By Craig Simpson
Regulators in New Zealand want to make it easier for investors in KiwiSaver and other funds to assess the risk of their investment or potential investment, and to allow for comparisons across funds.
We actively support anything that makes it easier for investors to compare funds and assess the potential risks. This is important because there is no consistency to how the funds themselves label their products which is both disappointing and frustrating.
Having looked at the new risk indicator guidelines from the Financial Markets Authority (FMA) in some detail we have some concerns. In a number of cases the calculated risk score is understating the potential future risk of the funds.
The variance of returns in a market environment that is stable will lead to lower levels of volatility. An all-equities portfolio, for example, should in theory be considered higher risk.
In some examples highlighted below several funds have the same risk indicator despite having very different asset allocations (see: The risk indicator in action below).
It is often advised - even by the FMA - that investors should realise past performance is no indicator of future performance. So why then is the FMA risk indicator using past returns over a reasonably short time frame (i.e. 5 years) as a potential indicator of the future risk of a fund? If the FMA wants to use an historical track record then it should make the time frame longer so it takes in more market cycles and economic conditions.
For our own KiwiSaver regular return calculations and risk categorisation we will continue with our own current approach as we believe it is more robust than relying solely on a quantitative measure based on short term historical return data.
A more detailed look at the FMA risk indicator system is provided below along with some real life examples of how the system is applied.
Risk indicator thresholds
In order to make the assessment of risk easier New Zealand's regulators have chosen to adopt the same process and principles promoted by the Committee of European Securities Regulators (CESR) in order to calculate a ‘synthetic risk and reward indicator’ for key disclosure documents. The stated aim is to 'simplify the risk assessment process and enable investors to make informed decisions'.
The illustration below shows the proposed risk category, volatility ranges and descriptions.
| Risk category | Annualised standard deviation | Description of volatility |
| 1 | 0% to less than 0.5% | very low |
| 2 | 0.5% or more, but less than 2% | low |
| 3 | 2% or more, but less than 5% | low to medium |
| 4 | 5% or more, but less than 10% | medium to high |
| 5 | 10% or more, but less than 15% | high |
| 6 | 15% or more, but less than 25% | very high |
| 7 | 25% or more | extremely high |
Investors may see the risk indicator displayed graphically in various forms in fund updates or marketing material.
The FMA warns that managers need to exercise care and ensure the fund name does not misrepresent either the funds risk category or volatility; or the types of products the fund may invest in. The FMA does not mandate specific names and states they cannot give definitive guidance on all names and circumstances that might be misleading.
Risk indicator must reflect future potential volatility
For those wanting further details on how the synthetic risk indicator is calculated you can refer to the CESR paper here. Workplace Saving NZ in conjunction with BNP Paribas Securities Services provide some more useful background information in their 2014 presentation here.
It is important that fund managers do not mislead or deceive the public with their risk indicators or fund name. In instances where the risk indicator has been calculated using the prescribed methodology and the result is in inappropriate for the fund or is likely to mislead/deceive the public, an alternative risk measure must be calculated that accurately reflects the future volatility of the fund.
The risk indicator in action
Below is the sample we found and although we have used reported and publicly available documentation, we have removed the fund manager(s) name. We have done this as some readers could mistakenly come to the conclusion that the risk indicators could be misleading or not accurately reflect the risk profile of the funds.
Remember the calculation methodology is prescriptive and based on historical data.
Looking at the risk indicator results below it is obvious to us the system does not offer investors a silver bullet solution to assessing a fund's risk.
| Tested fund ... | FMA risk category | Asset mix | Methodology adjustment |
| Fund 1 | #4 | 98% shares | not going 5 years so part market index and part fund returns for volatility calculations |
| Fund 2 | #4 | 10% cash and fixed income, <10% property and the balance in shares, and other investments | |
| Fund 3 | #4 | 25% cash and fixed income, <10% property and the balance in shares, and other investments | |
| Fund 4 | #3 | 45% cash and fixed income, <10% property and the balance in shares, and other investments | |
| Fund 5 | #3 | 65% cash and fixed income, <5% property and the balance in shares, and other investments | |
| Fund 6 | #2 | 80% cash and fixed income, <2% property and the balance in shares, and other investments | |
| Fund 7 | #1 | 100% cash or equivalents |
Observations
The first issue we see several funds all have the same risk indicator despite there being some considerable variance in the asset allocations.
Of particular interest for us is the classification of Funds 4 & 5. The asset mix suggests the funds could belong in the traditional Balanced and Moderate fund categories respectively. Our immediate concern is how can an average KiwiSaver investor accurately distinguish between these two funds based on the risk indicator alone? Simply, they can’t.
The investor would need to have some knowledge around the investment philosophy and asset allocation construction to distinguish that one fund could be more volatile (risky) in the future simply by its greater allocation to equities over cash and fixed income. Investor education around the risk indicators and what the underlying asset allocation and security selection will be paramount we feel.
In fact, all the portfolios we examined have an indicator of 4 or below. At the top end the portfolios with the greatest potential for loss has an indicator of medium to high risk associated with them compared to what the industry would traditionally consider to be high risk because of their high (above 80%) equity exposures.
On this basis we would consider the calculation methodology is leading investors to under-estimate the future potential risk of their investments.
Other flaws of the proposed risk indicator system we recognised include:
- The calculation methodology is not completely standardised so results may vary. Funds that price daily or weekly must use weekly data points to calculate standard deviation which is the starting point for the final figure that is assigned.
