$1 Million Isn’t Enough
I came across a post from yesterday from Mr Money Mustache, one of the oldest sage fountains of wisdom at the spiritual center of the FIRE movement. In it, MMM argued that an individual could comfortably retire on $1 million; in fact it seems as though he is fairly incredulous that anybody could question these assumptions.
I’m formally declaring my break with the Mustache-ian school of logic as I just can’t follow him where he is wanting to go with this post. I think his advice is extremely dangerous and likely to end extremely poorly for anyone that follows it. The reasons are that it completely ignores sequence of returns risk and current market valuations.
I fear MMM is selling water to the thirsty in the desert, but 10 years from now they will discover they drank a whole bunch of sand.
As with everything on this site, everything that follows is for informational purposes only and is not a recommendation to buy or sell securities.
The Mustache Assumptions
I’m summarizing the article briefly, but the main points are that:
- An individual with $1 million can comfortably spend $40,000 a year and is virtually guaranteed to never run out of funds
- The split between retirement accounts and after-tax accounts doesn’t matter
- Only 4% real returns (inclusive of dividends) are required to sustain this.
- The entire portfolio is in stocks (which makes Monte Carlo testing rather easy for me)
4% real returns are actually slightly lower than the real, inflation adjusted historical return for the S&P 500 (about 6% since 1962). However, for purposes of my own testing, I used the 4% numbers to match up with MMM.
Where Is The Volatility?
A brutally nasty consequence of retirement planning is sequence of returns risk. This is simply the risk that a major market downturn hits during the earlier years of retirement. For example, if a $1 million portfolio is cut in half overnight by a stock downturn, $40k of spending is now eating up nearly 10% of the yearly principal. Investment proceeds in the future cannot possibly counteract the much higher percentage drawdown, and the portfolio crashes to zero.
It is very easy to come up with point estimates following a simple 4% trajectory that magically work, but the volatility of the market doesn’t allow us that luxury in real life. As I read the article, the experience of the late 90s kept sounding in my mind, so I decided to put MMM’s assumptions to the test.
I obtained total return data for the S&P 500 back to 1901 courtesy of Robert Shiller’s web page. I then built a simple model to do the following:
- From 1987 to 2007, test the hypothesis that you could retire with $1 million of real wealth
- Since average US life expectancy is about 80 and MMM cites retiring at 35, this is left with 45 years to cover
- Each year, the algorithm calculates a 45 year portfolio simulation:
- Actual S&P 500 returns for 1987-2018
- For 2019+, MMM’s assumed 4% real return.
- Simulate 10,000 outcomes with 13.9% volatility (the historic volatility on the S&P 500)
- Note: My own beliefs as presented here are that long term returns are entirely predicted by starting valuation, but for now I will use the ‘conventional’ assumptions that stock returns are upward trending random walks.
- Calculate the percentage of portfolios in the 10,000 sims that ‘Crash Into the Mountain’ and end up with no money before the end of life
Probability of Being Broke
The columns in the chart above are the probability that the money runs out before end of life based on retirement date. If you took the MMM advice and retired with a $1 million portfolio in 1999 or 2000, there is a near-certainty the money is gone. The most likely case is that your money would last until 2030, at which point there is nothing left. I don’t really want to be a 65 year old with not a penny to my name:
The black line above represents a measure of market valuations proposed by Warren Buffett, defined as the market capitalization of stocks divided by the GDP of the country. When it is high, stocks are comparatively more expensive.
The fascinating thing from this analysis is that the results are highly correlated: the higher the beginning valuation, the higher the probability that the portfolio crashes. In fact, mathematically this correlation is amazingly strong, clocking in at 0.86.
The reason is the because long term valuations matter to a portfolio. As pointed out by Robert Shiller for his Nobel Prize, the relationship between stock returns and starting valuation is very, very clear in the long-run:
Just spitballing at a chart here, but the last time the Buffett Indicator currently indicates stocks are approximately at about the same levels as 1998 type prices. There is a 55% chance someone retiring in 1998 will have run out of money before end of life. That’s better than a pure coin flip, and surely a hell of a lot more risk than I want.
If you are retiring on $1 million right now, you are betting that Mr. Shiller’s chart is ‘just noise.’ Do you feel that certain?
I’m very concerned a lot of people in the FIRE movement are making some serious, serious mistakes in believing that the recent bull market is the rule and not the exception. As I stated in why are people retiring early, this is the exception, not the rule:
It is therefore certainly not guaranteed that $1 million will last. In fact the data over the last 20 years indicate the odds of running out are vastly, vastly higher than indicated by MMM.
If you are at the precipice preparing your leap at age 35 with a cool mill-sky in the bank, your choices as I see them are:
- Hope Robert Shiller and I are wrong and that there is no relationship between returns and current valuations
- Save more than $1 million to avoid sequence of returns swings
- Wait until valuations become more favorable (like 2003 or 2009). This is the thought process behind business cycle investing.
- Try to lower expenses if the market craters. This allows you to manage the sequence of returns problems but living off of $20k probably isn’t what people have in mind by ‘retiring’.
I’d give a serious, serious long hard think about this before pulling the trigger.
For more discussion of the sequence of returns risk and to see mitigations, please continue reading here.