I built a Roth conversion optimizer because retirement is the phase I’m in. That sounds reasonable until you notice what it leaves out: by the time anyone needs a conversion plan, the big decision (traditional or Roth, every paycheck, for decades) has already been made. So I built the piece I was missing: a Roth vs traditional 401k calculator that optimizes the working years and the retirement years together, as one plan.
It’s free and needs no signup. You enter your salary, when you plan to stop working, what you spend, which accounts you have, and how your employer matches. It tells you how much to put in each account this year, and every year after that.
The phase I wasn’t looking at
Everything in my Roth conversion calculator starts the day you stop working. You arrive with a traditional balance, a Roth balance, maybe a brokerage account, and the model figures out how much to convert each year and which account to spend from. I wrote it for my own situation, and it does that job well.
What I didn’t see for a long time is that I was solving the second half of the problem. The size of that traditional balance, the thing the whole conversion plan revolves around, is the result of thousands of small choices made earlier: pre-tax or Roth on every paycheck, whether the IRA money went in deductible or not, whether the extra savings went into the 401(k) or the brokerage. By the time you’re my age, those choices are history. You can only clean up after them.
That’s a bias, and a pretty human one (you optimize the problem sitting in front of you). But most people who search for this stuff are still working. For them the contribution choice isn’t history, it’s next payday. So the new tool starts there.
Why it’s one model and not a separate calculator
Most Roth vs traditional calculators compare two numbers: your tax rate now and a tax rate you guess for retirement. If the future rate is lower, go traditional; if it’s higher, go Roth. That rule isn’t wrong, but the “rate in retirement” isn’t one number you can guess. It’s whatever bracket each dollar actually comes out in, and that depends on decisions made later.
A traditional dollar saved at 35 might come out at 12% in the years between retirement and Social Security, when your income is low and you can convert to Roth cheaply. Or it might come out at 22% or more because required minimum distributions pile on top of a pension and Social Security. Or it sits in the account until the end and gets taxed at whatever rate your heirs pay. Which one happens depends on how much traditional money there is, and that’s the thing you’re deciding. It’s circular, which is exactly the kind of problem a linear program handles well and a two-rate calculator can’t.
So instead of building a separate calculator, I put the working years in front of the model I already had (the one I described in the Roth conversion math post). Each working year now has new decisions: traditional 401(k), Roth 401(k), traditional IRA, Roth IRA, HSA, and whatever is left goes to the brokerage. The employer match, contribution limits, catch-up contributions, FICA and the child tax credit are all in there. Then the same solver picks contributions, retirement conversions and withdrawals together, maximizing what you keep after tax at the end of the plan.
It also runs your inputs twice more: once forcing every contribution to traditional and once forcing everything to Roth. Both of those runs still get to optimize conversions and withdrawals in retirement, so the gap you see is purely the contribution choice. Limits are the 2026 figures from the IRS announcement ($24,500 for a 401(k), $7,500 for an IRA, plus catch-ups), and you can edit them.
What it said for a 35-year-old on $110,000
The sample scenario on the page is a single 35-year-old earning $110,000, with 1% raises above inflation, retiring at 60. The employer matches 100% of the first 4%. They have $60,000 in traditional accounts, $15,000 in Roth and $20,000 in a brokerage account. They spend $60,000 a year while working and $55,000 in retirement, and expect $28,000 a year of Social Security at 67. The return is 5% above inflation, the plan runs to 92, and anything left in traditional accounts at the end is valued after a 22% tax.
This year the plan puts $24,500 in the Roth 401(k) and $7,500 in a Roth IRA, collects the $4,400 match, and converts $11,800 of the existing traditional balance to Roth. Then it changes its mind as the career goes on. The first couple of years are all Roth, the middle years are a mix, and in the last stretch before retirement (from 52 to 59) everything goes traditional at the $32,500 catch-up limit. Over 25 working years about 61% of the employee contributions end up in Roth.
The late traditional money isn’t an accident. Once the paycheck stops at 60, the plan converts $66,500 in the first year and $44,209 in the second, filling the cheap brackets before Social Security starts. Over the whole plan it converts $178,254. In other words, it deliberately saves pre-tax late in the career because it already knows there will be cheap years to move that money later. That’s the interaction a two-rate calculator can’t see, and it’s the same logic behind how full to fill your tax bracket each year.
And the money involved: the optimal mix ends with $6,712,726 after tax (in today’s dollars). Always-traditional ends with $6,558,226, so the mix is worth $154,499 more (+2.4%). Always-Roth ends with $6,488,204, so the mix beats that by $224,521 (+3.5%). Not life-changing as a percentage, but it’s a six-figure difference for the price of filling in a form.
Sometimes the honest answer is “it doesn’t matter”
I expected the tool to always find some clever split. It doesn’t, and I think that’s one of its more useful answers. I ran a 25-year-old earning $45,000 who spends $33,000 a year and works to 65. That person sits in the 12% bracket now, and the plan keeps them in the 12% bracket in retirement too. The optimal mix, always-traditional and always-Roth came out exactly the same, to the dollar.
When the rate going in and the rate coming out are the same, the choice is a coin flip, and the page says so instead of inventing a recommendation. That’s worth knowing (it means you can stop agonizing and pick based on other things, like wanting Roth money you can reach early). The decision starts to matter when today’s rate and the later rate are different, which is the case for most mid and high earners, and for anyone with a long gap between stopping work and claiming Social Security.
What it assumes, and what it doesn’t do
Like the conversion tool, this is a model, and models simplify. The ones worth knowing before you trust a number:
- Everything is in today’s dollars. Returns, raises and contribution limits are after inflation, and the limits are held at their 2026 values.
- Spouses are assumed to be the same age.
- The employer match goes to the traditional account unless you tick the Roth match option.
- HSA money is assumed to be spent on medical bills, so it’s never taxed going in or coming out.
- The backdoor Roth assumes you have no pre-tax IRA money, so the pro-rata rule doesn’t apply. You can turn the backdoor off and use the income limits instead.
- After-tax (mega-backdoor) 401(k) money and the overall $72,000 annual additions cap aren’t modeled.
The biggest assumption is the one you type in yourself: the return and the tax rate on whatever is left at the end. Change those and the split changes, which is the point. This is a planning model, not tax advice, so check anything you act on with a tax professional.
One more thing that’s new for my tools: there’s an MCP server, so an AI assistant that supports MCP can run the optimizer itself and come back with the numbers instead of a rule of thumb. The same server pattern already works for my shift scheduler and route optimizer.
If you’re still working, run the Roth vs traditional optimizer with your own numbers. If you’re already where I am, the how much to convert each year post and the conversion calculator pick up from there.
Things that I use, like, and am affiliated with:
Mint Mobile offers great cell phone service for $15 flat, get $15 off using the link. Get discounted phones with service activation and no contract.
I never spend money before I check Mr Rebates or Rakuten to get cashbacks, rebates, discounts, coupons or cheaper gift cards.