Some experts, most notably Michael Kitces who recently posted this blog The Sorry State Of Risk Tolerance Questionnaires For Financial Advisors, are increasingly critical of risk tolerance questionnaires (RTQs), a tool many/most investment advisers and firms use when constructing portfolios for investors. Answers to the RTQ, among other things, helps advisers determine how much investors should allocate to stocks, bonds and cash, and help advisers categorize investors as conservative, aggressive, and the like.
But Derek Klock, an associate professor of practice in finance at Virginia Tech, thinks that RTQs are getting a bad rap, that not all RTQs are created equally, and that new RTQs are better than the RTQs of yesterday. Klock addressed our questions about RTQs by email.
Q: So, why, from your point of view, are experts such as Kitces and others being so critical of RTQs? Are RTQs fair game for criticism?
A: Let me begin by saying that I have tremendous respect for Michael Kitces as well as Dr. Michael Finke and others for their thought-leadership in this area. I’m just honored to join their conversation. The simple answer as to why I believe that Kitces and other are being critical is that criticism promotes change; and yes, risk tolerance questionnaires (RTQs) deserve some criticism as many are either outdated or outmoded in their construction. They were designed with the best intention and knowledge at the time, but the professional understanding of this topic has advanced.
But I think that other professionals’ blanket criticism of the general usefulness and application of RTQs goes a step too far. Psychometric testing of personality, beliefs, and attitudes is messy and ever-evolving, but some of the more accurately designed RTQs when administered, interpreted, and applied correctly offer a great starting point for a planner/client conversation concerning risk and volatility.
Q: You've suggested that there are good RTQs and bad RTQs. How can investors tell the difference from a good RTQ and a bad RTQ? Can you give us some examples of bad RTQs and good RTQs, and what makes them good and/or bad?
Great question, but let me first re-label the good and bad to older and newer. Many RTQs have been around for a long time; and seemingly older model RTQs, just like cereal, don’t tend to be replaced. We just add to the list. A big issue with the internet - old stuff seems to hang around long after its expiration date. Back to your question.
Let me try to provide some insight as to what makes up a good RTQ and hopefully let the readership make more informed decisions from there. Older model RTQs were very outcome or goals driven – in sort of an end-must-justify-the-means way – in that they asked about what the client wanted their portfolio to do: provide current income, outpace inflation, provide a stable value, etc. Newer model RTQs predominantly are meant to measure two things: an investor’s risk tolerance – (the person’s emotional reaction to uncertainty), and/or an investor’s risk capacity – (the portfolio’s financial ability to absorb losses without affecting the client’s current lifestyle or future goals).
Risk tolerance questions are generally more hypothetical and may even take a form that is completely non-financial in wording. But that isn’t to say that they should ask about an investor’s hobbies – like the often mentioned sky-diving or other risky habits questions. However, the questions may ask about the client’s feelings or attitudes regarding other situations where there is a binary choice, financial or not, between a known outcome and one of ambiguity. These questions are meant to gauge how comfortable with, or knowledgeable about, the investor is with uncertainty.
In contrast, risk capacity questions typically center on the time horizon until the money is to be used, the amount of loss that could be withstood without meaningful financial effect, or how willing the investor is to a change their lifestyle should a negative outcome occur.
So, what is bad? “Bad” exists on two levels – poor design and poor interpretation. If an RTQ continues to use questions that have been proven to have little predictive value then that would be “bad.” And while there has been precious little back-testing of the reliability and validity of RTQs as a whole, there have certainly been questions vetted enough to prove worthlessness. The aforementioned sky-diving questions falls (pardon the pun) into that category.
Additionally, questions centered on gambling could be less useful as, while they deal with money, the gambler may be engaged in risk-seeking behavior not for the expected financial gain but simply for the thrill of participation. In economic terms, the utility they receive from the gamble is greater that the disutility they suffer from the financial loss. It is still a rational behavior, as explained by economic theory, just not of predictive value for measuring financial risk tolerance.
