In a previous article
, we described the changes in the draft amended PRIIPs RTS applicable for structured products. The amended RTS proposal was finally approved by EIOPA’s board and now it is up to the Commission to determine the next step and whether this RTS becomes active by the end of this year.
In this paper, we focus on the amendments related to autocallables
. These products are known by various names in different countries including Kick-outs in the UK, Express Certificates in Germany, Austria, and Italy, and Autocall in Belgium, France, Luxembourg, Netherlands, Nordic countries and most of Europe. We limit the scope of this paper to the changes in performance scenarios. In a future paper, we will address changes relevant to cost disclosure.
It is noteworthy that the ESAs have decided to dedicate special instructions for a specific financial product. Perhaps this is due to the many questions raised by market participants regarding this product type or because of their popularity in recent years.
We must also point out that autocallables are not always considered as distinct product type. Rather, in some markets and countries, autocallability is a product feature. For example, a digital product (i.e. paying income based on a yes/no decision) might have an autocallability feature built in. We understand the ESAs meant to apply the new requirements on any product with an autocallability feature.
The body of this paper expects, on the part of the reader, a certain amount of technical knowledge of the PRIIPs RTS, autocallables and statistics. For readers lacking this foundation, we have summarized our main findings:
- The new requirements in the amended RTS proposal mean a major change compared to the current RTS.
- The new requirements will demand a significant development effort.
- The new requirements lead to several shortcomings in the performance scenarios table such as loss of comparability with other products.
- On the other hand, in a certain way the new requirements might simplify the performance scenario results for the end investors.
New Methodology in Detail
As mentioned in the executive summary, whether the ESAs intentionally meant that or not, the new requirements for autocallables mean a major change compared to the current active RTS. Some amendments are generic, for all products, and affect autocallables as well. Other amendments are specific for autocallables. In order to demonstrate the extent of those amendments and their implications, we will begin our analysis by utilizing a sample autocallable product.
For the analysis, we picked a relatively simple and popular product: a 5-year EUR denominated autocallable product, observed annually. In case the S&P 500® index level increased at an annual observation date (compared to the index level at the launch of the product) the autocall product matures early repaying full capital plus 7% coupon for each year since issue (we priced the product when the amended RTS was published during 2020 and the implied volatility was still high, therefore the high coupon of 7%). At final maturity (i.e. after 5 years), if the product did not mature early, there are 3 scenarios:
- If the index is equal or above its initial level (the index level at the launch of the product) – the product will repay full capital plus 7% coupon for each of the 5 years.
- If the index is below its initial level and above or equal 50% of its initial level – the product will repay the full capital (with no coupons).
- If the index is below 50% of its initial level – the product will repay the capital reduced by the decline of index.
To summarise, in professional terminology:
- 5-year EUR autocall
- Underlying: S&P 500® price index (Bloomberg: SPX Index)
- Autocall trigger: 100%, annual observations
- Risk barrier: 50%, checked at final observation date (i.e. European)
- Autocall exit coupon: 7% (with memory feature)
Running the 10,000 bootstrapped scenarios (with the 5-year historical drift as of 24 July 2020) results in the following distribution of redemption scenarios:
As the amended RTS proposal instructs, no reinvestment of the “amounts received before the end of the holding period” should be assumed. Therefore, the distribution of total repayment (without time valuation or capitalization) is:
In other words, at RHP the Unfavourable and Moderate Scenarios show a value of EUR 10,700 (early maturity after one year), while the value of the Favourable Scenario at RHP is EUR 12,100 (early maturity after three years). For the sake of simplicity, we will disregard the Stress Scenario.
So what is the “Significant Change” in the amended RTS proposal? Let’s begin with the Intermediate Holding Periods (IHPs).
Intermediate Holding Periods
Since the amended RTS proposal requires RHP/2 IHP to be included only if RHP is 10 years or more, we will focus only on the 1Y IHP.
One simple fact to realize is related to the Favourable Scenario: the scenario that pays 3 coupons and is autocalled after 3 years. For this scenario to happen, the underlying index must be below 100% in each of the two previous annual observations. Usually, negative performance of the index means a negative performance for the entire product.
Now let’s compare the active RTS and the amended RTS proposal in the paragraph that describes how to calculate IHPs for Category-3 PRIIPs:
ANNEX IV point 24 (c) of the active RTS
In other words, the active RTS requires the scenarios at the IHPs to reflect the 10th/50th/90th percentiles of the PRIIP valuation at each IHP
However, the amended RTS
proposal says (ANNEX IV point 37):
Therefore, the scenarios selected for IHPs should be ones leading to the scenario shown at RHP
This difference is not just a semantic change. Let’s have a look at the sample autocall. The performance of the underlying index in the simulated 10,000 scenarios is distributed differently after one and five years:
As observed, at 1Y IHP only 6,616 scenarios are above 100%, while at RHP 8,131 are above. This is a reflection of the positive 5-year historical drift of the S&P 500® index.
