November 12, 2023 – If I wanted to comment on every piece of bad advice in the personal finance community, my quiet, relaxed early retirement would be busier than the corporate career I left in 2018. So, I usually stay out of the daily Twitter/X spats. Last week, though, an incident caught my attention, and it was egregious enough that I weighed in. In a recent Dave Ramsey show (original video here, starting at the 1:13:50 mark, Twitter discussion here), Dave doubled down on his recommendation of the 8% safe withdrawal rate in retirement, calculated as 12% expected equity returns minus 4% inflation (his numbers, not mine – more on that later). And several people pinged me and wanted me to comment. Safe Withdrawal Rates are my wheelhouse, given that I wrote a 60-part series looking at the topic from almost every angle I can think of. So here is my analysis, more detailed than I could do in a tweet: Don’t use a 8% Withdrawal Rate! That recommendation is crazy in more than one way. Let’s see why…
Crazy 1: Historical, real average stock returns are lower than 8%.
Dave is wrong on the S&P 500 average returns. I’ve pointed that out in my 2019 post, “How to Lie with Personal Finance.” The arithmetic average of annual returns will consistently and substantially overestimate the average compounded portfolio returns. For example, if you invest $100 and your returns are -20% in year one and +40% in year two, then Dave Ramsey’s average return would be 10%. But your portfolio will drop to $80 in year one and then jump to $80*1.4=$112 in year 2. So you’re $9 short compared to your compounded $121 portfolio value had you realized 10% average returns. The significant recovery in year two only applied to the smaller portfolio of $80, not the full $100 starting capital. So you’re missing out on $20*40%=8$, plus another $1 from the compounding of the 10% over two years.
It is a general insight from basic math and statistics that the higher the volatility of returns, the higher the drag from this effect. With a standard deviation of S&P 500 annual returns of between 16% and 20%, we’d expect the portfolio’s true compounded annual growth rate (CAGR) to be between 1.25 and 2.00 percentage points below the arithmetic mean. Please check out the money chimp calculator to compare the naive arithmetic average and true CAGR. For example, between 1871 and 2022, the arithmetic average was 10.81%, but the CAGR was only 9.16%, i.e., 1.65 percentage points lower.
Adjusting for CPI inflation, I calculate an average return of 6.72% in the S&P 500 for 1/1871 to 9/2023, with dividends reinvested. Significantly below Dave’s 8% number. People have repeatedly raised this issue with Dave, and he usually responds with ad hominem attacks like, “You’re a dumb math geek who knows nothing about money and you live in your mom’s basement.” OK, who will tell Dave that the Stacking Benjamins show’s “Mom’s basement” shtick is not real? It’s just for show, man!
Crazy 2: Dave Ramsey wants you to hold 100% equities in retirement.
Another piece of crazy advice is that Dave suggests you keep your retirement portfolio in 100% equities. It’s certainly true that a 100% equity portfolio gives you a high average return. 100% equities may work well while accumulating assets; see my post on optimizing pre-retirement glidepaths. However, a 100% equity weight creates an unpleasantly high portfolio volatility, which is particularly dangerous in retirement due to “Sequence of Return Risk.” Therefore, most financial advisers recommend between 20% and 40% of diversifying assets (e.g., intermediate Treasury bonds) during retirement. For example, in the chart below, I plot the withdrawal rates over 30 years (assuming capital depletion) that would have generated specific failure probabilities. 0% is the failsafe, i.e., the highest withdrawal rate that would have succeeded in sustaining a 30-year retirement. The values at 1% would generate a 1% failure probability, etc., all the way to the median, where you would have made it with a 50% probability.
With the 75% equity portfolio, you can raise the failsafe, even though the median outcomes in historical simulations will undoubtedly suffer a bit from the smaller expected return. But also note that too meek of a portfolio (e.g., 50% stocks and 50% bonds) will do consistently worse than the 75/25 allocation. So, from a safe withdrawal planning perspective, there is a sweet spot in the stock vs. bond diversification spectrum. Keep enough bonds to take the edge off the highly volatile equity portfolio, but keep the bond allocation low enough not to torpedo your long-term average real returns.
Crazy 3: An 8% fixed withdrawal rate would likely not survive a 30-year retirement.
Let’s put the Dave Ramsey claim to the test. I use my Google Sheet (see Part 28 of my Safe Withdrawal Series for instructions) and assume:
- 100% equities.
- A 0.05% annualized expense ratio.
- A 30-year horizon.
- Capital depletion, i.e., the final portfolio target is $0 or more. Later, I will also use the even tighter constraint of capital preservation, i.e., you succeed only if the CPI-adjusted portfolio value stays above the initial level.
- No supplemental flows.
- Withdrawals rise in line with CPI inflation.
