One of the things I dislike about the 4% Rule / Trinity Study is the open-ended definition of success – if you have more than $0 after 30 years, great success!
Having $1 and $1 million both meet that criteria, but will have obviously different outcomes over the next 30 years.
Planning for a 60+ year retirement requires looking a little closer.
Planning for 60+ Year Retirements
Planning for a 30-year retirement is fairly straight forward. We can use a tool like cFIREsim or similar to explore different asset allocations and withdrawal strategies for our unique situation and temperament.
When we look at longer retirement periods failure rates typically rise – some of the portfolios that would last 30 years start to fail at 40, 50, or 60 years, and a reduced data set means a single failure carries more weight. This increased failure rate means many people choose to build a larger portfolio, meaning a longer career.
I think that is unfortunate.
A better way to plan for 60+ year retirements is to link two 30-year retirement periods end-to-end / consecutively. If the value after 30-years is equal to or greater than the starting value, then logically we could continue forward for another 30+ years.
Clarifying based on a trend in the comments: This means that the inflation-adjusted value after 30 years is equal to or greater than the starting value.
Linking Retirement Periods
To link two 30-year retirement periods, I looked at the withdrawal rate that would result in having the 30-year portfolio terminal value exactly equal to the starting value (inflation-adjusted), assuming constant inflation-adjusted annual withdrawals.
If we start with $1 million, in 30 years we also have $1 million. I previously explored this while planning the GCC Endowment Fund.
The following chart shows the distribution of withdrawal rates by starting year for a US-based 75/25 stock/bond portfolio. In 73% of cases, that value is greater than 4% (the thick black line segment.)
Withdrawal Rates for 30-year Parity, US-based 75/25 stock/bond allocation
For the retirement periods that had a 30-year terminal value of less than parity, I also calculated the average final value that would have resulted had we taken a traditional 4% withdrawal rate approach, e.g. 45% of starting value for the 2nd column which includes 1973.
3 years are highlighted in yellow, 2 “failures” and one extreme success. These are years that I explored in-depth:
- 1929 – What if it all went to hell?
- 1965 – The Worst Retirement Ever
- 1982 – The Best Retirement Ever
As we increase the percentage of the stock allocation, the mean and standard deviation shift upwards – a greater withdrawal rate is supported. (This is what we did with our own portfolio.)
This is shown in the following chart for a 90/10 stock/bond allocation. Starting years that had an increase in sustainable withdrawal rate due to higher stock allocation are colored in green. 1 year had a lower withdrawal rate, highlighted in red.
Withdrawal Rates for Parity after 30 years – US-based 90/10 stock/bond allocation
Chart Insights
I think this is a fascinating way of looking at the data, which makes it easy to reach some helpful insights:
- A traditional 4% approach is very solid – in ~75% of cases we have more money in 30-years then when we started
- The 1960s were terrible across the board. Studying that era would be helpful for building confidence.
- 99.9% of success is luck of the draw – it’s all about when you start
- so much for picking yourself up by the bootstraps
- Not so clear from the chart, but somewhere around 3.25% withdrawal rate results in a 30-year portfolio value worth at least 50% of starting value (90% of cases.)
- With SS providing 50% of the annual budget, a portfolio worth 50% of starting value with 4% wr is sustainable for another 30 years (See the Year 2000 study for details.)
Summary
Linking two 30-year retirement periods end-to-end, with a middle value equal to the starting value overcomes limitations of a traditional Trinity study style analysis. We benefit from the expanded data set and success is clearly defined.
With a 75/25 US-based stock/bond allocation, a 4% withdrawal rate achieves 30-year price parity (inflation-adjusted) in ~75% of cases. Increasing stock allocation generally increases possible withdrawal rates.
With a stock allocation greater than 75%, I would be comfortable targeting a 4% withdrawal rate.
For planning your own 60+ year retirement, consider giving this approach a try.
Thanks for the fresh look at this, I haven’t seen it analyzed this way before. I’ve always thought the same that having $1 at the end of 30 years doesn’t seem too “safe”. Another way that seems to be gaining some internet steam is the 1/PE10 approach to “guarantee” a SWR. I don’t recall whether you have published on that before, but it theoretically takes into account your given ER year’s potential to perform in the future…based upon past of course!
I think in actuality, there are really too many variables, including health costs, taxes, shifting economic situations world wide, etc to be “guaranteed” of anything (on either side of the ledger) based upon historical values. Much like the alternative energy debate, planning for retirement, especially significantly earlier than considered traditional, I’m thinking “yes to all strategies”. Plan for a SWR based upon what you’d like to spend, then a different lower safer SWR for your “must have” expenses in bad years, plus side gigs, plus plan and analyze when SS kicks in, etc, and most importantly be flexible!
For many people including myself and my other friends (all around 50, so not that young) targeting FI, we could likely “survive” indefinitely right now without earning another dime of income, but the reality is that we would all prefer to have more flexibility to do different things, choose where we want to live, and have a bit more security than current portfolio allows. For me currently, it’s about the tipping point of when sum of the burning desire to do something else and the “pain” of traditional employment is greater than the lack of sense of security of the portfolio’s ability to sustain us.
Even with nothing burning to do, if I feel secure enough, I’ll probably go and figure out something else to occupy my days. Conversely, if the pain of employment gets too great, ditto. It’s a balance which has not pushed me over the edge. At this age, there is also probably a momentum factor for many, including me. That is not so easy to overcome.
There are many different withdrawal strategies.
Using CAPE as a guide for the initial withdrawal rate, if CAPE is high you can probably spend 4%. If CAPE is low, then you can probably spend 5-6%.
