A Critical Look at Dave Ramsey's Retirement Strategy
This blog was written by our summer intern, Aaron Price. Aaron is currently a double major in finance and music at the University of Houston. If you have any comments, you can email him at aaron@kingspath.com.
Dave Ramsey, an internet financial planning superstar, is a name you are likely familiar with – and for good reason! He is worth hundreds of millions of dollars, helps the average person get out of debt, and provides unique financial insight. Recently though, Ramsey has sparked debate within the finance community among financial planners due to one of his latest recommendations to a caller around age 60: an 8% per year spending policy (against the industry standard of 4% per year) on a 100% equity portfolio.
In his thought experiment, if you invest in “good mutual funds”, and you get market average returns of 10%, even with inflation of 2-3%, an 8%/year spending policy is fine. Of course, 10% is a greater number than 8%, so it all makes sense. Right? If not, what should my spending policy be?
Let’s first dismantle the assumptions he makes: 10% market returns (1), mutual fund investment vehicles (2), a 100% equity allocated portfolio (3), an assumption of an arithmetic mean instead of a geometric mean (4) and an assumption of no volatility (5) :
The number 10%, even using an arithmetic mean (more on the correct average-calculating method later), is slightly ambitious. The arithmetic mean, not adjusting for inflation, using data from the last 80 years is close to 9.31%. 69 basis points, or BPs, makes a huge difference due to compounding interests and costs.
His return assumption assumes significant alpha specifically through mutual funds (alpha just means outperformance compared to a benchmark like the S&P 500 index). However, mutual fund performance and persistence has been and still is horrendously low. In other words, alpha from mutual funds is not only unlikely, but also laughably inconsistent.
Most investors should never have a 100% equity allocated portfolio at retirement, mostly due to the consequences of the 4th and 5th assumptions
Assuming a 100% equity allocation, an arithmetic mean approach to stock market data is unrealistic considering the compounding losses associated with a loss followed by a gain. A geometric mean accounts for the losses that are incurred whenever the market goes down then goes back up. In other words, a negative 10% and a positive 20% return does not reflect a 5% return, but rather a 3.92% return. Assuming a geometric mean, the nominal return of the S&P 500 is close to 7.41% rather than the earlier mentioned 9.31%.
Volatility is guaranteed in real markets, yet there is lack of acknowledgement of this in Dave’s thought experiment. In other words, there is no consideration of sequence of return risk.
The crux of the issue here revolves around the fifth assumption: volatility. To understand why, we must first understand an "average market return," which is the average of all the returns in an index such as the S&P 500. Of course, you will not get 10% every year, but rather your actual returns will seem much more dramatic. Sometimes, the returns can be as drastic as negative 30% or even positive 40%. These extreme returns are classified as “tail risks”, and they greatly influence the average return.
With tail risks in mind, let’s talk about spending policies.
The following graphs assume the following conditions:
100% equity allocation
Normally distributed volatility derived from the standard deviation of historical S&P 500 inflation-adjusted returns
CPI-adjusted annual expenses
Success is defined as having over a $0.00 balance at the end of 35 years. (Age 65-100)
10,000 trials were run for each scenario
This first graph highlights the experience of a retirement account at an 8% withdrawal rate and features a horrific 15.9% success rate. Notice the average case – or the “mean” case – with this scenario: you are expected to run out of money 16 years into your retirement. This is ridiculously unacceptable. For some more context, we were able to derive the conclusions of each of these graphs through Monte Carlo simulations – a statistics practice that allows an output of a percent chance of success considering thousands and thousands of possible combinations and trials. The “confidence” intervals you see on the graph, or the shaded regions, represent the likelihood of the account balance staying in that range. When confidence increases, the shaded regions increase, and vise versa. The darker the shade, the more likely the event is, with the mean event being the most likely.
This second graph exhibits the same conditions as the last chart, except the withdrawal rate changes to the industry standard of 4%, rather than 8%. The success rate of the account balance remaining above $0.00 at age 100 is 63.5% That is a significant leap from 15.9%! You are almost four times more likely to be okay. In addition, notice the mean scenario: more often than not, at age 100, your account balance remains steady at around $1,000,000.
This last graph represents a more realistic example in terms of planning- a 50% equity and 50% bond portfolio rebalanced annually. With an 80.7% success rate, this chart considers the fact that a 100% equity-allocated portfolio at retirement is unrealistic. The choice of a 50/50 allocation, or 50% stocks and 50% bonds, is arbitrary, and the actual allocation would depend on an investor's risk tolerance. Due to the decrease in risk and diversification of another asset class, you see the mean scenario balance increase over time, whereas in the other cases, it was either decreasing or stagnant.
You might notice an interesting effect with these charts: the mean is always closer to the 5% percentile than the 95% percentile. People might expect the average to be directly in the middle of the two, but it’s not! This effect is partially due to the sequence of return risk: a concept found in investing in which negative returns early on significantly impact the future of an account, and when you experience positive returns early on, the account will become significantly more resistant to the withdrawals throughout the lifetime of the account. The reason the 95% percentile line seems to skyrocket in all the scenarios is explained through this principle.
Another concern you might have picked up on is the fact that the best chart we have shown so far has only an 80.7% chance of success! 80.7% is not amazing, but it is certainly not as horrific as a 15.9% success rate derived from an 8% spending policy. Bringing that number up to 100% chance of success is quite a challenge, but fully attainable. The percent chance of success in these scenarios would be significantly higher if we took into account different strategies like variable withdrawals, decreasing the time horizon from 35 years to 30 years, choosing a more efficient allocation, etc. Moral of the story is that there are several ways to improve your likelihood of having a successful retirement and can be discussed thoroughly with an advisor.
Finally, let’s go back to the tail risks we discussed earlier. The reason these success rates are not 100% is purely due to the timing of tail risks and sequence of return risk. If the market goes down 30% in the first year of your retirement, your chance of success literally drops to 39.1% from 80.7% if you’re withdrawing 4% of the initial $1,000,000 50/50 allocated account.
A 41.6% decrease in the success chance, because of ONE bad year. Tail risks are brutal.
Notice with all these charts it is not a matter of whether the tail risks occur or not, it is a matter of when, and whether or not the balance is large enough to withstand a negative tail risk. The tail risks are represented by how “wide” the shaded area is, with the 5% percentile representing negative tail risks earlier, and the 95% percentile representing positive tail risks earlier.
A 4% spending policy hedges against this risk, as it allows the balance to accumulate in size faster, while also reducing expenses. This is why even though you will see tail risks occurring in 4% drawdown scenarios, the account remains supported.
Because of these facts, we highly caution against an 8% drawdown and strongly recommend something closer to 4%, and when it becomes time for you to consider retirement, engaging with a financial advisor to develop a robust financial plan that considers sequence of returns, variable spending strategies, investment allocation and more.
