Toward Normative Standards in the Advice Profession
Normative standards can be broadly defined as a consensus among givers of advice on what advice or service best suits a particular set of circumstances. Unlike most other professions, those standards don’t exist among financial planners. That may soon change.
If you catch a cold and go to a doctor – any doctor – you’ll get much the same diagnosis and advice. There are procedures in the medical community to determine if you have cancer, and everybody who is diagnosed with a certain type will get roughly the same procedural recommendations.
This is true in other professions as well. The advice you get from compliance attorneys is boringly similar, and the accounting profession is not (in theory, at least) allowed to deviate from generally-accepted accounting principles. Go to 10 or 15 dentists and you won’t receive wildly different recommendations when it comes to maintaining the health of your teeth.
Compared with those professions, the financial planning world is a wild west when it comes to what Dr. Nathan Harness, professor of financial planning at Texas A&M University in College Station, TX calls “normative standards.” A client who walks into 10 different planning offices is likely to get 10 different portfolio recommendations, very different evaluations of their current will and living trust arrangements, and a wide variation of advice about whether (or how much) to convert from a traditional to a Roth IRA. Some advisors will provide recommendations on long-term care insurance, medical insurance, disability coverage or student debt. Others will consider those areas outside of the financial plan.
Dr. Harness, who serves on the Council for Education at the CFP Board, has participated in the task analysis process that the Board goes through periodically, which he describes as a neural networking exercise where many advisors contribute to a consensus on the tasks that are routinely performed in an RIA office. This led him to consider the much broader question: When those tasks are performed – say, reviewing the client’s will – what does a competent advisor look for? What will the advisor recommend based on what she finds? What, ideally, should she recommend under different client circumstances?
“A lot of people in our industry came out of a specialized area,” says Harness. “As we moved to a more synergistic, generalized understanding of planning, the integration of each one of those never happened. People struggle to release their specialized areas.”
So where would we start to build strong normative standards for the profession? “We have to start with a clear policy of what financial planning actually is, before we get into the actual practice-level advice,” says Harness. “What do we do with clients? What are the procedures in which we operate? What is our policy?”
The CFP Board has identified 72 core topics in financial planning, broken out into domains, and Harness would start from there. “In each one of the domains of financial planning, the questions are: What are our procedures? What are the policies in each one of these areas?”
As an example, Harness has focused on estate planning. He noticed that the students in his capstone course and their mentors were all over the lot when it came to whether they looked at a client’s will – and if they did, what they looked for. “I talked with Roger Pine [co-founder of Holistiplan, a software company] about it, and he said that we could start with the checklist that he would go through when he was working with clients,” says Harness. “He went down through a list of things that he would look at.”
Having checklists specific to each domain in financial planning would, all by itself, will start normalizing planning advice across the profession. The next step would be to poll the financial planning community, asking advisors what they would recommend if they found this or that in the will, or if they didn’t find things that ought to be there. Some of this would be relatively uncomplicated – after all, nobody disagrees that the will should name the heirs. Other issues might be more complex, such as whether the will names all of a client’s assets, or whether some assets are included in a living trust. Who would be responsible for checking this? Should there be a power of sale in favor of the estate executor? Which clients should set up a testamentary trust, referenced in the will, to protect against estate creditors?
Harness tried to tap into the wisdom of his students’ mentors, but the initial process was unsatisfying. “What we discovered really quickly is that the questions we were asking were too general,” he says. “As a result, we were getting really general responses. I learned that you have to ask very specific case-level questions about these issues, and ask a bunch of small questions instead of a big one.”
Would a bigger sample set be practical? Who would conduct detailed research into crowd-sourcing the best advisor advice throughout 72 core topics? Harness notes that the financial planning software companies are responsible for the normative standards, such as they exist, in the very narrow sphere of retirement-sufficiency analysis. In one example, MoneyGuidePro (now part of Envestnet) tapped the considerable expertise of Harold Evensky, of Evensky, Katz/Foldes, to create its calculation engine and default assumptions. Harness imagines that future software firms will expand their inquiries into the best ideas of dozens, hundreds, even thousands of professionals, and gradually sort out the best two or three answers for each presented set of client “symptoms.”
Interestingly, that process is already underway with Pine, of Holistiplan. Pine and Holistiplan co-founder Kevin Lozer do what most software developers do: They listen to what their customers want out of the program and add features. But what’s different about Holistiplan is that the software isn’t a practice management convenience like CRM or portfolio reporting software. Holistiplan uses a specialized optical character recognition (OCR) process to read a client’s recent tax return, and pull out the information into a report that puts, in readable format, the client’s total income, total tax paid, effective rate, taxable income, marginal rate, itemized or standard deductions, short-term and long-term capital gains, qualified and ordinary dividends, a carryforward loss (if any) and the next year’s safe harbor tax payment.
