It’s time for another Safe Withdrawal Rate case study today! Believe it or not, but this is already the ninth installment of the series! Check out the other case studies here. Today’s volunteer is Mrs. Wanderlust (not her real name), a frequent reader of the ERN blog. She and her husband plan to retire in 2018 (more or less voluntarily) and asked me to run their numbers. One challenge in pinning down a safe withdrawal rate: large additional cash flows because they plan to purchase of an RV and then sell it a few years later. They will also have different budgets during different phases in retirement. And not to forget, a four-legged family member that’s factored into their planning. So without further ado, let’s start calculating…
Mrs. Wanderlust’s Situation
Ages: 43 (me), 42 (husband).
I’m getting laid off early next year (my choosing – this is my ER event and part of our financial plan – it’s a windfall of a severance). My husband will follow next September. We think we’re going to follow the glidepath investment strategy, but starting at 70% equities. My question is mainly around our SWR. Our unique situation is that we’d like to buy an RV (while keeping our home) and travel 6 months/year for 4 – 5 years. I have a budget worked out for this, but am nervous to have the added expense during the highest risk of sequence returns period. Our budget would drop after we’re done RV’ing (and the remaining equity in the RV would be put back into the market at that time), but would increase slightly after our dog dies (yes, our dog dying is written into our plan – stop laughing!). After he passes, we plan to slow travel for 6 months/year.
OK, I’m impressed! You plan for everything, even the life expectancy of your dog! I thought I had seen everything but this is very impressive! Way to go! Let’s look at the portfolio:
401k: $1,095,691
Roth IRA: $31,890
Taxable investments: $515,851 (80% is cost basis, 20% is capital gains)
Checking/saving: $22k
Your budget in retirement?
2019 (or 2020?) – 2023 (RV years): $65k
2024 – 2028 (post RV, dog alive): $60k
2029 – 2034 (slower travel, post dog years): $65k
2035+: $75k (401k becomes fully accessible in 2035)
Note – …we can reduce our budget by only 10% during the first 10 years- it’s pretty scaled back to start with. One of our questions – how strictly do the numbers say we need to be to the budget? Can we scale up by a few thousand? We do not have children so our portfolio goal is not preservation but spend-down.
Other information: Additional Cash Flows?
When I leave work, it will be via a lay-off. My company has had regular lay-offs for over 15 years and I’m positioned well enough (and with enough experience) that I can coordinate my exit with a lay-off. Yay me! My severance package will be around $107k net. Timing is sometime in the first half of 2018. Let’s call it a best/worst case scenario of having a last day of working on 2/28/18.
We also have a cabin/lake home we will put on the market next spring. After realtor’s fees and taxes, I’m estimating we’ll net $100k (note, this is quite on the conservative side).
We are budgeting $100k for buying the RV and $70k for selling it; the exact purchase timing is something we’re interested in. Do the numbers say we can do this in late 2018, or should we wait a few years, say until 2020? Stupid sequence of returns risk. 😀
Social Security:
- Mrs Wanderlust: Age 62 – $1,600 Age 67 – $2,280 Age 70 – $2,827
- Mr WL: Age 62 – $1,362 Age 67 $1,934 Age 70 – $2,398
A pension: It’s a lump sum payout pension, currently worth $2,577 with an established annual COLA change of 4%. Stop laughing at the amount. 😉
Definitely, the severance package and the sale of the cabin will help with cash flow crunch! If possible it’s always best to structure your departure from work as a layoff because of the additional benefits, both from the workplace (severance package, benefits, etc.) and the government (i.e., unemployment benefits).
For now, let’s assume that you buy the RV in late 2018 and start with your RV trip in 2019. Your RV vs. non-RV budget is only $5k apart ($65k vs $60k), so why delay?
I reconfirmed with Mrs. W. that the pension is not $2577 per year, but one single payout. A 4% safe return is probably nothing to sneeze at (better than a CD!) but in the big scheme, it’s just a rounding error. You should probably keep that as a reserve for repairs on the RV. I will not factor that into the calculations here.
