Across Europe, the retail investments industry (comprising structured products, funds and insurance investments) awaits ESMA’s proposed revision to the PRIIPs regime after the consultation process at the end of last year. These changes will have to be ratified by the European Commission before a likely implementation date of the start of 2022. This revision hit a major stumbling block when the ESAs could not agree on the draft revisions, as published in July 2020.
While this process runs its course it is interesting to examine how the calculation results have changed as a result of the market events of 2020. All major indices have been extremely volatile this year and this has started to feed through significantly to five year historical volatility which is one of the key determinants of PRIIPs output.
Index volatility history
The Eurostoxx-50 rolling five year volatility level was at 21% in mid 2015 and climbed a little further to nearly 23% before slowly drifting down over the last few years to 17%. The SRI (synthetic risk indicator) for an ETF linked to the Eurostoxx-50 would have fallen from a 5 to 4 on the PRIIPs SRI scale of 1 to 7, since the boundary of those two bands is 20%. This changes in risk rating has not been as significant because the last few years have been extremely calm, but the primary contributing factor is that older data has dropped off. The five year historical volatility at 2015 references data between 2010 and 2015, the early part of which was an extremely volatile time for the European market. As this window moves forward the volatile years of 2011 and 2012 fall away hence the volatility drifts down. During 2020, the inclusion of recent market events has started to move volatility back up again rapidly from 17% in March 2020 to 20.5% in July, putting it back as a 5 out of 7.
The FTSE-100 has shown a similar story, with its five year volatility rising from 13.8% in March to 17.8% in July. This is a pretty significant move, however because the risk band 4 covers a volatility range of 12% to 20% there would be no change to the published risk rating. It is part of the approach of the PRIIPs regime to keep a simplified presentation of the SRI but this can hide the true picture when it comes to risk.
The three months between March and June 2020 is likely to remain among the most volatile periods for the next few years which means that in March 2025, long after the events of this year have faded from immediate memory the five year volatility number will fall quickly from whatever level it finds itself at as the data drops out of the five year window. This is a disadvantage of using a fixed window compared to other schemes such as exponential weighting.
How risk ratings work
Risk ratings for structured products also use five years’ worth of historical data as inputs to generate simulations for the risk ratings and performance scenarios. The risk of a product is calculated using a Value-at-Risk approach to convert the magnitude of each product’s downside risk as defined by its payoff formula into a volatility equivalent measure in order to map back to the same risk rating scale.
A typical EuroStoxx-50 linked Autocall has shown a slight increase in risk level over the last few months. By examining the “exact” risk number that comes out of the calculations and not rounding to the nearest integer, we see that the Autocall has increased in risk by 0.2 of a rating point in the last nine months, compared to an increase for a Eurostoxx-50 ETF by 0.33. When we consider that a structured product is typically part equity and part bond in nature, this smaller increase is logical. While the published SRI number is always an integer, in order to track changes it is useful to consider fractional moves to get an idea of direction and magnitude. These will feed into the published number eventually.
Some numbers moved others have not
Some observers may be surprised that risk ratings would not have increased more significantly given that the market has been dominated by large moves and talk of the accompanying risk. By using relatively long windows of data (five years), the risk ratings do not react quickly to market falls such as has been seen this year. If the risk ratings were based on one year volatility for example they would move much more frequently.
Other indices and underlyings show a different story. For example, the S&P GSCI Crude Oil Index has seen its volatility jump from 38% to 58% over the start of 2020 to reflect the dramatic moves in oil prices. This is an increase of over 50% of its previous level which is clearly a big move, yet because the volatility band for risk rating 6 is 30% to 80%, the risk rating of this fund would not change.
We further note that issuer credit ratings did not change for most banks over this period which means that this contribution to the PRIIPs risk rating has not moved.
Making sense of performance scenarios
Some interesting results can also be seen in the performance scenarios. One of the main differences between PRIIPs calculations for funds and those for structured products is that performance scenarios for funds use the five year average index return to calculate the moderate scenario whereas most in the industry in structured products have elected to use risk neutral scenarios. This means the moderate scenarios for structured products will not have moved much but that fund performance will show some significant decreases as the market directional falls of this year feed into the data.
Interpreting PRIIPs results for advisers and investors remains a challenging process with so many inputs and some complex calculations. Keeping investors comfortable for what these results mean is important as the market waits for further detail on future regulatory change.