers.
Guidance

• We agree that more detailed guidance on the topics suggested in the Discussion Paper would be helpful. While we fully agree with the aims of the regime, the KID poses significant potential disclosure risk for manufacturers and any assistance in managing this is welcome. Guidance should also assist in ensuring conformity of approach between manufacturers and jurisdictions and without it there is the potential for significant divergence.

• In particular it would be helpful for guidance to be provided as to what information should and should not be considered “key”. It is important, however, that this guidance is conceptual such that the manufacturer retains ultimate discretion to determine what the key elements of a product are and how best to disclose them.

• Similarly guidance on how the product pay-off can be described in a summary format (including the elements a manufacturer may safely omit) will also be valuable as this will be a particularly challenging aspect of preparing a KID and one which entails significant legal/disclosure risk for the manufacturer.

• It will also be essential that the approach mandated is realisable in the space available. In light of the space limitations, it may be that some compromises will need to be made as compared to the disclosure standards that might otherwise be required if space was not at a premium.

• We agree that clear guidance on what should be considered “direct” and what “indirect” exposure is important. The ESAs should also note in formulating the RTS that whether exposure is direct or indirect may not be static and/or may be partial (i.e. the manufacturer may change its “hedge” during the term of the product or may only hedge in part). Further, the interpretation set out in the Discussion Paper does not specifically contemplate the scenario in which the manufacturer does not hedge at all (i.e. it takes the risk on its own book and does not purchase the asset(s) or other instruments to hedge the risk) and it will be important to clarify which category this would fall into and how it might be disclosed.

Prescribed statements

• Prescribed statements may also assist in ensuring conformity, however, given the broad range of products covered it may be difficult to formulate statements that are meaningful and insightful in all contexts. We would need to see specific examples of such statements to pass comment.

• Given the limited space available, prescribed statements will only be helpful if they add significant value to the disclosure provided. It will detract from the document if space that could otherwise be used for specific and insightful disclosure is instead taken up by more general prescribed statements.
• We agree that allowing cross-references to be made to more detailed information (be that within or outside of the KID) could also be of significant assistance to consumers. As well as a helpful summary of the product, the KID will have further value if it is able to assist consumers in navigating the other disclosure available to them.

• We believe that a fundamental aim should be to keep the description of the objective as short and simple as possible. To assist this, and as raised in previous representations, we believe that it would be helpful to include a “Product data” section in the “What is this product?” section that includes key data on the product (e.g. in a securities context, the issue price, nominal amount, issue date, valuation dates, payment dates and brief explanations of any key terms, in each case in line item form). We believe that presenting the data in this format is more user-friendly for consumers. It will be both more engaging than a longhand narrative format and will also facilitate comparison between products (line items offering a more readily discernible point of comparison than where data must first be extracted from a narrative description). By providing this more functional information in a “Product data” sub-section, this will allow the “Objectives” disclosure (and indeed the other elements of the “What is the product?” section) to be more concise, focusing solely on the substance of the investment proposition and relying on the “Product data” sub-section to complete the detail. We believe that this approach is consistent with the Level 1 text and will provide consumers with significant assistance in engaging with the KID. We would be happy to provide an example if that would be of assistance.
• We agree that further clarification would be helpful and that the avoidance of duplication (in particular with the disclosure contemplated in Article 8(3)(g)) and ensuring that the requirements are mindful of the space available should be a focus.

• Given the volume of other disclosure requirements contemplated in the Regulation that relate to other more essential product features, realistically it may be that there will only be space to disclose the ability to bear loss and investment horizon. These would in any case appear the two most pertinent considerations.

• We would also suggest that if the product targets specific developments in the underlying, to avoid duplication, it would be more appropriate to disclose this in the “Objectives” section.

• If further disclosure is to be required (beyond the ability to bear loss and investment horizon), we agree that consistency in approach with other regimes (in particular MiFID II (see Footnote 18 below)) is important to aid clarity and limit potential investor confusion, however, such additional requirements must be considered in light of the space available and the role and knowledge of the manufacturer and specifically:

(i) Detailed disclosure will not be feasible in the space available in the KID and providing heavily abridged disclosure may be of negligible benefit (or indeed misleading).

(ii) Given that a number of elements pertinent to determining the suitability of the product to an investor will be covered elsewhere in the KID, it is important that these requirements do not result in the duplication of information in this section.

