If you’re been reading here for a while, you may have sensed that we’re borderline obsessed with ensuring that we always have access to affordable health care, especially in the years between leaving our jobs and qualifying for Medicare. As lots of knowledgeable readers have shared, health care is nowhere near “affordable” for anyone earning in the higher five figures and up, and we care about keeping those costs down so much that we’re willing to shape our retirement income around controlling our health care costs. In other words: We’re willing to live on less than we can technically afford in order to gain some level of certainty on health care costs. This and keeping our house cold are our weirdo areas.
Spoiler alert: This is not a post about health care.
I bring up health care only because it plays into a larger issue that we’ve struggled to get our heads around, which we’ll call our optimal retirement income: a level at which we get a big Obamacare/ACA subsidy on our health insurance, we pay low taxes and we enjoy a comfortable standard of living. While some of this optimization may be more straightforward for investors who focus entirely on dividend investing, as opposed to our index fund and rental income strategy, anyone with multiple streams of passive income will have to confront this math at some point.
While looking at the ACA subsidy levels is a good place to start, unfortunately it’s only part of the equation, because income as defined by the taxman and retirement cashflow are not necessarily the same thing. (Shoutout to Justin at Root of Good for writing such a great ACA subsidy post that I link to again and again!)
What this post is really about: The income vs. cashflow discrepancy in retirement
To figure out our optimal retirement income, we have to look at both income and cashflow, how they’re each calculated, and how they each impact our taxes, ACA subsidy and overall quality of life. Let’s dive in!
This post gets into taxes and income, so all the usual caveats apply: We’re not financial planners or tax professionals. Consult the pros regarding your own situation, and know that we offer this to spur thinking, not to serve as financial advice. You’re grownups — you know the deal. :-)
Everything in this post comes back to the chart below. Don’t worry, we’ll zoom in and break it down.
The big picture point, which you can get at a glance from the different colored bars, is that sources of income may have a different impact on our cashflow than they do on our taxes, and they may have a different impact over time. In our quest to find our optimal number, we have to factor all of this in.
Optimizing Income for Taxes and ACA
The ACA subsidy limits correlate loosely to tax brackets, so we’re going to short hand “taxes and ACA” to just “ACA.” Basically, if we can get a big subsidy, then we’ll also be in a very low tax bracket. (Which we’re only okay with because we’ve paid a lot of taxes so far — we think of it as paying it forward.)
Because we don’t share our numbers, we’re talking about all of this conceptually, which may be easier anyway, since it won’t unintentionally anchor you to the targets we’re aiming for. Instead, you can apply your own circumstances.
The technical paragraph: The ACA subsidies are based off the modified adjusted gross income (MAGI) reported on your tax return for the year in which you receive the health benefit (meaning: if we receive a subsidy now in 2016, we’ll make sure everything is square when we file our 2016 tax return next year). The downside is that you have to guess, and if you guess too low, you could end up having to repay part of that subsidy on your next tax return. The upside for early retirees is that we have a lot more freedom than regular workers to reverse engineer our income to fit into the constraints we wish to fit. Unless you’re still making IRA or HSA contributions in retirement, which you can subtract when calculating AGI and then MAGI, MAGI in retirement will probably be pretty close to your full gross income. You don’t get to take out either the standard deduction or itemized deductions to determine your ACA subsidy eligibility, and if you claimed student loan interest in calculating your AGI, for example, you have to add it back in to find your MAGI. Not cool, I know, but then maybe that’s another incentive to pay that loan off ASAP.
So thinking about all of that, we know we have to fit all of our income streams into our MAGI and still stay under a certain limit for ACA and tax purposes. The black bar in the chart below represents the current subsidy limit that we’re eyeing. We know it could change, but we have to plan around something.
Within that limit, the two things that must fit are the two income sources we can’t avoid: our net rental income (rent received minus depreciation, mortgage interest, insurance and maintenance — determined on its own schedule within the tax return) and the dividends on all of our taxable accounts, which we’re taxed on whether we actually take those dividends out or reinvest them. But we also hope to have ample room within the limit to take some capital gains from shares we’ll sell, to earn a little income from fun work and, in years with bad dividends or market gains, to convert a few dollars from our 401(k)s into a backdoor Roth.