I display the historical simulation results in the table below. This is the standard output in my Google Sheet. Each row is for a specific withdrawal rate: 2.75% to 4.25% in 0.25% steps. I also added 6.5% (roughly the average real equity return) and Dave Ramsey’s 8% rate. Each column is for a different simulation history subset. We can look at all historical cohorts since 1871—or only cohorts since 1926 (the starting point of the Trinity Study). And then also slice by equity valuations: a) the Shiller CAPE below 20 (i.e., equities are historically relatively cheap), b) the CAPE above 20 (equities are relatively expensive), and c) the CAPE above 20 and the S&P 500 index at its all-time high.
Let’s look at the failure probabilities: With the 8% withdrawal rate, you have an overall failure rate of about 56-61%, depending on the simulation start date. Even when the CAPE is below 20, you still have a less than 50% success rate. Conditional on the CAPE being above 20, you have only a pathetically low 3% success rate.
What’s the verdict here? Given today’s elevated CAPE ratio, Dave’s advice will almost certainly wipe out your portfolio. As of Friday, November 10, 2023, the CAPE was 29.48 (traditional Shiller CAPE) or the more historically comparable 24.34 (see my post, Building a Better CAPE Ratio for more details on the adjustments I implemented). You can all but guarantee to run out of money using Dave’s recommendation and assuming future equity returns follow a similar pattern as during the past 150 years.
Also, notice that the unacceptable failure probabilities are not just due to Dave Ramsey’s slight miscalculation of average returns. Even using a 6.5% withdrawal rate, just under the long-term average real equity return, you’d run out of money about 37-40% of the time unconditionally. And 77-78% conditional on the CAPE above 20. And even with 4.00% to 4.25% withdrawal rates, you’d still generate mostly unacceptable failure rates in the double-digit percent range once you account for expensive equity valuations. Sequence Risk is a real headache!
Maybe the 8% Rule can still work because all the historical failures are simply due to running out of money after 27 or 28 years. So you could still be safe for most of your 30-year retirement. But that’s not the case. Here are the five stereotypical sequence risk victims starting retirement right at the market peak before big bear markets.
- 8/1929, right before the Great Depression.
- 11/1968, before the start of stagnant equity markets in the late 60s.
- 12/1972, at the market peak before the first oil shock.
- 8/2000, at the Dot-Com market peak.
- 9/2007, at the peak before the Global Financial Crisis.
If we track a hypothetical portfolio value time series, $1,000,000 at the beginning of retirement, then $80,000 in annual budget (modeled as $6,666,67 in monthly withdrawals), you would wipe out your portfolio well before the end of your 30-year retirement. After less than six years during the Great Depression, after around nine years in the 1968, 1972, and 2000 cohorts and about 14 years for the 2007 cohort.
Crazy 4: Preserving your capital is even less likely
If you thought Dave’s claims couldn’t get any crazier, note that he stresses in the video that your portfolio will not just survive with an 8% withdrawal rate, but you will even preserve your capital. So, let’s put that claim to the test and calculate how likely it is to maintain the portfolio value plus CPI after 30 years. Please see the table below:
The unconditional failure probabilities now rise to about 76% (all cohorts) or 71% (cohorts since 1926). Conditional on an elevated CAPE ratio, you have a big, fat 100% failure probability. So, historically, in retirement cohorts with similarly valued S&P 500 CAPE ratios, there hasn’t been a single cohort that was able to preserve the portfolio’s purchasing power after withdrawing 8% in the first year and adjusting withdrawals subsequent withdrawals for CPI inflation.
Crazy 5: An 8% variable withdrawal rate is not very useful either!
Maybe we misunderstood Dave Ramsey. Perhaps a way to salvage that unmitigated disaster that his advice would create in your retirement is to assume that we withdraw a variable 8%. So, imagine you start with a $1,000,000 portfolio and withdraw $80,000 in year 1. Imagine further that the portfolio falls to $800,000 at the beginning of year 2. Now, you only withdraw $64,000, equal to 8% of the new portfolio value, instead of the $80k plus CPI. The advantage is that we’ll never deplete the portfolio down to zero. But we could certainly suffer painful cuts in our portfolio value and thus the purchasing power of our retirement budget.
Let’s look at the unconditional distribution of withdrawals of that $1000000 portfolio using the 8% variable withdrawal rate. We start at $80k p.a., and then, due to the volatility of the portfolio returns, the realized historical portfolio values and withdrawal amounts fan out, as in the chart below. Noteworthy, some cohorts even raise their withdrawals due to outstanding returns. The 90th percentile stays above the $80k mark for 30 years. But the median slowly drops to below $50k, and the 25th and especially 10th percentile sustain substantial budget cuts, down to about $30k and $25k per year toward the end of retirement.
This strategy is unworkable for most retirees without the flexibility to cut their retirement budget in half. Also, for longer horizons, you will only exacerbate the spending cuts. So, FIRE enthusiasts with a 50-year or even 60-year horizon must keep cutting their retirement budget even more.