Social Security Timing
As I have said in these case studies before, it’s too early to commit to anything. But as a general rule of thumb, it’s best for the spouse with the higher benefits to wait until age 70. So, just because I have to make some assumption for the calculations, let’s assume that Mr. Wanderlust claims benefits at age 62 (when Mrs. Wanderlust is 63) and Mrs. Wanderlust who will likely have the longer life expectancy, claims benefits at age 70.
The Safe Withdrawal Rate Simulations
First, let’s calculate portfolio values as of 1/1/2019. You indicated that you still contribute a combined $14,950 into your 401k plans in 2018 and $10k in taxable savings. $107k from the severance package plus $100k flow into the Savings account. I also assume that in the latter half of the year, you use $100k from the savings account for the RV purchase. I also assume that you’ll need $30,000 in living expenses in the 2nd half of 2019 (half of your $60k non-RV retirement budget).
The $1,817,879 figure (projected NW in today’s dollars) is what we’ll use as our baseline.
—> Link to the Google Sheet <—
(Note on the Google Sheet: You can’t edit this sheet. You first have to download your own copy via File->Make a copy)
I also calculated all the supplemental cash flows (see chart below) that I like to take into account during the withdrawal phase. Most prominent is the one month where you sell your RV and you get back around $70k nominal (less in today’s dollars), over 3% of the portfolio value in today’s dollars. Also, let’s assume that the $65k is the baseline consumption level. We’ll factor in the lower-expense post-RV-years as a positive $5,000/year cash flow, the post-2035 years with $10,000 higher expenses show up as a slightly negative cash flow need (see the dip between months 193 and 240). Social Security for Mr. W. (after 240 months) and Mrs. W. (after 325 months) create two more small jumps later in retirement. The final long-term supplemental flow is actually quite substantial, around 1.8% of the initial Net Worth, adjusted for inflation. And that’s already after a 15% haircut to account for potential benefit reductions in the future!
Other assumptions in the Google Sheet:
- Asset allocation (static for now) is 70% stocks, 25% bonds, 5% cash/money market.
- 55-year horizon.
- Capital depletion, i.e., 0% final value.
Here are the results:
How much can you withdraw? Using a 3.9% SWR is pretty close to the fail-safe. Conditional on an elevated CAPE Ratio (20-30), a 3.9% WR will have a 95% success rate. 3.9% should sustain close to $71,000 annually. So, after taxes, that’s pretty close to your $65k target. It sounds like you budgeted just about right! Did you already perform the calculation at home and just had me check again? 🙂
(For full disclosure, there was one even lower SWR in 1901. So the overall fail-safe SWR is only 3.76%.)
How about a glidepath?
We like the idea of the glidepaths, though 60% equities seems too low for our start point. We think we are interested in starting at 70% and scaling up to 90%…but what do the #s say?
Great question! I wrote about that topic in Part 19 and Part 20 of the Safe Withdrawal Series. A glidepath would mean you increase your equity weight over time. This would expose you to less equity risk early on when you’re most subject to Sequence of Return Risk while scaling up the long-term equity share to make the portfolio last through the decades.
Currently, the Google Sheet is not set up to calculate the SWR for glidepaths for every single starting point in the backtest. But in the tab “Case Study” we can spot-check some of the unfortunate historic retirement cohorts, such as the November 1965 cohort that had the lowest SWR over the last 100 years (even lower than the Great Depression!!!). Here I compare the baseline static 70/25/5 allocation to one that starts with 65/35/0 and scales up the equity weight by 0.2 percentage points per month all the way to 100%. The initial SWR was 3.94% (picked to exhaust the portfolio under the glidepath assumption). The static allocation allowed for only a 3.81% SWR and would have run out of money after about 40 years when using the 3.94%, see below! I call this the SWR butterfly effect: A mere 0.13 %-point difference will shorten the life of the portfolio by 15 years!