Therefore, at 1Y IHP, for the Favourable Scenario
- According to the current RTS we should choose the 90th percentile. Assuming negative performance of the index means negative performance of the autocall, the Favourable Scenario at 1Y reflects an early call at 1Y, i.e. a value of EUR 10,700.
- According to the amended RTS proposal we should choose a scenario consistent with RHP, i.e. a scenario that will lead to an early call after 3 years. For a scenario to “survive” for 3 years, at 1Y the index should be below 100%, meaning the Favourable Scenario will show a negative performance at 1Y IHP. This is very counter-intuitive, as investors expect the 1Y Favourable to be the 10% best case after 1 year, irrespective of what happens at 3 years. Most investors in this autocallable expect the Favourable Scenario to redeem at 1Y with a value of 107%.
For the Unfavourable Scenario
at 1Y IHP:
- According to the current RTS we should choose the 10th percentile. As mentioned above 33.84% of the scenarios show a negative performance of the index after 1 year. Therefore, the 1Y Unfavourable Scenario will show a loss.
- According to the amended RTS proposal, “Figures for intermediate holding periods shall only be shown for scenarios where the PRIIP has not yet been called or cancelled before or at the end of that intermediate holding period” (ANNEX IV, point 17). Therefore, we should leave the Unfavourable 1Y IHP empty.
Performance Scenarios Table
Based on the analysis above, let’s see how the performance scenarios table changes from the current RTS to the amended RTS proposal. To ease the comparison, even though needed, we eliminate the RHP/2 IHP in the current RTS. Below you can see the results consistent with the current applicable RTS
Investment EUR 10,000
In the table above, the PRIIP value for each holding period (1 year, 5 years) and scenario is consistent with the requested percentiles
. The investor knows, for example, that after one year, the positive case would be a 1-year early call
. The investor can also easily understand that the 10% best case at the RHP corresponds to a 3-year early call at 121%. This also means that different holding periods in the same row can have different early maturities
– in this case it happens in the Favourable Scenario. These guidelines have been tested with dozens of distributors in the EU, and following their feedback the vast majority of manufacturers, members of EUSIPA national associations, have endorsed these principles for autocallables
(for more information, please see Recommendation 4 at: https://eusipa.org/eusipa-rts-implementation-advice-published/
Now, let’s see how the performance scenarios would look according to the amended RTS proposal
A few remarks on the table above:
- A counter-intuitive negative performance at 1Y Favourable Scenario: For the 1Y Favourable Scenario to be “consistent” with an early call after 3 years, the index should be below 100%, and therefore the product shows negative performance. This is counter-intuitive as it would lead investors to think that even in the 10% best case, they will lose money at the 1Y IHP.
- The amended RTS proposal doesn’t describe how a manufacturer should calculate the PRIIP value at IHP that is “consistent” with the RHP scenario. For the 1Y Favourable, for example, we have looked at all scenarios that led to an early call after 3 years (663 scenarios out of the 10,000) and chose the median product value at 1Y of those scenarios (and applied a 1% early exit cost). However, this methodology can lead to a stability problem. In case a rare scenario (having less paths in the 10,000) was selected for the Favourable Scenario (i.e. by selecting the 90th percentile at call or maturity), the value at 1Y IHP can fluctuate significantly due to the random nature of the bootstrap.
- Having empty cells at IHPs and multiple different periods at the rightmost column requires infrastructural changes at the industry level. For example:
- The EPT (European PRIIPs Template) format will need to add the relevant information (i.e. when each scenario is called) and allow empty IHPs.
- Recalculation of costs for IBIPs (Insurance-Based Investment Products) that embed autocallables (MOP and non-MOP) requires further examination. This will be elaborated further in our following paper that focuses on autocallables’ cost disclosure.
- The holding periods in the new table for autocallables are different from products with no autocallability. While in other products, the rightmost column always shows the RHP, for autocallables it presents the actual maturity (either final maturity or early call). This means there could be comparability challenges with other products.
- As the above example shows, several outcomes aren’t easy to understand, such as empty columns and negative 1 Year Favourable Scenarios. Before any new methodology is applied, in the form of amended RTS, Level 1 changes or Level 3 Q&As, this would obviously require a thorough consumer testing and a cost/benefit analysis to be determined conclusively.
The new methodology for autocallables requires significant efforts to implement. And yet, there are still many unclear specific details (especially regarding Intermediate Holding Periods) and several drawbacks, such as lack of comparability with other products.
I would like to thank my colleagues Eddie Korol, Shai Corfas, Stefan Marcus, Markus Diesing and Daniel Herzbach for their contribution to this paper.