In response to a recent blog on benchmarks and behavioural economics I have been asked to comment on the importance of benchmarks in point of sale documents. The following is the detail supporting that response, and (at the end of the post) a set of answers to a number of questions:
First of all some questions?
Should non financial experts be responsible for making what in many instances are arguably expert financial decisions?
If they are to be held responsible and they lack expertise and experience to make complex decisions, should these decisions not be restricted to the simplest possible set of choices? i.e. simple investible benchmarks!.
If they are responsible for the decision, irrespective of the ethics and the reality, should they not be provided with the information that can help them make that decision as a matter of ethical principle?
When you are building a portfolio you need to start off with an asset allocation which reflects your risk preferences, your financial demands, your investment discipline and style and current market and economic conditions.
This asset allocation, depending on the sophistication of the process should have a further layer that details how much you have in cash, how much in fixed interest (corporate/government fixed interest), the maturity structure and the global diversification, how much in equities and within equities the style profile, market cap, sector, yield and global allocation + other assets and alternatives.
The less sophisticated the process the simpler the asset allocation profile and the more likely the asset allocation profile will equate to a broad generic cash, bond and equity market allocation + international diversification.
Moving outside the simple requires justification (it also involves responsibility), because it is strategic and involves taking on additional risk. This is unequivocal.
Now, in terms of fund selection, if you have a sophisticated process you may have specific allocations to small caps, mid caps and large caps, to value and/or growth, to yield etc. This should be reflected in the fund selection used to build up the asset allocation. The benchmarks used to assess performance of these investments should likewise reflect the asset allocation and risk management objectives of the fund.
Now, you can fill the asset allocation profile with investments that are expensive, that are wildly different from the intended asset allocation structure and/or and that have a history of under performance, or you can attempt to fill the asset allocation structure with investments that match your asset allocation profile and that provide cost effective risk/return management.
I would prefer the latter of the two, but if we are limited to an at a glance document we need something that provides a simple and direct comparison. In other words a benchmark!
Benchmarks can be highly complicated, but what the consumer needs, if they are to be held responsible, is a yardstick which will allow them to compare the performance of the investment recommended for them against simple low cost passive investible alternatives. If an investible benchmark is not available then an appropriate index or sector average should be provided. I am sure a hierarchy of index selection can be easily determined.
And of course, it is much easier if we can relate the recommended fund via its benchmark (i.e. small cap value, mid cap growth, broad market value etc) to the recommended asset allocation and investment focus of the recommended portfolio.
Benchmarks are one way in which an investor can gain an insight into the relative performance of a recommended mutual fund or other collective. The closer an investment resembles the benchmark the more likely the performance is to track it and costs in this context will matter.
The less the fund resembles the benchmark the more likely the performance of the investment is likely to deviate, and at times significantly, from the benchmark. Advisors should take responsibility for recommending funds that deviate significantly from their benchmarks. Investors should be primed to ask questions where a fund has deviated significantly, positively or negatively. Indeed, the more sophisticated the fund the more explanation that will be needed.
Benchmarks are important not just for the comparisons they provide, but also because they should back an advisor’s recommendation.
A benchmark should be compared against a fund over its lifetime so that an investor can see the long term comparative return profile of the fund. It should also be compared on an annual basis against the benchmark in terms of returns relative to the benchmark – so the investor can get to see how the fund performs on an annual basis relative to the benchmark. Funds that deviate from the benchmark will see both +ve and negative deviations and will allow a representation of when a fund has significantly out performed the benchmark.
Now, in a sense a benchmark is the eye of the needle for the investment industry. If the consumer is meant to take responsibility then they need to understand the performance profile of the investment and the risks they are taking to access that profile. The ye of the needle is this: responsibility, whoever holds it, has a consequence and you cannot squirm out of this truth.
The responsibility of the relationship
The amount of information required to make a fund recommendation is far different from the amount of information needed to accept that recommendation. The problem in the retail advisory segment is that this responsibility is a fuzzy one and not in truth clearly defined: the advisor is responsible for collecting a simple set of parameters and delivering a “suitable recommendation” that fits those parameters but is not overtly responsible for much else. The big deal is that the investor is responsible for the decision to transact and this means that the investor needs to understand the physical characteristics and the risk and return profile of the investment. This requires at the very least a benchmark.