(iii) In line with ESMA’s December 2014 Technical Advice on MiFID II in relation to the manufacturer’s identification of the “target market”, a manufacturer will typically not know who the end investors will be and must therefore identify its target market “on a theoretical basis. This means they would determine the product’s compatibility with a target market without specific knowledge of individual clients but with a more general view of how the specificities of the product would be compatible for certain types of investors, considering their knowledge of the financial markets and their past experiences with similar products and investors. Distributors should use the manufacturer’s more general target market assessment together with existing information on their clients to identify their own target market for a product, that is the group of clients to whom they are effectively going to offer the product through the provision of their services” (see Footnote 19 below). Consumer types for the purposes of the PRIIPs Regulation should be identified in the same way as the manufacturer should identify its target market under MiFID II.

Footnotes

Footnote 18: As indicated in our response to Question 1, the MiFID II requirement to define a target market (Article 24(2)) and to indicate if the product is aimed at retail or professional clients (Article 24(4)(b)) should be satisfied by the requirement in Article 8(3)(c)(iii) of the PRIIPs Regulation to describe: “the type of retail investor to whom the PRIIP is intended to be marketed, in particular in terms of the ability to bear investment loss and the investment horizon.”

Footnote 19: See paragraph 6 on page 52 of ESMA’s December 2014 Technical Advice on MiFID II.
No comments
No comments
• We agree with the assessment presented in the Discussion Paper.

• In relation to the disclosure regarding relevant compensation or guarantee schemes, any guidance prepared should consider carefully the likely relevance of such schemes to each type of PRIIP and only require the manufacturer to disclose information that is specific, concise and knowable. Indeed it is arguable that either:

(i) limiting the disclosure required to providing the website of the relevant scheme; or

(ii) the ESAs preparing specific guidance mandating standard wording to be included in respect of each scheme that may be applicable (i.e. a different form for each scheme),

would be the best way to ensure consistency and assist manufacturers in managing disclosure risk. The RTS could also provide standardised wording to include where no guarantee scheme applies.

• We note that it would not be feasible to include any discussion of the insolvency regime that applies to the manufacturer in this section and this would go beyond what is required by the Level 1 text in any case.

• We note further that, with the exception of structured deposits (which may be covered by national deposit guarantee schemes), we are unsure which such schemes would apply to PRIIPs. Deposit guarantee schemes would not apply.
• We agree with the assessment presented in the Discussion Paper.

• Holding periods:

 - It will be important for the RTS to clarify that the manufacturer is not expected to recommend a holding period to investors. The appropriate holding period for any particular investor may be specific to their individual circumstances and would need to be assessed by an advisor in light of those circumstances and their investment goals. This cannot be addressed in the KID which is not investor specific. As such, the RTS should clarify that the term “recommended” holding period should be interpreted only to require the manufacturer to disclose if the product is specifically designed to be held only for a certain period (e.g. it is a hold-to-maturity product).

 - For open-ended investment or exchange-traded products, the RTS should provide that the recommended holding period can be disclosed generically.

• Exit fees/costs:

 - If RTS is to be developed in relation to early exit fees/costs the ESAs should bear in mind that such fees may be contingent or vary in accordance with parameters whose values would only be known at the date of exit (or otherwise following the date of the KID) and could be particular to a specific investor. It would not be possible to give a monetary value for such fees/costs in the KID and as such it is important to ensure that the RTS does not require this. In such circumstances we assume that the manufacturer would be required to disclose (i) the existence of the potential fee/cost and (ii) the factors that would be used to determine/impact the amount of such fee/cost.

 - To the extent that exit fees/commissions may be charged by the distributor/broker, the manufacturer may not know (a) the amount of such fees, (b) the factors that may determine whether they apply and indeed (c) whether or not they are contemplated at all. The conditions may also vary between distributors/brokers. It would be helpful for the RTS to clarify that in such circumstances the manufacturer would only be required to include a generic statement that exit fees/commissions might be chargeable by the distributor/broker and that an investor should contact the distributor/broker to establish whether such fees may apply and if so on what basis.
We believe that it is important that the RTS should clarify the following points:

(i) that (a) this requirement can be satisfied by simply providing consumers with a contact point to address complaints to and (b) specifically that a more detailed description of any complaints process is not required. Providing further details on the complaints process in the KID would not seem to us an efficient use of the space available and we do not believe that this is the intention of the Level 1 text in any case. We agree that a web link would appear to be the most appropriate medium for the manufacturer to disclose, whilst noting that if the manufacturer does not have a website (for example, if it is an SPV) it would be important for flexibility to be preserved for the website of another appropriate transaction party to be used or other information delivery routes. For the reasons stated in our response to Question 30 we do not believe that a requirement to disclose further contact information would be of sufficient benefit to consumers to justify the space it would take up. In any event, if there is a concern regarding consumers who do not have internet access and this has already been addressed in relation to Article 8(3)(a), we would argue that the address or telephone contact details disclosed for the purposes of Article 8(3)(a) should be sufficient to address this.

(ii) in relation to disclosure regarding the distributor (i.e. the person “advising on, or selling, the product”), we would ask that:

 - the RTS clarifies that the manufacturer should only be required to include a simple statement that “If you wish to complain about the person who has advised you on, or sold you, this product you should contact them directly”. For products with multiple distributors (or for which further distributors or sub-distributors may be subsequently appointed) it would place an unacceptable demand on the limited space in the KID to require disclosure on each distributor. Indeed, the manufacturer may not know the identity of one or more of the distributors or sub-distributors when preparing the KID; and

 - the RTS nonetheless permits the manufacturer the option, but not the obligation, to include specific disclosure on the distributor where this is known.

We believe that the use of the specific wording suggested above is preferable to more generic wording. There would be little or no benefit to investors in including generic information about making complaints to distributors generally.
• We agree that in many cases a link to a webpage may be the most efficient way to provide this information.

• We note, however, that there may be other arrangements in place to provide such information to consumers (e.g. in the context of a securities issuance making such documents available at the office of the principal agent or publishing them on the website of a relevant stock exchange). The RTS should allow the manufacturer to disclose whichever of those media as it sees fit. It should also provide that as long as one of those options is disclosed that should be sufficient to satisfy the requirement under Article 8(3)(i).

• If the manufacturer does not have a website (for example, if it is an SPV) it would be important for flexibility to be preserved for the webpage of another appropriate transaction party to be used or other information delivery routes.

• Given the space limitations, the RTS should make clear that if the webpage has already been disclosed elsewhere in the KID, the address should not need to be replicated in full in this section.
No comments
No comments
No comments
No comments
No comments
No comments
• It is key that the obligation to review/revise the KID is aligned with MiFID II obligations to ensure information provided to retail investors is “up-to-date” and to provide updated costs information (see our response to Question 1 for further details in relation to this).

• The requirement to review, revise and update the KID places a significant and costly regulatory burden on manufacturers. Experience with short form disclosure documents in other jurisdictions (e.g. Italy and Germany) demonstrates that it is key that the requirement to update the KID therefore contains clear and reasonable limits.

• As the KID is a pre-contractual document prepared by the manufacturer, it should not substitute the duties of the advisor/distributor to provide information to investors throughout the life of the product. JAC members are of the view that the KID should not be used to inform investors of any changes occurring after the product has been sold. The KID should only be updated to reflect changes likely to significantly impact investment decisions relating to future sales.

• The KID should only be updated if there has been a significant new factor, material mistake or inaccuracy in line with the requirements in relation to supplementing a Prospectus (pursuant to Article 16 of the Prospectus Directive) which would significantly impact investment decisions relating to future sales. We consider that it would make sense for the requirement to update a Prospectus and to supplement this to be broadly aligned with the requirement to review and update a KID. The reason for this is that the KID must not be misleading or inaccurate or inconsistent with the legally binding pre-contractual and contractual documents. If the requirements to update the KID and requirements to update/supplement the related Prospectus are not broadly aligned this could lead to inconsistencies between the Prospectus, summary and the KID.

• For products that are continuously marketed (for example, where there is a lengthy open offer period to retail investors or where the manufacturer makes active two way pricing in the product available to retail investors), the KID should be updated wherever there is a significant new factor, material mistake or inaccuracy within a specified number of Business Days. The manufacturer will require sufficient time to update the KID and it will not be possible for real-time/same day updates to be available.

