As Monte Carlo analysis becomes increasingly popular in retirement plans, financial planners are talking more and more about the probabilities of a client's success or failure. Yet in the end, most planners evaluate client goals, look at the probability of success (defined usually as not running out of money), and the client makes a decision about whether they like the result or not. Oddly enough, planners rarely take the next logical step: ask the client what probabilities they would like to see, and use that risk/success metric to determine what the other answers - such as retirement spending or the retirement year - could be.
The inspiration for today's blog comes from a demo I saw yesterday for retirement planning software analytics put together by Fiducioso (subsequently renamed as Income Discovery). They are doing some interesting work trying to blend together the very complex analysis of retirement planning's moving parts - not just things like retirement age and spending level, but also the use of systematic withdrawals, laddered TIPS, annuities, etc., and integrating those together with multiple types of accounts (taxable vs. IRA vs. Roth IRA) to evaluate the asset location strategies and the order of account spend-downs.
One thing that struck me in the Fiducioso demo, though, was remarkably simple, yet incredibly unique from what I've seen of retirement planning software. Not only did their software let you input things like spending levels and a retirement year, and find out the probability of success, but they also let you enter a targeted probability of success, and find out what spending level would be supported.
Such a simple concept, yet I realized that I have never seen a tool do this anywhere else! For all other software, the client's spending and time horizon goals are entered, a probability of success is determined, and if that is "too risky" then it is up to the planner to keep re-running new versions of the plan with different inputs until the outcome is acceptable.
Not that it's bad to drive to a result based on a client's goals for spending and retirement years, but many clients would rather just know "when can I retire" or "how much spending is safe" - in other words, in some cases it really is true that the primary goal is "show me a result that is safe" (i.e., has a low probability of failure) and that is the retirement the client wants to live. Yet when our goal is to live retirement with no more than a certain level of risk, answering how much spending and what of retirement will support it is a remarkably difficult question to address. It's just not supported by today's software.
What do you find drives the discussion in your retirement planning with clients? Is it the spending and retirement age? Or do at least some of your clients drive to their goals from the base of their probabilities of success?