On the other hand, if the investment advisor is responsible for defining the asset allocation profile and making sure that the vehicles selected are cost effective and efficient in managing risk and return, then they will have already done the necessary performance analysis and compared the performance of the fund to an appropriate benchmark. The investor is not responsible for the fund selection, neither is the investor responsible for the process that defined the asset allocation, nor the process that properly assessed the investors risk aversion and investment preferences. The investor will be responsible for accepting a specific risk and portfolio profile, but the advisor will be responsible for the integrity and transparency of this process. In this case, the point of sale document for the investor will explain in simple terms the asset allocation, the risk profile, management and other expenses and the performance of the investment. The investor does not necessarily need to fully understand each individual fund, as it is the risk and return profile of the overall portfolio that matters. The funds are merely building blocks and it is the advisor’s responsibility for making sure the blocks fit.
In this situation there is no reason why a fund should not have its performance compared against an appropriate benchmark. The recommendation is the asset allocation and the risk profile of the asset allocation which should more or less directly reflect the total asset allocation and risk profile of the individual investments.
If the advisor is not responsible for the above, but is only responsible for collecting a limited set of information (basic attitude to risk, time horizon, investment objectives for the capital at hand, expertise) and recommending an investment which fits these parameters, and regulation deems the investor to be responsible for the investment decision, then we are in a whole different ball game. The investor has to be able to relate the fund to the overall asset allocation, has to be able to work out what is the asset allocation and performance profile and whether or not the fund is an appropriate way of gaining the appropriate asset allocation. Again, the less expertise the investor possesses, in this paradigm, the simpler the recommendation.
The proof of the investor “expertise” pudding is in their ability to understand a benchmark (etc). How many investors are put down as experienced investors yet are unable to understand what a benchmark is?
The fundamental objective of a fund facts document in a transaction service process
In truth, what more information in the fund facts actually does is to create a greater burden of responsibility on the advisor to justify the recommendation, and provides the investor with more information as to the nature of the investment and its relative performance.
Information on what benchmarks are and why they are important also need to be provided to the investor, so that over time they have the opportunity to learn about the fund and the impact of costs and differences in style and asset allocation on performance. If the regulators are determined to make the transaction system work then they will have to provide access to more information.
A benchmark is also rod for the back of the advisor and will require advisors to justify their recommendations. This should also help police “suitability”. Over time it should become difficult for advisors to recommend funds that have consistently and significantly under performed their benchmark or that are too complex for the investor to understand.
Without a benchmark then there is no way you can make an assessment over costs or role and the effectiveness of the investment. You are buying an investment blind.
The lack of an effective benchmark is not the only thing missing
We really need a comparison in terms of asset allocation, yield and price of the fund relative to the benchmark. So that investors can compare the difference between the two and have the opportunity to ask questions. I would also suggest that the style classification of the fund is also provided.
In addition we need a summary statistic which shows the risk of the fund and the risk of the benchmark and a simple explanation as to how much riskier in terms of price movement the fund is relative to the benchmark.
Again, all the above reduces to a much simpler document if all the advisor is recommending is broad generic asset allocation funds. The complexity comes where the advisor is recommending active and more sophisticated strategies.
At the point a fund recommendation is made, the risk that we are talking about is the risk of the fund itself relative to the benchmark, which links back to the asset allocation profile of the portfolio. This should not be the point at which we enter into a discussion about the relative risks of cash, bonds and equities. The fund selection decision is not the portfolio allocation decision.
One of the problems with the retail transaction service process is that the two risk processes are often conjoined. This means that discussions of broader risks such as the risk of capital loss and the ability of assets to meet needs in the event of significant market and economic risks are often lumped into the fund selection decision.
The fund facts document should not be discussing wider risks, but the specific risks of the investment relative to the benchmark. In this sense, standard deviation and the Sharpe ratio are acceptable measures of risk.