But it gets even worse. If we now focus on the historical cohorts with similarly richly-priced equity valuations, we get the spending distribution chart below. The median retirement budget is now at only $18k, almost 80% below the initial. Even the 90th percentile is down to under $25k p.a. The 10th percentile has wiped out over 90% of its portfolio and retirement budget, with a final value of only about $6,400 p.a.
Summary so far
You’re asking for trouble if you use an 8% initial withdrawal rate when equities are moderately overvalued (CAPE>20, as they are today). Historically, you would have run out of money if you withdrew a fixed amount. Or you would have melted down the portfolio after 30 years to a degree that you’d have to live on severely constrained terms. I would not recommend this retirement strategy to anyone I know and care about!
Attempting a diagnosis
Don’t get me wrong. Dave Ramsey provides some good content. Folks in debt have benefited from his tough talk and gotten their finances and often their lives back in order again—credit where credit is due. I also don’t think Ramsey is pushing his 8% Rule advice out of malice. If I had to diagnose the problem, Ramsey is the poster child of the Dunning-Kruger Effect. It’s a cognitive bias established in numerous empirical studies whereby people are often overly confident about their abilities, especially when they have only very shallow subject knowledge. At that point, you’re on top of “Mount Stupid,” see the diagram below. With more experience, you will recognize your shortcomings and blind spots and tread more cautiously. Only later you’d gain more confidence and actual expert status. Also noteworthy, even experts remain relatively humble, and their confidence remains below the Mount Stupid level.
The Dunning-Kruger problem often surfaces when people pick up a few fun and intriguing insights here and there and believe they’re now subject experts. For example, Dave Ramsey knows a thing or two about finance, like average equity returns and inflation rates. He likely has a lot of deep subject knowledge in entirely unrelated areas of personal finance. But he has never run a single safe withdrawal rate simulation in his life. He is not familiar with Sequence Risk. He doesn’t know what he doesn’t know. And what’s worse, he doesn’t even want to learn more and brushes aside every critic who points out his fallacies. In the video, his daughter Rachel points out that the 4-5% Rule isn’t “that stupid” (around the 1:17:27 mark), but Dave just completely bulldozes over her, and she backs off and goes with the program afterward. So, the tragedy here is that Dave Ramsey seemingly wants to stay at that “Mount Stupid” level. And it’s troubling that he calls other people stupid for noticing!
Most people going through this Dunning-Kruger cycle will eventually acquire more experience and expert knowledge. I indeed went through this cycle during the last ten years planning for early retirement, starting with enthusiasm and overconfidence. Then doubts crept up once I looked into the rabbit hole of safe withdrawal rates, noticing that the Bengen and Trinity Study work shouldn’t be applied to all FIRE folks. Bengen and Trinity only talk about unconditional failure/success probabilities, completely ignoring equity valuations. But with more research, I gained enough confidence to retire comfortably in 2018. I don’t argue that I know everything. Quite the opposite; I might be a moron, too, and I got a lot more things I want to understand. I just try to be a smaller moron than Dave Ramsey.
Dunning-Kruger is everywhere!
And, of course, my blog post here wouldn’t be complete without offending some of my fellow FIRE bloggers. Before we unload too much on Dave Ramsey, remember that the Dunning-Kruger Effect is also alive and well here in the FIRE community. For example, the following (paraphrased) claims are circulating in our community:
- You can take your withdrawal rate to 7% if you’re flexible.
- If your portfolio survives a 30-year retirement, it also survives the next 30 years. We can use the Trinity Study results and use them for early retirement.
- The 4% Rule can’t fail if we raise the weighted dividend yield to 4%.
- … and many more.
Like Dave Ramsey’s crazy rant, all these claims are the product of the Dunning-Kruger effect. They are based on folks casually reading a Jack Bogle book and/or the Trinity Study and making up some additional sh!t on their own, without ever bothering to answer basic questions like, “How deep of a budget cut would that flexibility entail?” or, “how long do you need to stay flexible?” etc. Or “Why would higher dividend yields automatically boost returns?” The appeal of writing what people want to hear, confirmation bias, and the tendency of blog posts to go viral when using overly optimistic estimates for withdrawal rates are all around us.
To wrap up, I liked my first-grade teacher back in Germany. She was a significant influence at that time in my life. I never felt the urge to reach out to her while working on my doctoral dissertation in economics or preparing for early retirement, though. If she had reached out to me and offered advice and told me that I shouldn’t listen to the math geeks around me, I would have politely told her to stay in her lane. Similarly, most of us in the FIRE community have graduated from Dave Ramsey. Or even better, many of us, myself included, never even required his services. We should all safely ignore his 8% withdrawal rate advice now. But I feel sorry for the Ramsey listeners. I hope they are smart enough to get a second opinion elsewhere before implementing his crazy, unhinged advice. But, for the love of God, please stay away from Suze Orman!