In some previous articles, I’ve discussed longevity risk, withdrawal rates, and Dave Ramsey’s 8% withdrawal rate recommendation. These topics can quickly get mind-numbingly complex.
The financial side of retirement has a lot to do with the math; it’s a big math problem we need to solve as best we can. We do this thoughtfully, humbly, and wisely, recognizing our finite abilities and the lack of absolute certainty in our calculations.
Most importantly, we realize that our future provision is ultimately in God’s hands:
”But I trust in you, O Lord; I say, ‘You are my God.’ My times are in your hand;” (Ps. 21:14-15a, ESV)
Retirement planning, in general, is a discipline filled with financial jargon as well as complex mathematical calculations. But as I have stated many times on this blog, my goal is to simplify these concepts so that you can understand them and make wise decisions about your retirement stewardship.
Because these topics are crucial to your long-term retirement stewardship, I want to continue the discussion by exploring the “probability of portfolio ruin” or p(ruin) concept.
No, p(ruin) isn’t some fancy casino game or a dramatic end-of-the-world movie plot; it’s a part math, part economics, and part emotions problem that’s a reasonable concern of retirees. It’s the opposite of the “probability of portfolio success,” which certainly sounds better.
This topic comes up quite often in the retirement planning literature, and though it sounds like a simple term (and a little scary), it’s not as straightforward as it may seem. Moreover, some think it’s of vital importance, while others believe it isn’t ultimately.
What’s the “probability of ruin” (and can we please call it something else?)
Picture this: you’ve saved, invested, and planned wisely for decades, and now it’s retirement o’clock. You’re all set to buy that pop-up camper for some fun in the outdoors (too bad there’s no such thing as a “pop-up Holiday Inn Express room”—I might buy myself one) and take up soap carving (because why not? you need something to do in retirement and it is a ‘thing’).
But wait! There’s a potential risk looming on the horizon: the “probability of ruin.”
“Oh, thanks for raining on my retirement parade (and my soap is melting),” you say, “and what exactly is ‘probability of ruin?’ ” you ask. Well, it’s the chance (expressed as a percentage) that your retirement savings might not last as long as you do. Your money might decide to bail on you when you need to ensure you have a lifetime supply of soap for your carving hobby.
Soap carving aside (you’re welcome), the probability of ruin, or p(ruin), is closely related to the ”safe withdrawal rates” (SWR) and “sequence of return rates” (SoR), concepts we’ve discussed earlier. Your chosen SWR affects your p(ruin) based on the relationships between the amount of your savings, your withdrawal rate, and how long you need them to last.
Sequence risk affects p(ruin) a little differently. It doesn’t mean you can’t have any negative returns, or really bad negative returns in a given year, for that matter. Having a long sequence of years of negative returns has the most impact. Though such periods are relatively rare, according to Morningstar, there have been quite a few. Depending on your SWR, a long period of primarily positive by low real (after inflation) returns can have a similar effect.
Perhaps you’re familiar with the so-called “Lost Decade” of the 1970s? It wasn’t a science fiction TV series; it was a period of higher inflation combined with a stock market that cratered and then plateaued for a while. But then the Fed clamped down on inflation and later dropped interest rates, and stocks launched into a multi-decade bull market. Timing is everything.
Simply stated, p(ruin) tries to answer this question: “What’s the likelihood (expressed as a probability percentage) that your money won’t last as long as you do?” If, based on whatever calculation method you decide to use, the probability is low, then you can rejoice. You may need to return some of that soap if it’s high.
Despite its shortcomings, p(ruin) has become the de facto metric for retirement planning, mainly due to the now famous “4% Rule,” which was never intended to be a rule but, as Captain Barbossa said, speaking of the Pirates Code in “The Pirates of the Caribbean” movie after reneging on an agreement said, ”the code is more what you’d call ‘guidelines’ than actual rules.”
This “guideline” suggests that retirees with a balanced portfolio (40/60, 50/50, 60/40 stocks/bonds) should be able to withdraw about 4% of their savings each year, adjusted for inflation, to have a 95% chance of not running out of money during retirement. You may even be using it or some variation of it.
