One Risk Score Should Not Rule Them All (Goals, that is)

Mar 19, 2015

I really enjoyed Raef Lee’s latest post on modular planning. (But not that whole British thing. I think he is faking Soccergoalit, accent and all.) It seems that modular (and goals-based planning) have been very hot topics recently and I have spent a lot of time discussing and debating their finer points. I think that the traditional comprehensive plan, with all the time spent gathering data, crunching numbers, and assembling impressive thick binders of analysis and reports actually pushes more people away than it attracts.

Lately, I have really been discussing how investment risk tolerance is, in my opinion, wildly misstated in most software and planning questionnaires. So this is my push for modular risk tolerance questionnaires and software.
Case in point: In our recent white paper, The Next Wave of Financial Planning, Raef, Michael Kitces and I argued that changing demographics, “robo” advisors, and increased collaboration between advisor and client is transforming our business. Financial planning is really becoming the value hub of an advisor’s business. One of the exciting areas of our business right now is the growth in technology that allows advisors to co-plan in real time with their clients. The challenge, however, is that with all the new planning technology, most firms (if not all) still think that client risk can be summed up by a single risk score.

In 2004, when Dan Nevins wrote his groundbreaking white paper, “Building Goals-based Investment Strategies,” I heard him give an example of how the investment community treated investor risk tolerance. Dan would say something like:

“This morning, I poured myself a cup of steaming hot coffee and some cold freshly squeezed orange juice. If you created a questionnaire about my beverage choices, asking me to rate on a scale 1 – 10 the temperature choices that I prefer, then added them up to create a “temperature choice score,” you would deduct that I like all my beverages at room temperature or just score it a ‘5.’”

Of course, this is an extreme example, but why then would an online questionnaire or planning tool try to come up with one score for investors? As consumers, we all have different levels of risk in mind for different goals, right? More importantly, would a robo advisor be able to differentiate or push back on a risk score if a millennial scored a “low risk” portfolio, even if he/she was investing for retirement some 30-40 years away?

Who gets to define risk anyway?
When I think about risk in a typical family’s investments, I don’t think about the Dow or the S&P. (I guess no one should worry about the Dow anymore; AT&T doesn’t.) Instead, I think about the risk of not achieving goals. Let’s use my family as an example:

  • In our boys’ 529 plans, the risk is not having enough to pay for our sons to go to college
  • In our short-term investments, liquidity is key and stability has to be there
  • In our retirement, what will our lifestyle look like and will we have enough?

If you think about it, one risk score probably won’t cover the total needs of the Anderson portfolios. If, in Dan Nevin’s example, I scored a “5,” the longer term portfolios would probably have too much cash to achieve the higher returns needed, and the shorter term portfolios may have too much volatility if I needed to withdraw money in an emergency.

Adding thought to technology
So if technology is not going to make it easy, advisors need to add the human touch to the portfolios. Advisors can use the scoring guides to see if all the accounts together reach the target allocation, but you should consider a modular approach to determine risk, rather than an almighty risk score.

Next time you are sitting with a client and determining allocations or portfolios, try assessing the risk at each goal level. If the software (or answer sheet) says they are a moderate investor and the pool of assets has a 20-year timeframe, remind the client that the questions pertain to a specific pool of money, rather than their total investments (and document the answers so you both can review it often).

Clients tend to think in buckets (or pools of assets) for specific purposes. Traditional investment theory and technology tend to think in terms of a single overall score to determine a client’s risk. Put the human element (you) into the mix, and I think you have the best of both worlds.

Image courtesy of

Share Button
John Anderson

John Anderson

John Anderson is the creator and lead author of Practically Speaking blog and Managing Director of Practice Management Solutions for the SEI Advisor Network.

Learn More About John Anderson



HNW Whitepaper

Recent Tweets