This is the third part in a series of posts that discuss strategies to make retirement plans “better.” If you haven’t already, please read the intro which discusses what makes a plan “better,” Part 1 which describes why financial planning models may limit your retirement, and Part 2 which describes how bucketing your investments such that equities rise over time enhances your spending and legacy prospects.
This section focuses on how we frame “risk” in a financial plan.
Retiring is scary. You go your whole life with an income to pay your bills, and then one day that income is just gone, and you have to hope that all that hard work and saving was enough to last the rest of your life. It’s no wonder risk of running out of money is on the forefront of every retiree’s mind.
Research from Blackrock suggests that the majority of retirees plan to never spend down the principal in their savings. After about 18 years of retirement, most people have more than 100% of what they had when they first retired.
There’s nothing wrong with this, per se, but it’s rarely stated to us as a goal. In fact, many clients have told us that they’d like to “die broke,” meaning enjoy every dollar they’ve accumulated for themselves. This makes sense! Most client’s children will be retired themselves when they die, and likely won’t need the money.
We have a mismatch here. Clients want to spend more of their money, but many resist spending down their principal.
It goes back to risk. “I’d spend more but I don’t want to run out of money.” “I’d give more now but I might need those assets at some point.”
Our fear is that traditional financial planning models are reinforcing client’s fears. In Part 1 we talked about some of the assumptions that make plans too conservative, but now I want to talk about how risk is defined in financial plans.
The financial planning industry has gravitated to a single number that defines the riskiness of a plan, Probability of Success. That is, if we run 1000’s of scenarios, and 800 still have money at the end, that’s an 80% Probability of Success. The stat itself isn’t terrible. It does a good job of capturing the volatility of investment returns, providing different sequence of return scenarios, and accounts for the unique spending patterns of different clients. The problem is in how it’s framed.
Why a 50% Probability of Success is Actually the Ideal Target for Retirement
First, probability of success being on a scale from 0-100% leads people to target a level consistent with a good grade in school. A “great” plan should be an “A” or above 90%, a good plan is a B or 80%, and an ok plan is C, or above 70%. Most financial planning software makes this worse by highlighting high grades with one color, and low with another. This reinforces in client’s minds that unless they are achieving a “passing” grade, they are failing.
Framing it this way completely misses the mathematical reality of what probability of success is measuring. If I were to solve for the best estimate of what a client should spend to accomplish their goals, I’d use time value of money math (what’s used in loan calculations, investment valuations, bond pricing, and more). In fact, this is how retirement planning was always done prior to the proliferation of probability of success. Without going through a long explanation of why, solving for the appropriate level of spending with time value of money month is the equivalent of solving for a 50% probability of success.
How Adjustments Make All the Difference in Retirement Planning
The main issue with probability of success is that it’s measured at a single point in time, and assumes you are unwilling to adjust regardless of how bad things are going. Don’t you think if you saw your wealth down 30%, you’d tighten up the budget a bit so it can recover?
As the rest of this article will go on to show, if you are willing/able to adjust over time, it’s impossible to run out of money. As such, what we should be solving for is the probability of making an adjustment, which is 1- probability of success (or 30% if you had a 70% probability of success).
I said before that a plan with a 50% probability of success (aka 50% probability of making an adjustment) is, in fact, the "best guess" for how much someone can spend in retirement to accomplish their goals. That may initially strike some readers as wrong, and risky. If we ran your plan, and then didn’t look at it for 30 years, I’d agree with you!
In the real world, we run the plans at least every year. If you are behind, that means either spending was higher than we planned for, or returns have been worse than predicted. When we see that, we can either cut back spending until the plan gets back on track or hope for better future returns. Hope is not a strategy, so let’s focus on spending.
There’s two ways to adjust spending, abruptly or gradually. The method you take will depend on how much of your spending is discretionary. If a large portion of your spending is discretionary, meaning you could forego it for a year or so, you might just do it all in that year. Skip the vacation, don’t remodel your kitchen, etc. This is what people do naturally to some degree anyway, and for some people it works.
