In a letter
sent by the ESAs (the European Supervisory Authorities, i.e. EBA, EIOPA and
ESMA) to the European Commission, the ESAs have included a draft amended PRIIPs
RTS. As the letter points out, this amended RTS was not
supported by EIOPA’s board though it did receive the support of ESMA and EBA. So
while the future of the RTS as a whole is still uncertain, we do believe that as
far as Structured Products are concerned, we should not expect too many changes
to this RTS going forward. Therefore, we have decided to start analyzing its
content and summarize the main changes.
We reviewed the differences from the current active
PRIIPs RTS (including the Q&As) and not from the latest Consultation Paper,
as we would imagine most readers are interested in the expected changes from
their current PRIIPs KID implementations.
In this article, we will focus on the changes
applicable for Structured Products. The changes listed below are not an
exhaustive list though we believe these to be the most substantial changes. RTS
amendments that appear to simply be minor wording corrections have been
important change has been applied to the SRI calculation criteria. The amended
RTS allows manufacturers to increase the SRI if they feel it “does not
adequately reflect the risks of the PRIIP”, while asking the manufacturer to
document “such an increase.” We think this addition is a result of the UCITS
exemption expiry. Usually, UCITS SRRI are higher than PRIIPs SRI. The ESAs
would like to eliminate the confusion caused to investors when a UCITS instrument
falls into scope for PRIIPs and their risk indicator is suddenly reduced. Subsequently,
we suspect this change might be less relevant for structured products.
As far as structured
products are concerned specifically, there are only several minor changes to
the SRI calculation criteria:
1. VEV buckets: it wasn’t clear for a
specific bucket whether the range is included or not. For example, MRM class 3,
VEV range of 5% - 12%. Are 5% and 12% are included or not? The new RTS
clarifies it is greater or equal to 5% and lower than 12%.
2. In the VEV formula (ANNEX II, point 17 for Category 3 products), the
1.96 has been moved out of the square root:
However, it is important to mention that the
1.96 was already out of the square root in the delegated regulation (published
8 March 2017) and in the PRIIPs flow diagrams, and it just looks like a typo in
the version that ended up in the official journal. Therefore, it is safe to
assume most manufacturers are already in line with the amended RTS.
3. In the SRI narratives (ANNEX III),
the amended RTS adds several clarifications regarding when each narrative
should appear. Manufacturers will need to verify they are in line with those
instructions. Still, we have discovered several noteworthy typos in the amended
RTS. For example:
SRI determination for structured products is relatively untouched, the amended
RTS suggests numerous changes to the performance scenarios calculation and
presentation. Some of those changes warrant a closer look. We intend to further
elaborate in a set of dedicated articles. For the time being, here is a summary
of the changes:
1. Category 2 methodology has
completely changed. The Unfavourable, Moderate and Favourable Scenarios are no
longer based on Cornish-Fisher expansion. However, since there are only a few
Category 2 Structured Products (such as Actively Managed Certificates tracking
a portfolio), we intend to further elaborate on this topic in a dedicated
article focused on funds.
2. The ESAs decided to leave as is the
historical drift and abandon the methods they suggested in order to deal with
the pro-cyclicality problem (e.g. risk premia for equities). Instead, they
allow manufacturers to use percentiles lower than 10th, 50th
and 90th for the Unfavourable, Moderate and Favourable Scenarios when
they believe the scenarios are too optimistic. There are several challenges with
- From the legal perspective, it is
transferring the responsibility to treat too optimistic scenarios to the
- From the technical point of view, it
is unclear how to quantify “inappropriate expectations”. It is simply too underlying
and payoff dependent.
- Another technical issue: it is not a
trivial task to lower the percentiles and reach an appropriate expectation. Not
to mention there is simply no guarantee of finding one.
3. In the Consultation Paper, the ESAs
suggested to replace the Stress Scenario with a Minimum Scenario. However, in
the amended RTS the ESAs decided to keep the Stress Scenario, and to require,
in addition, the minimum scenario. In the Minimum Scenario, the minimal value
in monetary terms should be specified. And for structured products with an RHP
longer than a year it gets a bit more complicated, as the minimum is only
guaranteed at maturity and not in all Holding Periods. In this situation, the
amended RTS requires further narratives.
4. Intermediate Holding Periods (IHPs) –
the ESAs decided to keep 1Y IHP as of today (i.e. for RHP longer than a year)
and discard RHP/2 for products with an RHP less than 10 years. For products with
a 10-year RHP or more, there will be three Holding Periods, similar to the
current RTS. Another change in IHPs is tricky to spot. The ESAs do not describe
this change in chapter 4 (where they list main changes, feedbacks and
considerations), but readers are expected to realize this change by manually
comparing the amended RTS with the active RTS (a simple doc compare won’t help
here, too many differences in performance scenarios annexes). For Category 3
products, the ESAs give up the active RTS requirement to “choose underlying
values consistent with the 90th, the 50th, and the 10th percentile levels… and
use these values as the seed values for a simulation to determine the value of
the PRIIP.” Instead, the amended RTS requires values at IHPs to be “consistent
with the estimation at the end of the recommended holding period”. This
“semi-hidden” change that is just merely embedded within the RTS is rather a
significant one. Sorting the 10,000 scenarios just by their RHP and choosing
their respective scenarios at IHPs would lead PRIIPs with callability (either
autocallability or issuer callability), for example, to show values at IHPs not
at an ascending order (e.g. Favourable 1Y IHP can be in many cases lower than
Moderate 1Y IHP).
5. Monetary unit rounding – the amended
RTS required rounding the values to 10 EUR (or equivalent in other currencies),
unless “it could be misleading to round the figures”. In these cases, the RTS
requires rounding to 1 EUR. It is unclear what those “misleading” cases are. In
addition, with the current low rate environment, 10 EUR rounding, especially
for 1Y IHP, is not negligible.