This probability, expressed as a percentage (e.g., with a 95% success rate, says you have a 5% p(ruin)—p(success) would admittedly be a nicer way to express it—is often used to determine whether you are using a “safe” withdrawal rate.
It’s also used to set a savings target; i.e., with a 4% withdrawal rate, you’ll need 25 times your required income. For example, if you need $20K per year, you need to save $500 (25 x $20K = $500K, and $20K / .04 = $500K, and $500K x .04 = $20K).
It’s proven to be a helpful guide, and I’m sure Captain Barbossa would agree. However, the 4% “rule” and this idea of p(ruin) have faced criticism and debate, especially in recent years. Year-over-year studies by Morningstar has come up with a SWR of 3.3%, 3.8%, and most recently, back to 4.0% in 2021, 2021, and 2023, respectively.
It seems we’ve, in effect, come full circle since William Bengen’s original SWR study in 1994.
Estimating “p(ruin)”
The work in 1994 by William Bengen that came up with the “4% rule” was based on actual historical market returns. (Did you know we have stock market records from the beginning of the Dow Jones Industrial Average in 1896 and the S&P 500 since 1923?)
He modeled the performance of different retirement portfolio stock/bond allocations for 51 years between 1926 and 1976 for minimum 30-year periods. Bengen found that 4% was the withdrawal rate when no failures occurred when simulating possible outcomes for minimum-30-year periods. Thus, 4% was the maximum percentage that was deemed “safe.” A subsequent study in 1998, now known as the Trinity Study, backed up Bengen’s initial findings.
The use of historical market returns makes a big assumption: that the capital markets will behave in the future like they did in the past. It assumes that we have a relatively limited number of 30-year historical periods of data, which is all we need to know to project future market returns with a reasonable degree of certainty.
Many investors assume that the future will look much like the past; they’re “positively biased,” This seems logical since the markets have generally done well over long periods, but if you think about it, it may not be wise to take the past and project it well into the future. As Yogi Berra once said, “Making predictions is hard, especially about the future.” Yogi’s right; the future will look like the future, which may or may not be similar to the past.
When it comes to global economics and personal finance, it’s more likely that we’ll experience something we haven’t seen before in our lifetimes (like a worldwide pandemic) than not. In my investing lifetime, there have been a couple of “black swan” events and at least as many “grey swan” ones.
In other words, 50 or 100 years of historical data can give us trends but have little predictive power. It can only tell us what the future might look like if the same trends occur as they did in our limited past.
Because many think they won’t, more recently, retirement income researchers have used Monte Carlo simulations with the assumption of lower expected returns than we have seen historically. (And no, Monte Carlo has nothing to do with casinos, gambling, or a car; well, not too much.) The honestmath.com tool I used in the article about Dave Ramsey and his 8% withdrawal rate uses Monte Carlo simulations.
Monte Carlo analysis is not based on historical returns but on assumptions about future asset returns, which can only be proven correct in hindsight. This uncertainty about future returns (and their sequence) is the essence of risk in retirement models.
The same can be said of other assumptions we must make to run such simulations, such as inflation and taxes.
These simulations can also model extreme market events (called “long tails” by statisticians—an odd term, for sure) and “black swan” events such as the 2020 pandemic. Studies like the ones done by Morningstar using Monte Carlo simulations have resulted in revised SWRs of between 3.0% and 3.8%, somewhat lower than Bengen’s original 4%. However, more recently, due to high yields on fixed-income securities, it has increased to 4%.
It’s important to remember that SWRs are based on a 90% portfolio survival rate over 30 years. That translates into a 10% p(ruin).
Can’t we do better?
The question many retirees want answered precisely is almost impossible: “What can I spend each year for the next 30 years and be 95 to 100% certain that I won’t outlive my savings”? (This implies a 5% or less p(ruin)).
It’s impossible to answer with 100% certainty because we can’t predict future market returns or how long you and your spouse will live. Nor do we know exactly what we will have to spend. And we surely can’t know what “black swan” events may be ahead—that’s why they call them black swans (it turns out they’re scarce).
