$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:
  • 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

retiring with 1 million

A Very Dangerous Plan

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:

Your Portfolio Dies Long Before You If You Retired in 2000

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?

Conclusions

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:

  1. Hope Robert Shiller and I are wrong and that there is no relationship between returns and current valuations
  2. Save more than $1 million to avoid sequence of returns swings
  3. Wait until valuations become more favorable (like 2003 or 2009). This is the thought process behind business cycle investing.
  4. 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

 

21 Responses

  1. Mr. CC says:

    Great insight, and I appreciate that writers are highlighting the risk of the 4% Rule for the long-horizon retiree. I think MMM inspired quite a few to retire with 25x expenses in previous years, but it seems more and more are coming around to the 4% “rule of thumb”. Is anyone really retiring and not securing even a modest income anymore? 5-10k in income is a major hedge against sequence of returns risk.

    • FatTailed says:

      Agreed absolutely, it helps dramatically.

      If this blog accomplishes anything, I hope people will look into Dr. Shiller’s work about the valuation/return longterm tradeoff and reconcile how their investing goals align with that.

  2. This is a good counter-argument to what most people are taking as infallible (the Trinity Study), and quite relevant too. I hope that people aren’t retiring today and assuming 4% will definitely work out for them indefinitely, considering sequence of returns risk and the current point in the economic cycle.

    I think the rebuttal you will hear from the FI community goes like this: “When we ‘retire’, we will still earn some income from other sources, such as passive investments and side hustles. And if worst comes to worst, we can go back to work or pick up some part-time employment opportunities.”

    • FatTailed says:

      Certainly that is a fair counter and one way to manage the sequence of returns risk.

      I guess the point of my post is to try to correct a misunderstanding of how returns work. I will probably do this more explicitly in a post, possibly for Friday.

      I always see people using the long-run average equity return in their projections. Or they say “I’m investing for the long-term, I don’t need to sweat how I’m invested today.”

      That logic is fine if you are starting with nothing. However, if you have a huge pile of assets, the only thing that matters is your return forward from TODAY. That is going to be hugely different, even over a 20 year time period, from the average return based on where valuations are. It’s why 2000-2018 annualized returns are absolute dogshit compared to 2002-2018 annualized returns.

      My view (as advocated throughout the blog) is that 1) valuations matter in the long term and 2) we have a pretty good idea of where we are at in the business cycle. I think its needlessly fatalistic to just the hands up and say there is no need to manage risk because no one knows what will happen.

  3. Miguel says:

    This is amazing analysis, especially given the quick turn around.

    I always figured MMM was just overly optimistic to get people hooked. It can be dangerous if people follow it to the letter, but hopefully by the time someone reaches that point at 35, he’ll have the years and knowledge behind him for that level of careful consideration.

    Ultimately though, the purpose might be to get people interested and involved more than give them a roadmap. Just like seeing an Olympian might make you join a gym, even if following their training regimen would be ludicrous.

  4. Seth K says:

    Good read.
    Big ERN has the most robust analysis on this IMO
    50 year retirement with 4% WR and 100% stocks has a 90% success rate historically
    https://earlyretirementnow.com/2016/12/07/the-ultimate-guide-to-safe-withdrawal-rates-part-1-intro/

    He also has analysis related to started at a high CAPE. Having said that, I think many in the FIRE community underestimate how much it would suck to be in that 10% failure group, old, and with no marketable skills. However, other than somewhat naive 20 year olds looking for a quick exit, my experience is most people realize the 4% rule is a north star, a starting point, not a definitive answer. Plus, it assumes no social security, never making a penny again, no inheritance, etc. So yes, the MMM article lacks nuance and could be dangerous if folks only scratch the surface. I think most reasonably inquisitive people who are capable of saving 50% or more of their income dig deeper. We’ll find out in 20 years!

  5. Adam says:

    How are you accounting for the fact that dividends are around 2%, half what you need in the US, and presumably would not drop as much as the market, or stay the same, so you are not selling 4% each year?

    Also it seems unlikely that in a 50% drop people would blindly keep pulling 4% income. Everyone has some margin of error / fat in the budget.

    • FatTailed says:

      Numbers are inclusive of dividends for the S&P 500 for the parts of the history for which we have market returns. For years in which those are not available, I just use MMM’s 4% real total return numbers which he states are also inclusive of dividends.

      Lowering spend is certainly one response, but at a 40k/year payment how much room is really left to cut?

  6. Max Corder says:

    As a financial advisor, I work with a number of people who are retired. Some have been clients of mine for 25 years, so we have been planning for this for a long time. I know a few people who have retired “early”(late 40’s), and those are the ones who have built a business over many years and sold it for a lot more than $1million. I have never met anyone who worked and scrimped and saved to accumulate $1million by age 35. If you have children and live a “normal” life you can’t do it unless you are in unique two-earner high-income professions (in my opinion). For my clients who are now retired or who are nearing it, there are several things we take into consideration; (1) Taxes. No matter where you live, you are going to pay Federal, State and Local taxes on your income, whether it is ordinary or capital gain. (2) Social Security. For the typical couple over 65 these days, Social Security yields $30,000 to $40,000. (3) Medical expense. You will visit more doctors than you thought possible as you age, so having adequate medical insurance and savings for uninsured expenses is important, and not cheap. (4) Overhead. Even if your home is paid for, it is probably too big for easy maintenance, so you have to hire it done. It is expensive to maintain, so think of selling it and downsizing. If you decide to stay, plan on high and higher property taxes, insurance, maintenance, power bills, etc. These expenses can easily run more than $10,000 per year.