However, the one thing that I can unequivocally offer as a sure sign of poor design is that the risk category outcome should not change dramatically on the basis of one question. For example, a questionnaire may have a single question pertaining to the risk capacity of the taker, while the other questions focus on the risk tolerance. If the questionnaire provides qualitative categories for the taker’s risk preference, the resultant categorization should not change dramatically based solely on the taker’s risk capacity.
Then there is poor interpretation. If a financial professional is using a RTQ that he or she believes is constructed to measure risk tolerance, but it is actually better suited to measure risk capacity then the results will potentially be misinterpreted. It is extremely important that the administrator and interpreter of the RTQ be knowledgeable about the differences in risk tolerance/aversion, risk capacity, loss aversion, and risk perception. All of which leads directly into your next question concerning implementation.
Q: Could there be cases where investors use a good RTQ, but not necessarily work with an adviser who knows how to implement the information gleaned from an RTQ? What can investors do to make sure they are working not only with a good RTQ but a good adviser?
A: Of course it is possible. Finding a good adviser, one that not only uses the most valid, up-to-date RTQ available but also has the knowledge to interpret the results, is paramount in this discussion. However, it also needs to be said that interpretation and implementation of psychometric testing is really tough. We have talked at length about the input and composition of RTQs; let’s now discuss the output.
Some RTQs result in a scaled numeric score, say 27 on a scale of 10 to 40. Others offer a categorization class, such as moderately-conservative, on a scale of conservative to aggressive. Still others provide both. But in all cases it is up to the financial professional to take the number and/or category and turn that into a portfolio with appropriate risk and return levels matched to the client. This is why some RTQ and/or risk-profiling providers attempt to automate the interpretation and implementation by helping the advisor make the leap from the questionnaire to the portfolio.
A notable example of this approach is Riskalyze. This platform asks the clients a series of binary choice questions to determine their risk profile. The output then takes the form of a number (scaled 1 to 100) and a portfolio with a risk score (standard deviation) that matches the client’s volatility preferences.
As to the second part of your question concerning a good adviser, I would refer to the CFP Board of Standards “letsmakeaplan.org” website as it contains a search engine for finding CFP professionals who have met at least a minimum educational requirement for practicing financial planning, taken an exam on that knowledge, and continue to update their education. Additionally, the site provides a good list of ten questions everyone should ask when vetting a financial planner.
Q: One criticism of RTQs is that they tend to focus on the short term, and investor’s answers change with the markets. So, critics say what’s needed is an RTQ that doesn’t change with bull and bear markets but rather focuses on the long-term. What say you?
A market-neutral RTQ? Why would we want not to simply periodically retest? Some research supports the assertion that investor’s risk tolerance changes with current or recent market conditions. Some research suggests that the relationship to recent market conditions is much greater in declining markets than improving ones. Still others suggest that the riskier those conditions the more myopic (short-term, near-sighted) the investor gets because of increased vividness. I’m not positive that there is a good solution, but I’m also not sure if this is really the problem. According to the research, many factors influence investor risk tolerance over the lifespan such as age, education, and wealth. Why would we not expect that risk tolerance is going to change? Advisers understand that when markets decline they need to stay in closer contact with clients and possibly adjust clients’ portfolio allocations. Why not take the opportunity to re-measure the client’s current risk tolerance to make the most accurate adjustment possible.
I might add that there has been some recent discussion attempting to focus on investors’ risk perception (how they see their immediate environment) separately from their risk tolerance. Maybe this will turn out to be the key to separating an investor’s base-line risk aversion, from the current sway of circumstance – market forces, mood, age, wealth, etc. However, until then rather than try to apply a single test over an extended period, advisors should consider periodic retesting.
Q: Part of the problem with constructing portfolios is that many investors and advisers for that matter don't have a clear understanding of such terms risk aversion, loss aversion, risk tolerance, and risk capacity. Can you give us a plain English definition of those terms and how each and all play into constructing an investor's portfolio?