This is a cool idea, to look at the Trinity study this way. Certainly helpful for early retirees!
But sorry, to say this, I think you’re WAY overestimating the chances of success in the specific way you do this:
– Having the starting portfolio amount after 30 years is very unsafe, because of inflation! If you assume 2% inflation per year, your withdrawal amount after 30 years will be about 80% higher than at the start. This means that if you’d retire in that year, you’d want your 80% higher withdrawal amount to be 4% of your portfolio. This means that your portfolio should be 80% larger than the starting value in order to have a situation analogous to when you started (withdrawal amount is 4% of portfolio value).
– You assume that reaching the ‘analogous’ situation after 30 years results in 100% success. It would be better to assume the same failure rate as for the first 30 years. For example, if your portfolio failed 10% of the time in the first 30 years, you’d expect the same to be true in the second 30 years. This would give a total failure probability of almost 20%, almost double compared to the 10% you’d assume in this case.
Not really. Everything is adjusted for inflation, using the actual historical CPI as is the case with all of these 4% rule type analysis. End value = starting value means equal purchasing power.
Technically speaking you have a conditional probability, but you would need to do a weighted probability. In 75% of cases spending 4% you will hit the 30 year point with more money than you started with (also inflation adjusted) so the probability of success going forward is 100%.
I like the concept but wouldn’t the mid-point need to be starting value plus 30 years x some inflation percentage? Equaling the starting value doesn’t seem to account for inflation.
It does. Everything is adjusted for inflation.
While I enjoyed using the financial simulators leading up to retirement (and still sneak a peak at them now and then) I also was never comfortable driving my ending balance after 30 years to at or near $0. I am too conservative financially to ever be content with that, and I would like to leave a decent amount to our only child.
Kudos to people who have to plan a 60 year retirement scenario. Could not have done it myself due both to my psychological makeup as well as the fact that I made the bulk of my $ during the last 15 years or so of my working career that ended at age 60. Best wishes to those who are making a go of it!
The cool thing about the 60 year retirement is that the difference in withdrawal rate between having $0 in 30 years and having the same amount as when you started (inflation-adjusted) is very small. The charts in the endowment post spell that out in an interesting way.
I like the 4% rule, it’s a good guideline. I’m not sure if 60 year is a good period to plan for, though. That’s too long. For us, we’ll just be more conservative until we’re 55. Then, we’ll go to 4% rule at that point.
3.25% + part-time work until you’re 55 is a good plan, IMO.
Winnie will probably be retired 60 years or more.
We all know the 4% rule is more like a guideline. 30 years is also just a guideline, so stacking two 30 year retirements is a good place to start. if you were to look at 3 “20 year” or 4 “15 year” scenarios you would have an even greater set of data but I suspect the results would be similar.
I think the goal here is to encourage us to look at ER earlier rather than later and not worry so much about SWR. The key to survival is flexibility. If you can be flexible in what you spend, you will likely have a successful financials outcome.
Personally I’m looking to hit the 4% “rule” based on current spending and a 50+ yr retirement but I could easily take that down to 3% if needed and still be just fine. Probably even lower with geoarbitrage and other risk mitigation.
Yes, guideline is the perfect word.
Using shorter segments pushes the withdrawal rate down – in the extreme case if we did 60 1-year segments the volatility is too high to get start/end values equal.
Great post Jeremy. Reading through these comments (and some of your previous article that share historical values) you might want to consider adding a big disclaimer to say that your numbers are (always?) inflation adjusted. This might save some confusion from the readers. Beside that, it is too early for us to see where our portfolio will be in 30 years as we reach our financial freedom only 18 months ago and have been enjoying since them a similar nomadic lifestyle that yours (thanks for the inspiration btw). We think that we have enough contingency plan in place that if we need to make more money we can pull one of them :-)
The conclusion I’ve reached is that inflation is a nebulous concept that nobody truly understands.
Another great analysis, GCC. I really appreciate your number crunching posts the most–both for tax-planning and for asset allocation strategies and projected returns.
Thanks boss. Any topic requests?
Nice analysis. I might have misinterpreted, but it seems as though you implicitly assume that the second 30-year period was the same as the 1st 30 yrs? If so, that seems a bit limiting to me … the first and second 30-year periods are independent from each other. A more realistic approach IMO would be to look at all pairs of 30-year periods, or random combinations via Monte Carlo. It’d be interesting to see how the success rates change.
The 2nd 30-year period starts with the funds you have remaining after the first 30 years. That’s the opposite of independent.
The key point here is doing whatever it takes to arrive at the 30-year point with at least as much money as you started with. Then you are adopting the same level of risk as a standard age-65 retiree using the 4% rule at that point, which is close to zero.
Another wonderful analysis Jeremy, thank you! It would be awesome if you could also please provide some analyses, tax minimizing techniques, etc for individuals planning for/working towards high retirement incomes (aka “Fat FIRE”) who are also living in high tax states such as California. It seems most of your strategies such as capital gain harvesting, Roth conversion ladders, etc would unfortunately not apply to folks with high retirement incomes. However, any recommendations you could provide for this subset of the FIRE community would be most appreciated. Thanks again!
The best Fat FIRE strategy is to find peace with the fact that you’ll pay lots of taxes.
If you like, you could spend a year abroad to get things started on strong footing. See kickstart your retirement.
First comment here! Thanks for the straighforward analysis, I appreciate how simple you make this concept, generally ignoring numerous other factors that just get in the way (social security, spending less as you age, etc). If X was good enough when you start the first 30 years, then it’s going to be good enough when you start the second 30 years. Simple.
Nice job.