• Also it should only be necessary to update those sections of the KID which have materially changed as opposed to the whole KID. It should be possible for such updates to be by way of a (one-page) supplement to the KID. Otherwise, manufacturers are likely to find the volume of KIDs that require updating and the amount of work involved in updating each such KID, overwhelming. We appreciate that this update method (using a one-page supplement) could raise issues of comprehensibility however this method of updating is one that is used for Base Prospectuses in the context of the Prospectus Directive regime. It would also make sense to have the option of using a (one-page) supplement to update a KID as in some circumstances this may be more comprehensible, for example, where a KID has already been provided for a live offer and new details need to be disclosed.

• In relation to updates to the risk indicator, JAC members advocate that for products that are continuously marketed (for example, where there is a lengthy open offer period to retail investors or where the manufacturer makes active two way pricing in the product available to retail investors) this should be updated based on the same conditions set out by CESR in its guidelines on the methodology for the calculation of a SRRI (CESR/10-67) i.e. (in summary) only where the volatility of the PRIIP has fallen outside the bucket corresponding to the previous risk category on each weekly or monthly data reference point over the preceding 4 months. For products that are not continuously marketed, as indicated in our response to Question 50, no updates to the risk indicator should be required.
• A distributor intending to use a KID to market a product should take responsibility for checking with manufacturer whether KID is up to date. A distributor must not proceed to market a product actively if the manufacturer indicates that a revised KID is required.

• Where products are not actively marketed/there is no open offer period/liquid secondary market, requiring a manufacturer to review and update the KID would be disproportionate and inappropriate. Most structured products, closed funds and OTC derivatives would fall into this category.

• If a manufacturer facilitates a buy-back of an investment product from a retail investor wishing to exit a product (e.g. by providing a bid to allow a single investor to exit a product) there should be no requirement for the KID to be updated. Any such requirement may be counterproductive in that the costs involved in updating the KID may deter the manufacturer from facilitating such buy-backs and so increase liquidity risk. Only if the manufacturer makes active two way pricing in the product available to retail investors should the requirement to update the KID be triggered (as per our response to Question 49).

• For Prospectus Directive PRIIPs, the requirements of the Prospectus Directive and the Transparency Directive should already provide sufficient protection to investors for the purposes of making them aware of updated financial information in relation to the manufacturer and significant new factors, material mistakes or inaccuracies in relation to the information included in the prospectus which is capable of affecting the assessment of the securities.

• Unless clear and reasonable limits are set in relation to (i) when the KID must be reviewed and updated; and (ii) what aspects of the KID must be updated, it will be very challenging (or, indeed, impossible) to keep the KID up-to-date. Some products, for example any leveraged product, may be very volatile and may have substantial intra-day price fluctuations which could lead to a change in the risk indicator on a daily basis. In the absence of clear and reasonable limits in relation to updates, there is a risk that the KID will never be fully up-to-date as the risk profile may change constantly.

• As previously indicated, it is important that only costs that are within the knowledge and control of the manufacturer are included in the KID. It would not be appropriate for distributors/advisors/third party costs to be included.
No comments
• JAC members have expressed strong views that it would not be reasonable/practicable to require manufacturers to actively inform investors of changes to the KID as they typically have no direct relations with the end investor. Even the distributor may not have a continuing relationship with the end investor after the initial sale of the product. Provided the initial KID states where updates to the KID will be made available (e.g. on a specified website or at the office of the principal agent or on the website of a relevant stock exchange, noting the importance of preserving flexibility in relation to where such updates will be made available) this should be sufficient.

• If an active communication model were required, the entity (if any) with a direct relationship with the investor should be the one responsible for ensuring this communication. In most cases this would not be the manufacturer.

• An active communication model is not required for updates to the UCITS KII. It should not be required for updates to the KID.
• From a practical perspective it is important to note that requirements on timing of delivery of the KID must take into account the fact that it must be consistent with legally binding pre-contractual and contractual documents and the information required by Article 8 of the Regulation, hence it is unlikely that it can be finalised much in advance of any Prospectus, particularly taking into account the fact that it must include, for example, information on costs which may not be available much in advance of the issue of the product (if at all)

• The criteria for determining the meaning of “in good time” in Recital 83 of MiFID appear to be a useful model provided that the practical constraints on timing of delivery of a KID consistent with the content of the Prospectus and containing information on, for example, costs, are taken into account
• No
• We believe the Discussion Paper offers a well-reasoned analysis of the potential benefits and drawbacks of the ESAs preparing templates.