One big change for us in retirement will be how we treat dividends. We’ve always reinvested them, but because we’ll be paying the taxes on those dividends anyway, we might as well use those dollars to support our lifestyle instead of us paying more tax for more capital gains because we sold more shares while reinvesting dividends. That feels a little like double taxation, which doesn’t fit with a good tax optimization strategy.
Also notable in the “income” calculation is that “rental income” doesn’t necessarily equate to “rental cashflow,” because most of that money goes to pay the rental mortgage for another decade-plus, and cost basis (what we paid for the shares we’re selling) isn’t taxable, meaning our taxable income is skewed compared to our actual cashflow on at least those two measures (too high on rental income, too low on cost basis).
So based on the ACA subsidy limit we’re eyeing, we do have one number (our maximum taxable income), but it’s not THE number, the number we’ll actually live on. To calculate that, we have to go another step, from income to cashflow.
Optimizing Cashflow in Early Early Retirement
We will often talk about our “income in retirement,” when what we really mean is our cashflow in early retirement. As this chart represents proportionally, more or less, the biggest chunk of our retirement cashflow, at least in the first 10 years or so of early retirement, isn’t income at all: cost basis. Really we’ll just be getting back what we put into our investments, with some percentage of growth taxed as capital gains. But at least in the early years of our early retirement, those gains will be a small percentage and the cost basis itself will be a bigger percent.
Rental income will also be a smaller share of our cashflow than it looks like in our taxable income breakdown because the income calculation doesn’t factor in the full cost of mortgage payments, which eat up most of the rent.
Finding THE Optimal Number
Lots of people determine their retirement cashflow number by using the 4 percent rule, and that’s certainly a valid way to go, assuming your funds are all accessible when you need them, and not locked up as tax-deferred funds you can’t touch without penalty until age 59 1/2. (In our case, we think the math is more complicated, and have done extensive number crunching to figure out how much we need for phase one of our early retirement, before we turn 60. Plus we just sleep better at night knowing we’ll be mostly leaving alone our tax-deferred funds until our later years.) But even without the easy 4 percent rule to help, we still know roughly what we can safely spend in phase one of retirement, and we can plug that into this formula:
At a basic level, you want to ensure that everything you plan to live on — the “income,” the cost basis of the investments you plan to sell and any adjustments you need to account for — all fit within what you can comfortably afford to spend each year without running out of money. You can certainly go deeper than this, and do some projections like these, but you’ll have to make some guesses about likely interest rates and when the market might tank, which is beyond the scope of this or any post.
Optimizing Cashflow in Late Early Retirement
Just for fun, let’s take a look at our late early retirement (approximately age 50 to age 60).
Though a ton could change between when we pull the ripcord and then, we expect to see two main shifts: 1. the rental property will be paid off, so the rent will be actual cashflow for us, and 2. we’d ideally see a higher proportion of capital gains and lower cost basis in the mix because our investments will have had more time to grow. The second factor is highly market-dependent, because how many shares are left to sit and grow depends on how much we have to sell in our early years. The proportions in this chart are what we would hope to see!
Recap of Income vs Cashflow
We’re 1600 words in and almost there! Recapping the whole idea, we see that “income” in retirement will be smaller than cashflow for anyone whose retirement plan includes selling shares of stock or bonds, but also that income doesn’t fully account for every aspect of cashflow, namely cost basis. Also, over time, the proportion of different cashflow or income streams can change based on market conditions and other factors like paying off mortgages. Fun!
Once We Get to “Traditional” Retirement Age
Because we’re building our income and cashflow to fit around ACA limits, we know that we’ll have a bit of a Rube Goldberg system in place. The good news: We can cut it all out once we reach Medicare eligibility at age 65.
Medicare does make some adjustments for income, but the different limit levels are far less punitive for high earners, so once we both hit 65, we plan to cut loose a little more. At the very least, we won’t have to be so careful about constantly engineering our income to stay within the bounds of the ACA box. Or — who knows? — maybe everything will change in the next 25-30 years! And if that happens, then we’ll have another monster post on our new calculations and plans. :-)
Is anybody else thinking about the income vs. cashflow discrepancy? Have you factored what you expect to live on each year and how that equates to income, which in turn affects your subsidy eligibility? Anything we missed in this post? Any thoughts we spurred? Let’s get into it in the comments!
Want extra Our Next Life content? Get the e-newsletter!
Subscribe to get our periodic newsletter with tons of top secret, behind-the-scenes info we'll never share here on the blog.