6. The amended RTS clarifies that
intermediate cash flows, such as coupons, will not yield a return. This is in
line with EUSIPA’s Recommendations 4.1.A and 4.1.C available at: https://eusipa.org/eusipa-rts-implementation-advice-published/
7. The ESAs finally clarified what
formula should be used for the “Average return each year”. It is simply:
“(scenario value/ initial investment)^(1/T) – 1”. Therefore, all Manufacturers
using non-compounded returns and other variations will need to adjust their
implementation. For RHPs shorter than 1Y, the amended RTS explicitly states
that the return should not be annualized, similar to what have been already published
in the Q&As.
8. The ESAs require adding a brief
description (up to 300 characters) of the presented scenarios. There are a few
more changes to the current existing narratives.
9. Finally, for dessert, the amended
RTS includes a special dedicated performance scenarios table for autocallables.
The ESAs’ suggestion contradicts EUSIPA’s recommendation (available within the
same link provided above) that is based on the active RTS requirement to keep
consistency with percentiles 10th/50th/90th at
each IHP. As mentioned above, the amended RTS replaces this demand with a
requirement to be consistent with the value at RHP. We believe the ESAs’
suggestion requires further clarifications, as it will face several challenges,
- Losing the order at IHPs. In many
cases, an autocallable’s Favourable Scenario at RHP is called later than 1Y
IHP. With the common autocall trigger at 100%, this would typically mean a
PRIIP value lower than 10,000 Euro at 1Y IHP. Assuming the Moderate was called
at 1Y IHP, its value will be higher than the Favourable. This is
- It raises comparability issues, as
the rightmost column for autocallable products represents the product
value at the actual redemption date (which might be different for each
scenario) as opposed to the product value at RHP as for all other
types of products.
the performance scenarios section, the costs section includes plenty of
amendments on transaction costs, because, unlike funds, few structured products
actually bear transaction costs (e.g. Actively Managed Certificates), we intend
to further elaborate on this topic in a dedicated article focused on funds.
As for the
presentation of costs, the ESAs decided to keep the two cost tables. Similar to
the current RTS, the first is the “Costs over time” table and the second is the
“Composition of Costs” table. Still, their contents have changes as specified below.
1. Similar to
the performance scenarios table, RHP/2 should appear only for PRIIPs
with a recommended holding period of 10 years or more.
2. The RIY
concept is still used, but with a few changes:
It has been
renamed “Annual cost impact”.
A simple description explains the concept in a narrative below the table, showing the return
at RHP before and after the costs.
Moderate Scenario is used for the RIY calculation only for RHP and RHP/2 (if
presented). For 1Y IHP, a net performance (i.e. after costs) of zero is used to calculate the RIY. This is mainly for achieving better
alignment with costs presented in MiFID II.
shorter than 1Y, the amended RTS explicitly states that the RIY should not be
annualized, similar to what has been already published in the Q&As. In this
case, the RIY at RHP should be based on zero net performance and the label
should say “Cost impact” instead of “Annual cost impact”.
narratives below the table should appear, in the event the manufacturer
remunerates the distributor.
4. A special
“Costs over time” table should be drawn for autocallables (again, like in the performance
scenarios, not for products with issuer callability) - The table should have only two Holding Periods: first autocall date and
recommended holding period, i.e. final maturity.
1. In this table, it seems like the scenario to be used at the first autocall date, as the basis for the RIY calculation, is an early call scenario at this first autocall date. However, in some rare cases, this early call scenario might not be deterministic. For example, consider a Phoenix with a conditional coupon prior to the first autocall date and no memory. Was it paid or missed?
addition, it seems that even if the first autocall date is less than a year,
still the RIY should be annualized, since the non-annualization depends on the
RHP to be less than a year. Therefore, this case might show an exaggerated RIY
for short term first autocall dates (monthly, quarterly). Of course, maybe the
ESAs just missed this specific case, and hopefully they can fix it.
3. Finally, under an autocallable's performance scenarios table, several new narratives should mention it is uncertain when exactly the product is expected to terminate and that it depends on how the market evolves. “Composition
of Costs” table
will include three columns:
1. The cost
item name - performance fees and carried interests are combined to one line,
but these cost items are irrelevant for structured products.
– a flexible description of the cost. The RTS suggests a few wordings, usually
including the raw costs. For example, for entry cost: “x% of the amount you pay
in when entering this investment”.
3. Instead of RIY at RHP (which is kept only for IBIPs, i.e. insurance-based investment products), for other product types, including structured products, the third column should show costs in EUR if the investor exits after one year (or RHP if shorter than a year). This poses two challenges for structured products: (a) The minor challenge: since there is no special “Composition of Costs” table for autocallables, the costs in this table might not sum up to the “Total costs” presented at the first table (different Holding Periods: first table – first autocall date, second table – one year). (b) The major challenge: as most of the Manufacturer’s work is done at the launch of the product, Structured Products usually charge an entry fee. Apart from an early exit cost in the event the investor exited the product early, the entry fee is often the only cost charged by the Manufacturer throughout the entire life of the product. Showing costs at 1 year exaggerates Structured Products’ costs. Moreover, investors usually hold classical Structured Products till maturity and do not pay early exit cost. Costs at 1 year are then exaggerated twice. This presents these products unfairly.
In sum, at first glance, readers of the RTS (and maybe even the ESAs) might believe there are almost no changes in this RTS relevant for structured products. However, as we have demonstrated above, there are indeed numerous changes that often place structured products in an unfavourable light.I would
like to thank my colleagues Eddie Korol, Shai Corfas, Emanuel Pollak and Stefan
Marcus for their contribution to this article.