The issue is it’s not realistic for everyone. Depending on the severity of the market crash/recession, the model might want you to cut spending 20-30%+ in one year. That’s why it's reserved for those with lots of “discretionary” spending. A better approach for most could be to adjust gradually.
The Power of Flexible Planning: Adjusting Spending Year by Year
Let’s say we have a client with $2 million and a 30-year time horizon. We assume the portfolio returns 5% real, with volatility consistent with our moderate benchmark (9.5%). The client’s goal is to try and spend as much as they can in retirement and “die broke,” which we hear often. Using time value of money math, I find that they can spend $116k per year, growing with inflation. As I said before, this is consistent with a plan that has a 50% probability of success.
Since the client wants to die with nothing, we want to try to hold probability of success constant throughout the plan, (meaning don’t let it drift up). So, if they dip below 40%, we decrease spending gradually until they are back within our target window, and if they get above 60%, we increase spending until they are back in our target window.
It's not an easy thing to model, (since it requires running monte carlo simulations in every year of retirement that adjusts for the prior year’s returns), but we can do it! For this blog, I’ll do it 10 times. (Too many more and the charts get a little crazy, plus it takes my computer about 10 minutes just to do the calculations). Here are the resulting probabilities of success over time. 
So far so good, the clients’ probabilities of success are fluctuating between 40% and 60%, as we defined.
What’s happening with spending? Below are the annual spending amounts for each scenario. Anything above the dashed line means we adjusted spending up from the initial plan, and anything below means we adjusted spending down. In most cases, spending was actually greater than what we originally planned for.
Remember, this client wanted to die with nothing. Here are the resulting portfolio values:
Exactly as expected! All portfolios are heading to $0 by plan end. Since even in the final years, we are still trying to maintain a 50% probability of success, we end up with plans that NEVER run out of money before the plan ends, and NEVER have money remaining after.
Of course, we’d want to be VERY sure that a client is going to be gone in 30 years, otherwise this was too aggressive. There are simple remedies. We could add a legacy or “buffer” goal so that there’s money at the end for one.
More realistically, we’d just stop forcing clients to spend more when the probability of success gets above the upper target (60%) at some point. If we did that, you’d see the probability of success gradually creep up as the plan gets closer to the end. This is likely what would happen anyway, as people naturally reduce spending as they age.
A Better Way to Plan
This is dynamic, multi-year financial planning that models not only what kind of returns we could have in the future, but how we might react to those returns as well. As you saw above, a 50% probability of success plan never failed, and in many cases increased spending to an unrealistic level. This isn’t captured in today’s planning software.
What does it mean for you? If you have a plan with a higher probability of success than 50% (which is everyone), you have flexibility in your plan. We should act on that.
If you want to spend more, let’s create an extra “discretionary” spending goal that we can adjust over time, and make it large enough such that it gets us closer to 50% probability of success. We call this flexible spending. This “extra” bucket is adjustable like in the simulations above. At worst, we’d have to reduce it until you are back at the spending level you have today. If you don’t want to spend it, lets give some of it away, or put it in a higher growth portfolio to maximize your legacy.
Reframing risk to properly account for spending adjustments (and all the other conservative assumptions built into plan), along with avoiding withdrawals from your risky investments to mitigate sequence of returns risk will dramatically increase how much you utilize your assets. This is a “better” financial plan.
In light of the improvements these adjustments make to the financial plan, we need to stop thinking of investments as just a tool to get us to the finish line, and as means to specific ends. A client’s investments should be tailored to the individual goals of the financial plan, which we’ll cover in the next part of the series.
At WJ Interests, we help you move beyond traditional retirement planning to maximize your wealth and live out your goals. As a trusted resource in Sugar Land, TX, we’re here to guide you toward a confident, meaningful future and lasting legacy.
Please Click Here for the disclosures regarding the WJ Blog.
