How We Calculated Our Numbers for Each Phase of Early Retirement

today’s post comes out of several requests we got in response to our reader survey (which you can still take if you’re interested!), asking how we arrived at our magic numbers for the amount we need to save to retire early. and even for those of you who’ve built loads of spreadsheets like we have, we know that it’s often helpful to look at other people’s calculations to test your own thinking. this is probably our most math-heavy post ever, so please let us know — how can we make mathy posts like this as helpful as possible? we’d love the feedback!

in some ways, people who retire at the traditional retirement age have it easy. they really just have to calculate one number: how much do i need to save for retirement? that can happen pretty painlessly if you know approximately how much you’d like to spend each year and apply the 4 percent rule (which, in practical terms, means multiplying your yearly budget by 25). voila: retirement magic number. (of course it can be much more complicated if you expect to have income from multiple sources, or have other circumstances to factor in, but you can still look at all your money as a whole if you’re planning to retire past age 60, which is not something early retirees can necessarily do.)

in our case, if we looked at how much we hope to spend each year in retirement and applied the 4 percent rule (25x), it would look like we could already retire already, or at least we could once we finish paying off the mortgage on our house, because the combined balanced of our 401(k)s and taxable savings is roughly equal to what the 4 percent rule says we should have saved. there’s only one little problem: a large majority of our cash is saved in our highly restricted 401(k)s, which we can’t legally touch without penalty until age 59 1/2. and sure, we could do the roth conversion ladder trick, except 1. as risk-averse investors generally, we like knowing that we’ll have a bigger pot of gold waiting for us in our later years, so don’t want to dip into our 401(k)s very much or at all, and 2. the conversion ladder wouldn’t let us move over enough cash each year to make enough of a difference, or it could mess with our obamacare subsidy eligibility. if we just retired today and started spending according to the 4 percent rule, we’d run out of accessible cash long before we can legally get at our 401(k) funds. so we needed a way other than the 4 percent rule to calculate what we’d need in each phase of our retirement.

 

defining the phases of our retirement

the chart below will look familiar to regular readers. this is how we believe our income will break out over the years of our early and traditional retirement:

essentially, we think of our retirement as happening in two phases:

phase 1: until mr. onl turns 59 1/2 (let’s call it 60), when we’ll be relying primarily on taxable accounts (this post has more detail on the allocations within those accounts), and later on a combo of taxable accounts and rental income once our rental property mortgage is paid off. we think of phase 1 as our “dirtbag years” (borrowed lovingly from eat the financial elephant), when we’ll live and travel more cheaply (think lots of camping, staying in hostels and pensiones abroad, being willing to work in exchange for services and/or food, etc.).

phase 2: after mr. onl turns 60, when we’ll live off of our tax-deferred accounts, currently our 401(k)s, but which will become rollover iras after we leave our careers. (and hey, maybe we’ll get some social security, too, but we’re not banking on that.) these are the years that we hope to make our “golden years,” living and traveling with a little more comfort since, if all goes as planned in phase one, we’ll be darn tired from circling the globe and adventuring, and in need of some pampering by the time we reach phase 2! ;-)

building our projection

to figure out how much money we need in each phase, we need to have several pieces of info handy:

1. how much we plan to spend per year in each phase (we expect a lower total during phase 1 — the dirtbag years — than in phase 2), in 2016 dollars
2. where that income will come from in each phase or sub-phase (like our rental income kicking in partway through phase 1)
3. a guess on how we expect the markets to perform between now and then, based on historical averages, factoring in historical averages for inflation as well

we know there are lots of calculators out there like cfiresim, which let you enter different assumptions and forms of income that will kick in at certain dates, but we wanted to build our own projections that we could keep handy and fiddle with.

a sample projection

because we don’t share numbers (though we do share charts!), we’re sharing a sample projection to show how we went about thinking through our retirement math. it’s based on fictionalized scenarios that loosely mirror our own situation, but with different numbers. the whole calculation is based on the following assumptions:

1. an “income”/spending target of $30,000 each year for the first part of phase 1, until our fictionalized rental mortgage is paid off, then dropping to $15,000 per year for the rest of phase 1, with rental income making up the difference.
2. an increase to $60,000 a year in our more baller “golden years” of phase 2.
3. everything is represented in 2016 dollars for the sake of understanding, and interest rates are decreased accordingly. given that the s&p 500 index has historically returned between nine and 10 percent a year, and that inflation is typically around three percent a year, that means that a high-end assumption of six percent growth (9 percent historical return minus 3 percent inflation) is a good starting point. however, because we’re somewhat conservative as investors, we also built out the projection to include lower returns of four percent and five percent. (and you could just as easily go lower still — not a crazy idea, given recent market returns, and the more pessimistic future projections of some analysts.)
4. a decent amount in tax-deferred accounts in this scenario — let’s say $500,000 just to have something nice and round to include. those will grow without withdrawals until age 60.
5. we’re okay spending down our taxable funds by age 60, when we switch to tax-deferred. (not counting an emergency fund, which isn’t factored in here.)

here’s what a projection with these made-up numbers might look like at different interest rates:

**PRETEND NUMBERS BUT REAL INTEREST RATES USED**

in the taxable column, which funds phase one through age 59, for each of the first nine years, we subtract $30,000 and then add the applicable interest rate for that projection to see what we’d have leftover at the end of the year. for the next ten years, we only subtract $15,000 before adding interest. by fiddling with the starting numbers, we can see how much we’d have to save to have funds to sustain us until age 60. in the tax-deferred column, we let the fixed starting amount grow at the assumed interest rate for the first 19 years, and then begin taking out $60,000 worth of 2016 dollars per year before adding interest.

we see, for these imaginary numbers and assumptions, that we’d want to have at least $275,000 saved in taxable savings before retiring, assuming those funds need to get us through 19 years supplemented by rental income, but that number assumes a six percent return, which is basically the historical average, adjusted for inflation, for every year of our early retirement. that feels a bit aggressive to us. to be safer, and to allow for the possibility that market returns may be declining over the long run, we should probably save $305,000 or more in this scenario, which only relies on a four percent return (seven percent minus inflation). in our real life numbers, we’re saving quite a bit more.

the tax-deferred side is where we see what a giant difference interest rates can make. a two percentage point difference could mean a variance of more than $400,000 by the time we reach age 60, almost equal to our starting balance when we stop contributing. and if you could get a consistent six percent return (nine percent minus inflation), you would never deplete your principle at that spend rate. even at five percent (eight minus inflation), the principle depletion is minimal. the tax-deferred projections are largely academic to us, since we already have more than we expect to need in that column (again, not our real numbers in the chart, but we’re well padded in that column). but if you need to figure out how much to have saved in 401(k)s and iras before retiring, you could structure that half of the chart the same way we set up the taxable side, so that you can fiddle with the starting number to figure out what you need to save in your tax-deferred accounts before calling it quits.

just for fun, we zoomed down into the true golden years, with those same interest rates, and assuming we have $500,000 in tax-deferred funds when we early retire, as well as a $60,000 a year (in 2016 terms) annual withdrawal from those accounts:

this is just a continuation of the same chart above, and you can see that with a four percent return (really seven percent market average, but adjusted for average inflation), we’d run out of money in this imaginary scenario when mr. onl is 85, and i’m 82. i’m still not convinced that immortality is off the table, but in any case, i prefer to see our money last until i’m 100 (making mr. onl 103), just to be safe. crazy, isn’t it, what a huge difference a single percentage point can make over time? the difference between a seven percent market return on average over those years compared to an eight percent return (represented as 4 and 5 percent on this chart) is the difference between a $1.5 million deficit and a three-quarters of a million surplus. the difference between eight and nine percent is a whopping five million inflation-adjusted dollars. if that isn’t a case for making sure you’re not paying 1% fees on your investments, i don’t know what is.

obviously we hope the last scenario comes true, so we can do some major charitable giving and leave big, fat bequests to some worthy groups. but if it’s looking like the first column is more in line with reality, we’ll have to adjust our spending accordingly (or maybe we get social security or sell a house or any number of other contingency measures). in our actual planning, we will calculate a new spend total for our “true” retirement/phase two based on the 4 percent rule, once we get closer to age 60 and have a bit more clarity about what else is out there in 20+ years (social security? medicare? soylent green? universal socialism? the zombie apocalypse?). only time will tell!

has anybody used a projection like the one we made (and then remade here with pretend numbers)? we haven’t done such a math-heavy post like this one before so let us know: was this helpful? what other info would be helpful to think about your own projections and planning? share, share away!