As stated previously, current estimates from various models and assumptions range from about 3.5% to about 4.0% of your portfolio’s starting value based on a life expectancy of about 30 years. That percentage is increased by the rate of inflation each year, so it’s hard to predict what it will be in 30 years. It could be 10 percent of your remaining balance or higher.
If this sounds like rolling the dice with your retirement savings, that’s because it is, to a point. From the world’s perspective, it all seems random and based on chance. But we know God is sovereign and reigns over all, including the financial markets (Lam. 3:37–38, Isa. 45:7–9). How exactly he does that and what he plans to do in the future aren’t things he’s chosen to reveal to us.
Still, you may wonder why there isn’t one fool-proof sustainable withdrawal rate that everyone can use with a 90% or higher confidence factor. The problem is our inability to predict the future of market returns; it’s impossible to prove that any individual model is correct and will hold fast over 30 years.
The idea of p(ruin) comes from the uncertainty of investment outcomes. That’s why newer models are run using state-of-the-art computers to analyze tens of thousands of potential outcomes. (HonestMath.com runs their Monte Carlo simulations on Amazon Web Services, AWS.)
But despite that capability, different models and assumptions produce disparate estimates of sustainable rates.
This is proof of what we already knew: using sustainable withdrawal rates is risky; how risky it is depends on how much you diverge from the strategy.
What to do?
Based on the best available information, I recommend following a withdrawal strategy with a percentage starting at the low end of the range if possible because that would be the most conservative. If you do, that will also significantly reduce the amount you can spend, or it may not be enough to pay the bills for some people.
As Christians, we bring some perspectives to this problem that can help us deal with this future uncertainty about market performance, our spending needs over time, and other factors outside our control. These include humility, wisdom, self-control, and trust in God.
Humility tells us while we can plan, we can’t control the outcomes—only God knows and controls all that will happen (James 4:13-16). So, we know that any financial concept that uses “probabilities” based on trends, modeling, and simulations is limited at best and poor at worst in predicting future outcomes.
Wisdom will help us realize that while these things may help frame the issue and provide some guidance early on in selecting a strategy, they don’t offer absolute certainty (Prov. 18:13). A wise steward would monitor how well their system is working (perhaps with the help of a trusted professional advisor) and then adjust as the situation warrants.
Self-control would cause a retiree who sees that it’s increasingly likely they’re going to run out of money to make those adjustments, even if they are a little “painful,” because they will almost always mean reducing spending, which may result in a lower standard of living (depending on lifestyle) (Titus 2:12)
Trust in God’s love and faithfulness to us, which includes providing for our day-to-day needs (which has a 100% probability), even with the possibility of a reduced standard of living or the sudden need for it, gives us great hope and assurance (Matt. 6:31-34).
Of course, the possibility of a reduced standard of living later in retirement, and to what degree, is difficult to estimate. We might imagine the following (worst-case) scenario: “What would my lifestyle be if I exhausted all my retirement savings?”
One way to look at this is to confess that no matter what happens, we have Christ (Rom. 8:38-39). All we need can be found in him, the All-Sufficient One, through whom the grace of God has been given us in abundance (2 Cor. 9:8).
Considering our financial situation, my wife and I would be left with only our Social Security benefits. We could cover our bare necessities with Social Security only, but that would be about it. This scenario is what many retirement planning professionals want us to envision to encourage us to put in place a “backup plan” to mitigate, perhaps one involving income annuities.
If someone were to ask you what your ”acceptable” p(ruin) is, you would probably say ”zero” or something closer to it than 5% or 10%. So would I.
If I knew with absolute certainty that our p(ruin) was 10%, I might have concerns, but it would mean that the things I listed above—humility, wisdom, self-control, and trust—will be more necessary than ever.
The problem is that we don’t know how to process even a small probability of a significant life-impacting event, so I think I’m going to ignore trying to determine the exact probability altogether, do what I can to avoid it and trust God to provide for my wife and me for as long as we live whether our chance of ”ruin” is 0% or 100%.