Then (this gets to the normative standards) the program makes a number of observations about the client’s tax situation. The report notes that the tax form shows an underpayment penalty that could have been avoided. The client has business income; does that business have a retirement plan? There’s no HSA deduction; has the client considered setting up an HSA? At the current tax bracket, the client might consider a Roth conversion.
Lozer and Pine started with the observations that they would have made in their planning careers. But their users quickly offered suggestions about what they would have observed and discussed from certain tax data. That safe harbor tax payment was recommended by a user. Others recommended that the software extend into modeling future tax rates and obligations based on this year’s return plus inputs from the client about what would change in the coming year or years. (Inheritances? Bonuses? Exercising stock options?) Then somebody recommended that the program calculate the optimal amount that a client could convert from a traditional to a Roth IRA without moving into a higher tax bracket.
That, of course, opened up a new can of worms. “If you do the Roth conversion and it adds to your AGI,” says Pine, it starts impacting things other than just your marginal federal tax rate. It affects how your medical expenses are deductible; it raises your Medicare Part B premiums, and how your Social Security is taxed. We had to get it down to where it shows the total tax impact of the conversion, and what would happen in each of those areas, in in total, if you added $1,000, or $2,000 to the conversion, so you can see which areas are impacted and by how much.”
Another suggestion from a user: Add the clients’ birthdays as an input alongside the OCR data retrieved by the software.
“When we built in that very simple feature,” says Pine, “it opened up this whole other set of recommendations and observations that we could spit out.” The program now tells advisors which clients should prepare for taking their maximum Social Security benefits, when they should start signing up for Medicare, when they have to start taking required minimum distributions, when they qualify to make qualified charitable contributions. “Initially, we input 15 different threshold ages that would trigger something that you might talk with a client about, just off the top of our heads,” says Pine. “Then we introduced the feature, and started getting feedback from our users, and literally within eight hours, I had 30 individual notifications that an advisor should talk with a client about at a given age. And they even caught one of my mistakes. In eight hours!”
Some of them are quite subtle. One Holistiplan user pointed out that the Medicare Part B premium – the so-called IRMAA adjustment amount – is calculated based a client’s two-year-old tax return. So three years before a client turns 65, the advisor might look at tax planning opportunities to reduce the IRMAA surcharge when the client signs on to the Medicare system.
Holistiplan regularly interacts with 600 users, who are encouraged to go to a designated area on the company’s website and offer their views of what the suggestions to clients should be based on – and virtually all of the suggestions sound like the normative standards that Harness is talking about. “Fundamentally, what we’re talking about is all of us [the users] putting our heads together to say: What is the best possible way to handle this fact pattern?” says Pine. “The next stage is: Let’s talk about qualified charitable distributions (QCDs). How do you go about describing and explaining the options to a client? I know clients who give monthly donations to their church; how can you adapt the QCD concept to their charitable intent?”
Pine and Lozer are working on additions to Holistiplan that will incorporate best practices in cash flow planning, and once again there will be enthusiastic input from users. Pine likens the process to open source software, where many people contribute to the product, and many people ultimately benefit as it improves. “It feels like the early days of NAPFA,” says Pine, “where you’re sitting at a roundtable, except now there are 600 people at the table, sharing their collective knowledge with all of us.”
Harness imagines a future where somebody, somewhere, will start building the checklists that can go into a central data repository, where any advisor will be able to pull out the best checklist on any aspect of planning. And then, he says, somebody, somewhere, will start tapping the wisdom of the crowd, asking about the recommendations about this, and then that, and then another item on this checklist, and the next one. No one advisor will have all the answers, but many advisors will have a piece of the overall “best” answer. That growing body of shared information, the results of an active neural network, could become the raw material that any ambitious software company could draw from.
“Ultimately, the whole idea of normative standards is, just like any other profession, to give people the best and most helpful advice possible,” says Harness. “There may not ever be one ‘best’ solution to any particular client situation,” he says, “but as we get more and more people contributing, we’ll get closer and closer to some optimal recommendations” – which, he says, will have to evolve if financial planning is ever to become a profession.
Bob Veres' Inside Information service is the best practice management, marketing, client service resource for financial services professionals. Check out his blog at: www.bobveres.com.