The net-net: Don’t expect miracles from a glidepath. Even with a pretty aggressive path that glides all the way to a 100% equity share in the long-run you’ll increase the SWR by only 0.13 %-points (from 3.81% to 3.94%) for the 1966 cohort. For a $1.8m portfolio that’s $2,340 of additional sustainable withdrawals per year. Helpful but not that significant!
Cash Flow Simulations
I like to simulate how the account balances evolve over time and what’s the best withdrawal order and whether there are any issues with running out of the taxable account before age 59.5 or over-accumulation of 401k accounts and a nasty tax surprise with RMDs.
Here are the assumptions:
- We don’t want to reveal where exactly Mrs. W. lives, so I assume a 6% flat tax rate on the state level without revealing the exact state tax brackets. The calculations will be very close to reality! I also assume that the Trump/GOP tax plan will pass in its current form with a $24k standard deduction, $19,050 in the 10% tax bracket and 12% up to $77,400 in 2018, adjusted by 2% assumed inflation every year.
- Let’s use 100% equities in the taxable account and Roth. 60/40 in the 401k. Notice that the initial allocation is pretty close to 70/25/5 and will slowly shift more into equities due to the Roth Conversions! There’s your glidepath!
- I would keep a pretty low cash balance considering you are running low on taxable account balances. I assume you withdraw $20k per year for several years. After you sell the RV (proceeds go into MM account) and you get down to $30k (in today’s dollars) you let the cash balance grow with inflation (2%) and just withdraw the interest income.
- For the first 11 years, do the Roth Conversion ladder. You fill up the standard deduction and 10% bracket with conversions. In the table below, this will show up as a withdrawal from the 401k and a “negative” withdrawal, i.e., a contribution to the Roth.
- Initially, you draw down the taxable account to balance the budget. In 2030, after 11 years, that account should be pretty much depleted. You then leave that account alone and withdraw only the dividends.
- For the years 2030-2034, you’ll keep doing the Roth conversions as before but you also withdraw previously contributed principal into the Roth. That will be quite a big number. For example, in 2030 you convert $53,552 and withdraw a total of $33,348+$53,552=$86,900 from the Roth (see the two cells marked with the red edge, second panel). The net withdrawal is only $33,348, but you actually fund almost your entire budget from the Roth. This goes on for 5 years until you reach age 59.5!
- Starting in 2035, you can finally withdraw from the 401k penalty-free. (I know, the SEPP/72t option exists, but let’s use that only as a last resort!!!) Let’s assume you withdraw the 2% dividend income from the Roth and the rest from the 401k to balance the budget.
The plan seems to work. Even with relatively modest nominal return assumptions, you draw down the portfolio only by a bit. You still have $1.5m (in today’s dollars!) at age 70 when Social Security kicks in and you’re home-free at that time. Also, notice that you’ve engineered a pretty nice transition of assets out of the 401k and into the Roth, so there is no issue with RMD (Required Minimum Distributions) and a big tax bill later in retirement. Of course, everything could be different if we have a repeat of the 1965 cohort, but let’s cross our fingers that this will not happen.
Conclusion
This is probably one of the closest calls in the history of Big ERN case studies. The SWR study reveals that the $65k annual budget comes pretty close to the maximum amount recommended for your situation. The withdrawal strategy would utilize the taxable account first while doing the Roth Conversions, then withdrawals of Roth IRA principal, then the 401k and Roth once you reach age 5905. It’s a bit tight and could be thrown off balance by a major bear market early during your retirement. But you still have several levers to deal with the unexpected. A spending reduction (and it might be feasible to reduce spending by more than 10%, check out Mr. Money Mustache for tips) and the SEPP/72(t) penalty-free withdrawals from your 401k.
So, it looks like you’re good to go! Best of luck and enjoy well-deserved early retirement!