• Given the broad range of products that are covered by the PRIIPs Regulation we foresee significant difficulties in preparing a sufficiently broad suite of templates to cover all contemplated products and indeed that would leave sufficient scope for future product innovation.

• We are concerned that the provision of a non-exhaustive suite of templates may compromise the discretion and increase the risk for manufacturers whose products do not easily fit within the templates provided.

• Notwithstanding the foregoing, if the ESAs are able to commit sufficient resources to this work to address these concerns we believe that such templates would be very valuable both to consumers (in driving conformity of approach and hence aiding comparability) and to the industry (in providing greater legal certainty and assisting with the significant challenges that implementation will pose). If the ESAs do opt to provide templates we would note:

(i) This should be addressed within the RTS and not at Level 3. It is likely that any Level 3 measures would be published too late: manufacturers having already made and financed the necessary IT and infrastructural developments to produce KIDs.

(ii) The use of the templates should be optional (even for products which correspond exactly with one of the templates). As discussed in the Discussion Paper, it will be particularly important to reserve flexibility for manufacturers to adopt alternative approaches where, in the manufacturer’s determination, that is more appropriate for the product in question. Given the very wide range of products covered by the regime, even with a comprehensive templating initiative it is inevitable that some products may not fit easily into one of the prescribed templates. Allowance must also be made for future product innovation.

(iii) It will be important to work closely with the industry in formulating the templates as they will:

(a) need to be of a nature that the industry can in practice implement (including being workable for the vast majority of products); and
(b) be most useful if they are able to address all (or a significant majority) of the uncertainties (and hence documentation risks) raised by the regime.

We note that the experience in respect of the guidance provided for the UCITS KII regime was that the concerns raised by the industry were not always addressed. The PRIIPs regime is likely to produce better outcomes for both consumers and the industry if that can be done.
• A regular payment option should in all cases be treated as distinct to the product itself (as it is for UCITS) and should be addressed separately (e.g. in a separate pre-contractual document specifically and solely covering that payment arrangement). As noted in the Discussion Paper this would be consistent with the PRIIPs Regulation.

• As an exception, we would agree that for products for which regular payments are inherent in the contractual form (and hence a single payment option does not exist) it would seem logical that the KID should be prepared on the basis of regular payments. In such case, assuming a single payment is made would be misleading.
• The choices that will be made as to the use of narratives or graphical illustrations for the performance scenarios will have different cost implications for manufacturers, the latter being more costly to produce.

• Similarly, the criteria for reviewing and updating a KID will have significant impact on costs, as they will result in a higher or lower proportion of KIDs to be reviewed and updated.

• The methodology for the calculation of a risk indicator may also have cost implications if requires manufacturers to obtain access to data not already available to them or to develop models not already used by them. It is paramount that the risk indicator methodology is based on data and models already available to manufacturers to avoid creating barriers for smaller manufacturers (for example, we do not consider that it would be acceptable for some manufacturers to have to rely on a third party calculation of the risk indicator).
We agree with the drivers identified for the Level 2 work. We would add that we are concerned that benefit is forensically analysed and all limitations of any approaches/features are given due weight. This relates particularly to areas such as the probabilistic modelling of likely return scenarios and benchmarking of returns and costs where the theoretical benefit of such approaches (and potential perceived benefits expressed by consumers through consumer testing) needs to be weighed against the real concerns regarding reliability of this modelling in light of the level of assumptions that will need to be applied to allow application across all PRIIPs. As discussed, the limited space available in which to explain the nature and effect of such assumptions (which in reality means that it will not be possible to explain these limitations in full) and the limited capacity of consumers to understand and weigh up the significance of such limitations must be factored into any determination of their benefit (resulting in the risk of incorrect conclusions being drawn). This is particularly the case given the likely significant costs in implementing such measures and monitoring the inputs and updating the outputs of such models. Such significant cost will inevitably have a disproportionate impact on smaller manufacturers.
• JAC members have voiced concerns that there is a risk that any assumptions could potentially be misleading for investors. As such it is imperative that assumptions are kept to a minimum and relate only to matters that retail investors are easily able to understand. For example:  - Holding period: to maturity  - Rates of return: none  - Amount invested: a nominal EUR 1/100  - Rebates (passed to the customer via the distributor): cannot be disclosed in the KID as not within the manufacturer’s knowledge/control
+44 (0)20 3088 3507