Ok, wow! Big question! Obviously these terms are interrelated but are also independent. It may be easiest to first categorize these terms as pertaining to interior finance factors or exterior finance factors. The interior finance deals with the person, the exterior finance deals with the money.
First the interior finance factors. Most define risk aversion or tolerance (these modifiers are used interchangeably) as the emotional, psychological, or physiological response people experience when faced with an uncertain outcome. It is that pit-in-the-stomach, cold-sweats, heart-racing, butterflies feeling you have riding a roller-coaster for the first time. The reason this is important to understand is that an adviser wants to know where the market-volatility induced panic-point is for his or her clients.
Loss aversion, also a physiological response, is a term that gained prominence as a result of Daniel Kahneman and Amos Tversky’s work in behavioral finance. It is the concept that a loss of a certain value feels worse than a gain of identical value feels good. This leads people to become risk averse, because they would rather not lose what they already have (and feel badly) in an attempt to gain something that they do not (and feel good). This element of the risk profile I believe is extremely important, as it has been shown that people are generally more likely to take risk to avoid loss than they are to seek a gain. This means that acceptance of risk is not static, or constant, and therefore risk tolerance is often dependent on situation.
Finally, the exterior component – risk capacity. This is a much more tangible, quantifiable part of a risk profile and has both a near-term and long-term component. The near-term component, most believe, deals with the effect of market volatility on the portfolio’s ability to financially support the client’s current lifestyle. Said in simpler terms, how much could a portfolio lose in value before it starts having a profound effect on a client’s financial means. This is one of the major reasons so many planners are proponents of an emergency fund or cash cushion. It protects the client’s current financial lifestyle from near-term market volatility. (This tactic also may help with risk aversion as it may distract the client from the market volatility.) But many advisers also consider risk capacity to have a longer-term component as well. This portion considers whether or not a client has time to let the market, and the portfolio, recover the lost value before it effects goal attainment. In summary, capacity deals with dollars and time rather than with feelings and attitudes.
But while these concepts are all important and all play a role in the overall risk profile of a client, it is extremely important that an adviser not confuse a client’s risk tolerance with the client’s portfolio risk capacity. While clients may have the time to recover from a loss, if they exit the market due to sheer panic and desperation because their portfolio was too risky and the market change too much to bear, then all the time (and market recovery) in the world will do them no good.
Q: Some say there is a difference between risk tolerance and risk aversion, but you seem to suggest they are synonymous, one in the same. Why is that? Isn't there a difference?
A: I don’t believe there is a difference. The classic economic position is that risk-tolerance and risk-aversion are two sides of the same coin. Said another way, your risk tolerance percentage and your risk aversion percentage should add to one. Of course all rational investors are going to be risk averse on some level, some exceptionally so. I could as easily call this latter group highly risk averse or slightly risk tolerant. I think people have developed a natural inclination to refer to a dominant characteristic, for example, right-brain dominant rather than left-brain submissive. Additionally, in the context of risk, my personal belief is that the use of the two terms came about because of framing. Risk-tolerance seems to have a more positive connotation than risk-averse.
Q: What else do investors need to know about RTQs that we haven't covered?
Yes, thank you. I would be remiss if I didn’t mention that while RTQs aren’t the end-all, be-all of portfolio construction, portfolio construction is hardly the only useful outcome of RTQs. They could prove very helpful in understanding client attitudes or behaviors in most other areas of financial planning - debt management, tax planning, insurance planning and estate planning.
I would close by encouraging all advisers to periodically review their risk assessment instrument, of course seeking the guidance of their compliance team, to ensure that the tool chosen is the best one available and that they are making the best use of the information the RTQ provides.
Robert Powell is editor of Retirement Weekly , published by MarketWatch. Get a 30-day free trial to Retirement Weekly . Follow Bob’s tweets at RJPIII . Got questions about retirement? Get answers. Send Bob an email here .