    These folks have been used to a nice lifestyle while they were working. Their living, traveling, automobile, helping adult children and also grandchildren easily costs $50,000-$60,000 per year. That’s on top of the overhead I mentioned earlier, medical expenses and income taxes.

    Many of my clients have outside income from rents and other passive investments. An investment portfolio of at least $2million would be a good start. And that is probably not enough if both members of the couple live into their 80’s-90’s and one or both of them need assisted living. I know for a fact that costs at least $60,000 per year for my mother-in-law, and in come areas of the country more than $100,000. And the expenses for the person remaining in the home continue.

    Retiring is great, but getting old isn’t for sissies or the person without adequate resources.

  7. John says:

    Reading bigERNs analysis above makes me believe that 3.3% would survive irrespective of the starting CAPE value and the duration of retirement. Do you agree?

    • FatTailed says:

      I do not. I don’t fully know the ins and outs of ERN (just looking at the model tonight) but from what I can gauge in ERN, his model does not suggest 3.3% would survive either, at least with the MMM parameters of 4% real return and 100% stocks.

      – I’ve tweaked the ERN portfolio to be 100%/0% with 4% returns for stocks.
      – It seems like this shows a 3.4% withdrawal rate would have a 25% failure rate for CAPE>30
      – I’m assuming he is averaging some of the 90s and early 2000s years when CAPE>30 but not as extreme as 99-00 when CAPE got over 40

      If I test my model against a 3.3% SWR from 2000 on, I get a 43% failure rate. From 99 on it’s a nearly 10% failure rate (sequence of returns is a bitch).

      Directionally the ERN and my model appear to be giving a similar result, if I were to average the CAPE > 30 periods I reckon I’d end up around a 25% failure rate as well.

      Right not the comparison is apples and oranges; I do need to dissect the ERN sheets to compare apples to apples.

  8. Nayan Modi says:

    Great insight. How would I go about doing something similar for my Indian markets ?

  9. Kit says:

    I agree that a lot of people appear to blindly rely on the 4% (I’ll ignore the $1MM since the 4% amounts to the same thing, but is adjustable for different levels of spending). As you pointed out sequence of return risk can play a huge factor especially in the 9 year bull market we have had.

    I am considering quitting/retiring/being lazy when we hit 25x our annular income. However, it isn’t a binary state. I am not going to look at my personal capital one day see that we have crossed $1.5MM (our number) and just stop going to work. I think I would probably work a year beyond that point and evaluate how our investments are doing and if we our comfortable pulling the trigger at that point. At $1.5MM we have enough flex that we don’t need to consistently pull out $60k each year. If we don’t need that much then we pull out less, if there is big correction early on we adjust.

    I think the 4% is great especially combined with flexibility. Anyone who hit exactly 25x their annual spending and then continued to withdraw using the 4% rule without adjusting to current market conditions wasn’t planning properly. If you retire at 35 and within 5 years that bear market has wreaked havoc on your portfolio you don’t continue to draw down you find some supplemental income (e.g. get a job of some sort (it doesn’t have to be full time)) to boost your portfolio back up and/or cover your expenses to allow your portfolio to regain its losses.

  10. Justus Pendleton says:

    I also think that MMM is wrong but I don’t think your simulation is quite as dire as you suggest. Your simulation shows that for someone who retires at age 35 there is a 96% chance they make it to age 62, at which point they would claim Social Security and be fine. Okay, maybe not FINE, since you’d be looking at something like $27,000 from Social Security, not the $40,000 they had planned on. But they’d stay retired, never be eating cat food, and have spent 45 years with their loved ones. I think a lot of people would reckon that’s pretty good odds, especially since 4% real withdrawals are extraordinarily pessimistic.

    If the simulation doesn’t take Social Security into account then it is missing an extremely large piece of the puzzle.

    • FatTailed says:

      This is very true, but how much are you going to be earning from social security if you only contributed up to age 35?

  11. Chris says:

    I suspect people underestimate how difficult it can be to find work if you are over 40, been out of work a while and in the kind of economic situation which results a severe downturn in stocks. My father was made redundant in his mid 40s during a recession, and basically never worked again despite gaining another degree during that time. He faced a perfect storm of being too skilled for entry level jobs, and out of the workplace too long for his profession when the recession ended. I’ve seen similar things happen to recent graduates, except they get stuck in low paying dead end jobs – and never quite manage to escape them after a recession.

    The flipside is I suspect most FIRE budgets could flex. A bit more frugality here and there – and you might spend 4% or even a little more during the good years, and rather less during the bad ones.

  1. December 10, 2018

    […] the space of nearly one week, my rebuttal to Mr. Money Mustache’s claim that $1 million at age 35 was sufficient to retire forever has very rapidly become one of the most read posts on this site. A reader of the article asked for […]

  2. December 14, 2018

    […] $1 million isn’t enough – Fat Tailed and Happy […]

Leave a Reply

Your email address will not be published. Required fields are marked *