PART III: Morningstar investigates the drawback of the SRRI presented by the room for interpretation within the ESMA fund type definitions.
Since the European Securities and Markets Authority (ESMA) first proposed to create a Synthetic Risk Reward Indicator (SRRI) with the aim of providing ordinary investors with a simple measure of risk, Morningstar minds have been debating the limitations of such a measure and investigating the additional risk posed by oversimplification of a complex construct. In this, the third in series of our discussion on the topic, we outline the difficulties of defining funds by ESMAâs categories and put forward Morningstarâs own plans for providing investors with a reliable and universal risk measure.
What Are We Talking About?
Calculating the SRRI according to the specification from ESMA begins in fairly straightforward manner--the basic measurement is the annualised trailing 5-year volatility of the fundâs return. This measurement determines which risk score the fund will receive based on a proscribed scale. For instance, a fund with an annualised 5-year trailing volatility of more than 5% but less than 10% gets an SRRI of 4. The scale begins at 1 and goes up to 7, the higher the number the higher the historical volatility--easy and sensible enough.
However, things start to get a little more involved when you get past this basic concept. First, what if the fund/share class has less than five years of history? What then? The answer to that question depends on the answer to the next question: Which type of fund are we talking about? Further still, that answer depends on the investment strategy of the fund. ESMA has proscribed five types of funds: market, absolute return, total return, lifecycle, and structured funds (more on ESMA fund type definitions later in the article).
Without going into all the details and permutations, suffice it to say that things get a whole lot less straightforward when you move past the market funds. But this is not rocket science; we have Morningstarâs best quantitative minds working on implementing this calculation, plus we have years and years of experience collecting data and doing these sorts of calculations.
Due to these considerations regarding the specification from ESMA, and as we have been working through the details around calculating the SRRI, some of the same questions keep coming up both internally and with clients:
-- How many share classes have more than five years of performance history?
-- How many share classes fall into each of the ESMA fund categories?
These are easy-enough answers for us to come by given the depth of our European fund database. You may well have asked yourself some of the same questions. So we did the maths to add some precision to our intuition; below we provide some clear answers.
Crunching the Numbers
To answer the first question, regarding the number of share classes that have more than five years of performance history, we started by filtering our European fund database to show only UCITS funds. This cut the list down from 78,583 share classes to 55,667 UCITS share classes, or 71% of the total.
The next step was to count how many share classes have more than five years of performance history. This was a simple query over the inception date field for all remaining share classes. A bit to our surprise, the split is about 50/50. That is, just a little less than half of all UCITS share classes have a 5-year history. See the table below for the breakdown of the Top 10 countries.
The second question, of how many share classes fall into each of the ESMA fund categories, is a little more involved but still not that difficult. It involves mapping the funds in our database into the fund types specified by ESMA--market, absolute return, total return, life cycle, and structured funds. Based on our initial mapping, the below table provides a good indication of the relative breakdown.
Apples and Oranges
At this point things again become a little less clear-cut and open to interpretation. To highlight why, we should look at the ESMA definitions of the fund types:
Market - UCITS that are managed according to investment policies or strategies that aim to reflect the risk and reward profile of some pre-determined segments of the capital market;
Absolute Return - UCITS that are managed according to investment policies or strategies that envisage a variable allocation of the portfolio of the fund across asset classes, under the constraint of a predetermined risk limit;
Total Return - UCITS that are managed according to investment policies and/or strategies that pursue certain reward objectives by participating, through flexible investments, in different financial asset classes (e.g. in both equity and fixed-income markets);
Lifecycle - UCITS that are managed according to investment policies or strategies that imply a shifting over time of their portfolio allocation from equity to fixed-income assets, according to some predetermined rules as a target maturity date approaches;
Structured Products - UCITS that provide investors, at certain predetermined dates, with algorithm-based payoffs that are linked to the performance, or to the realisation of price changes or other conditions, of financial assets, indices, or reference portfolios.
As you may be able to tell from these definitions, the ESMA fund type definitions leave room for interpretation. That is, each fund house will be taking a view as to how their fund range maps to these categories. It is not difficult to imagine a situation where two similar funds are categorised differently by different houses.
For example, âTotal Returnâ could be read as any multi-asset fund or as any fund that retains the ability to move from being wholly invested in one asset class to wholly invested in a different asset class. Similarly, there are many different definitions of âAbsolute Returnâ funds across the fund industry. These are not merely academic questions because they have direct consequences for whether or not a fund will have an SRRI and the basis for it.
Things Just Got a Whole Lot Easier
What is clearly obvious, yet still interesting in its own way, is that more than 90% of all UCITS funds available for sale in Europe are so-called âmarket fundsâ. This is probably a good thing, at least from a calculation standpoint, because the maths for calculating the SRRI is straightforward for these funds as described above, unlike the proposed calculation methodology for the other fund types, especially the Lifecycle and Structured Products funds.
Even when a Market fund has less than five years of history, the ESMA specification calls for back filling the actual history with either â...the fundâs representative portfolio model, target asset mix, or benchmarkâ to create the 5-year history. Once this is done, it is easy to compute the annualised trailing 5-year volatility.
At Morningstar, we have been doing these types of calculations for many years. We call them âtrack record extensionsâ. In our methodology we first look for a share class of the same fund with a full 5-year history so as to use the time series to back fill the performance history (by using an older share class in this way, we can extend the history for another 10% of the fund universe). By using another share class instead of a passive benchmark, we do not lose the active risk portion of the return. We feel this offers investors a better indication of the historical riskiness of the fund.
Our research into the SRRI calculation is ongoing and we will continue to share our findings. In the end, our objective is to calculate a Morningstar SRRI score for all funds in our database regardless of their ESMA category or the length of their performance history. We hope this consistent methodology across the entire European fund universe will serve as an additional aid to investors in selecting funds that fit their risk tolerances and help them achieve their financial goals.