- The monthly calculation of standard deviations (volatility) often results in a downward bias as the data points are less frequent and returns are smoother. This may cause such funds to produce a lower risk indicator figure than is truly the case and mis-represent the risk-reward profile of the fund from a forward looking perspective.
- The calculation doesn’t take into account any diversification. One of the goals of traditional portfolio construction is to include assets that have low or negative correlation with each other.
- The measures are purely quantitative and enable a more consistent approach, however they may not reflect all the relevant factors that may influence the risk of the investment e.g. quality of the fund manager, key person risk, characteristics of the asset class (e.g. low historic volatility of property funds).
- Researchers note that the risk indicator scale between 1 (low) and 7 (high) uses absolute thresholds which remain static over time and do not move with shifts in the overall market volatility. In the real world volatility will rise and fall with markets or portfolio asset composition and is not static.
- Several important risks are not captured in the risk indicator - credit risk, liquidity risk, counterparty risk, and operational risk, for example.
Unintended consequences
A possible and undesirable outcome from the FMA's risk indicator system could be that the FMA sees an increase in the volume of switches due to changes in risk indicators or an investor's reassessment of their risk profile based on their perception of where they fit under the scale in the diagram above.
If we see a period of continued volatility there is a distinct possibility the risk indicator for the more aggressive funds will change and members who are targeting a specific risk profile or risk indicator number will be forced to switch funds or providers.
In the financial year to March 2015 the FMA noted in its annual report, approximately 93,000 Kiwis switched provider. Roughly 1,000 members made three or more switches over the same 12-month period. The switches involved approximately $1.4 billion worth of funds. Further detail on the switches made can be found in Appendix 7 & 8 of the FMA report here.
While the switching process is not difficult, it could be costly for investors as some managers impose a switching fee. If the industry saw an influx of switches, those managers who have the ability to charge a switching fee, but elect not to charge it will surely start imposing this and try to recover some of the additional administration costs from the customer.
Because the transfer between managers is done via an electronic transfer of cash, your investment in manager A is liquidated at whatever the prices are on the day, and the proceeds are sent to manager B who reinvests them back into the market, the investor is out of the market for a period and thus missing out on a potential return for the period the funds are in cash. This is obviously not an ideal situation especially if markets are rising.
We are yet to be convinced that the risk indicators are the right mechanism and would prefer a more detailed approach that takes into account quantitative and qualitative factors as well as an examination of the overall quality of the investment portfolio and strategy.
interest.co.nz's approach to risk indicators for KiwiSaver
We will not be relying solely on the FMA risk indicator when we come to assign KiwiSaver funds to the various categories. The risk indicator will be just one tool available to us and we will continue to look through the fund and take into account the overall asset allocation, regional exposure or style tilts and security holdings as part of our overall methodology and risk assessments.
We will however report on what the manager has calculated as the risk indicator for each fund when these come on stream as part of our ongoing assessments and analysis so readers have some perspective on where the funds sit on a historical volatility basis.
Over time if the risk indicator methodology proves to be robust and accurately reflects what we deem to be the real risk of the fund, we will review our stance.
6 Comments
As much as MBIE or the FMA (or anyone else) wish to say that the risk indicator methodology is "based on" the CESR methodology, the reality is that it is a carbon copy - albeit with a change of title.
So why should that be a problem? Well, the CESR methodology was designed for pan-European UCITS funds and delivered as part of the UCITS IV directive. UCITS IV didn't just appear out of the ether, it was a development of UCITS I (1985) and UCITS III (2002).
Each iteration of UCITS built on what had already been developed, for example UCITS III introduced the concept of UCITS management companies being required to have risk management programs covering all funds.
This is important because it is linked to the UCITS IV Synthetic Risk and Reward Indicator methodology. The starting point for CESR when creating the methodology was knowing that all UCITS management companies already had robust risk management frameworks in place. This is not the case in NZ.
A couple of other minor differences between pan-European UCITS funds and NZ is that we have a predominantly fund-of-funds setup dominated by multi-sector funds (conservative, balanced, growth etc) whereas UCITS is dominated by single sector funds where most have direct market exposure. What this means for risk indicators is that you are more likely to get wider spread on the 1-7 scale when you are dealing with single sector instead of multi sector funds.
The other "minor" difference comes in the form of a key UCITS concept - investment restrictions. UCITS funds may only invest in transferable securities and money market instruments. Physical property is specifically excluded. Given that the CESR methodology was designed specifically for UCITS it was never intended to be used for funds with physical property exposure. Why do I mention this? There is no such investment restriction in NZ but we have copied a methodology designed with these restrictions in mind. Anyone who has visited the Disclose website may have noticed there are funds there that only invest in physical property and have unusually low risk ratings.
Unfortunately, this is what can happen when you simply copy a methodology developed for one market and don’t adapt it to cater for the idiosyncrasies of your own.
Thanks for the additional insights. Agree with the comments re property and this is one area that concerned me as the volatility is often lower because of the valuation methods adopted and this skews the indicator for multi-sector funds to the downside.
There is no silver bullet when it comes to risk assessment of funds as many are backward looking and don't provide insight into the future.
I think we will see managers having to over-ride their risk indicator scores with their own assessments of the likely future risk and use this so they don't mislead the public.
Corporate debt maturity over the next 5 years is huge,
http://www.bloomberg.com/gadfly/articles/2016-02-29/-9-5-trillion-debt-…
We welcome your comments below. If you are not already registered, please register to comment.
Remember we welcome robust, respectful and insightful debate. We don't welcome abusive or defamatory comments and will de-register those repeatedly making such comments. Our current comment policy is here.