So the benchmark is not there to educate investors over the risks of investment per se, but the risks of that particular investment. So take out the benchmark from the fund facts document and you take away the opportunity for objectivity in the decision. You are effectively saying that risk and return (and costs) are not important.
Most investors are not going to understand a benchmark for the first time and without direction from their advisor. So start off with simple generic funds or spend time with the investor explaining the rationale of the selection and the information provided in the benchmark. Over time, investors will either learn about benchmarks and allow for a more diverse set of investments to be recommended, or not, in which case fund selection and asset allocation profiles will be broad and simple – the fund will be the benchmark.
If regulators are committed to keeping the current system in place investors need to be educated and benchmarks and other necessary information have to be provided in fund facts and other point of sale documentation.
Questions and answers bearing in mind the above:
1a) Why do you think benchmarks should be published in Fund Facts?
Answer: In order to make sure that the fund is a competitive and cost effective method of allocating to a specific component of the portfolio its risk and its return need to be compared to an appropriate benchmark. Taking out a benchmark is asking an investor to make a decision that a professional would not be able to make. There is no informed decision.
1b) Is it sufficient to have them available, or at least cited, in other regulatory filings such as the Management Report on Fund Performance, produced semi-annually?
With respect to being cited or referenced they need to be provided before the sale because they need to be discussed with the advisor. Whether this is in a supporting technical sheet is not important. Regulations state that the investor is responsible for the decision. It is these regulations which effectively enforce the requirement for this information to be provided before the sale.
2) What BM information would you propose be required in Fund Facts? By this, I mean what time periods? Calendar year by year? Compound annual returns over 1, 3, 5, 10?
Answer: a simple graphical performance of the fund relative to the benchmark since inception + a bar graph of annual performance of the fund and annual performance of the benchmark. This should allow investors to see the compound impact of asset allocation and cost differences as well as annual differences that may impact the cumulative performance. One year of extraordinary performance should be picked up.
3) Are you concerned about end-date bias in the reporting of returns vs. BMs?
Answer – discrete annual performance comparisons to the benchmark should highlight significant annual deviations.
Moreover, the fund selection decision is not the point at which a discussion as to the relative risk and return merits of cash, bonds, equities or other investments should take place. In other words, a strong 5 year performance from a market low should not influence an excess allocation to equities or a heavy allocation to a very risky fund. Such an outcome would be attributed to a flaw in the advisor’s suitability process.
Such a flaw is unfortunately symptomatic of a transaction driven service process where the fund selection and the asset allocation decision are comingled. The over allocation to investments with strong recent performance is likely to be advisor influenced and the only way that much of this risk can be circumvented is via the introduction of best interest standards which should hypothetically separate out the asset allocation and the fund selection decision processes.
4) What about risk benchmarks, such as standard deviation for a particular index? Should this type of info also be in Fund Facts.
As I have said, the fund facts document should not be the time or the place to discuss the various merits of the cash, bonds and equities, but an assessment of the specific risk and return profile of an investment.
Standard deviation and a Sharpe ratio for instance would be appropriate. I would prefer that a bar chart of the monthly standard deviation of a fund relative to its benchmark also be provided on a year by year basis so the investor can see the relative risk of the investment against the benchmark – the objective being to create discussion and to generate questions.
Education should also be provided on what standard deviation means. Again, to exclude standard deviation is asking an investor to make a decision that a professional would not be able to make. Standard deviation relative to an appropriate benchmark provides a summary of the how the fund is impacted by risk relative to its benchmark.
Investors who do not understand the complexity of standard deviations should be recommended simple, broad generic allocations to the benchmark.
Obviously you can see where this is all leading
If regulators are determined to keep the transaction service process in situ then it needs to become more sophisticated with investors screened more carefully for what they currently do and do not understand. This will influence the complexity and sophistication of the recommendation meaning that those who understand benchmarks and standard deviation may be recommended investments where this is a requirement to understand the investment. Otherwise, broad generic index allocations that match simple broad asset allocation profiles should be recommended. Education is the path along which investors can progress to more sophisticated solutions, if this is considered appropriate.
But I know where my preference resides. Stop this charade now, introduce best interest standards and get rid of embedded transaction returns for advisors.