[FN-04] The Order-of-Operations Map

Which account gets the next dollar while you are still earning — answered for someone who intends to retire, which changes two of the rungs and adds a constraint the standard waterfall does not have.

Pillar: Foundations · Applies to: Anyone still earning and saving toward a retirement date, on time or early Last verified: August 2026 · Refresh cadence: Evergreen (the order); contribution limits via FN-02 Related: ER-05 Accessing Money Early · ER-02 Roth Conversion Ladder · HC-04 HSA Mastery · TX-02 Roth Conversion Strategy · TX-03 Withdrawal Sequencing · SS-05 Coordinating SS and Medicare · ER-07 Coast and Barista FIRE · TR-01 Five-Year Runway

Not advice. This page stops at your retirement date. Everything about taking money out — the order, the penalties, the MAGI consequences — belongs to TX-03 and ER-05, and the two questions have different answers. It is also deliberately narrow: it covers the account-choice decisions that shape a retirement transition, not saving in general.

---

TL;DR

• The generic savings waterfall is written for someone retiring at 65, and its only input is the current-year tax bill. For anyone retiring earlier it is missing a constraint.
• That constraint is accessibility. Retiring at 50 on $70,000 means bridging to 59½ — $665,000, or 44% of a $1,500,000 portfolio, that has to be reachable without penalty.
• A saver who maximised the 401(k) and nothing else reaches $150,000 of it. The shortfall is $515,000, and contributing more does not fix it — composition does.
• Traditional-now-and-convert-later usually beats Roth-now for this audience, because an early retiree has low-income gap years by construction. In §4: $54,731 against $37,139, a 47% advantage.
• Where the map ends: the retirement date. TX-03 and ER-05 take over, and they answer a different question.

---

1. The standard waterfall, and where it fails

The conventional order for the next dollar is uncontroversial and mostly right:

	Rung	Why it is there
1	Employer match	An immediate 50–100% return; nothing else on the list competes
2	HSA, if HDHP-eligible	The only triple-advantaged account, and the retirement healthcare fund (HC-04, HC-07)
3	High-interest debt	A guaranteed return equal to the rate
4	401(k)/403(b) to the limit	Large, tax-sheltered space (FN-02)
5	IRA, backdoor if income-limited	More sheltered space, and the most flexible one
6	Mega-backdoor Roth, if the plan allows	Rarely available and very large where it is
7	Taxable brokerage	No limit, full liquidity, step-up at death (EP-05)

One trap inside rung 5, because the corpus covers conversion ladders and not this. A backdoor Roth is a non-deductible IRA contribution followed by a conversion — and the taxable share of any conversion is prorated across all your traditional, SEP, and SIMPLE IRA balances at 31 December of that year ✅, not just the account you converted. A single legacy rollover IRA can therefore make most of a backdoor conversion taxable. The usual fix is to roll the pre-tax IRA balance into an employer plan first, if the plan accepts incoming rollovers ◻️ — and note the tension with the rest of this page: that same IRA money is conversion-ladder feedstock (ER-02). Decide which job the balance has before doing either.

Two things about this list are load-bearing for this wiki and absent from the usual presentation.

It ranks by tax efficiency alone. Every rung above 7 is more tax-efficient than the one below it and less accessible before 59½. For someone retiring at 65 that trade never binds. For someone retiring at 50 it is the binding constraint, and the list is silent about it.

The HSA’s position is understated. It sits at rung 2 for its tax treatment, but for this audience it is also the account that funds the HC-07 gap Medicare never covers, and it cannot be funded at all once Medicare starts ✅ (HC-04, SS-05). It has a deadline the others do not.

2. The constraint the waterfall is missing

Retiring at 50, spending $70,000, with $1,500,000 saved.

Years to penalty-free access at 59½	9.5
Bridge requirement	$665,000
As a share of the portfolio	44%
Reachable if everything went to the 401(k)	$150,000
Shortfall	$515,000

The money is not missing — it is in the wrong container, and the containers were chosen years earlier by someone optimising a tax bill.

Note what does not fix it: contributing more. A larger 401(k) balance does not become reachable. The fix is composition — taxable holdings, Roth contribution basis, a governmental 457(b), or a conversion ladder started five years before the money is needed (ER-02, ER-05).

And note what does. ER-05 ranks ten access routes, several of which are built during the working years and cannot be built afterwards: Roth contribution basis, a 457(b) left unrolled, taxable lots with high basis, HSA receipts. The accumulation decision and the access decision are the same decision, separated by twenty years.

The practical version of this is one question per dollar: what will this dollar’s access profile be at my retirement age? Ask it alongside the tax question, not instead of it.

3. What each account becomes

The map’s second job is routing — every container has an article that governs draining it:

While working	In retirement it becomes	Governed by
401(k)/403(b)	Rule-of-55 money if you separate at 55+; otherwise locked to 59½	ER-04, TX-03
Governmental 457(b)	Penalty-free at any age after separation ✅ — the best bridge asset there is	ER-05
Traditional IRA	Conversion-ladder feedstock, then RMDs at 73/75	ER-02, TX-04
Roth IRA	Contribution basis first, then seasoned conversions	ER-05
HSA	The dental/vision/hearing and LTC fund; no contributions after Medicare ✅	HC-04, HC-07
Taxable brokerage	The 0% capital-gains bracket, and a step-up at death	TX-01, EP-05
Employer stock	A possible NUA election — decided at separation, once	TR-05

Two rungs are worth more to an early retiree than their tax treatment suggests. A governmental 457(b) has no early-withdrawal penalty at any age after separation ✅ — and rolling it to an IRA destroys that. Taxable brokerage, last on the tax-efficiency list, is first on the accessibility one.

4. Traditional or Roth, decided by the retirement plan

The usual rule is to compare today’s marginal rate against your expected retirement rate. Correct, and it under-specifies the retirement rate badly — the corpus has three pages on what actually sets it (TX-02, SS-03, HC-06).

The version that matters here: an early retiree has low-income years by construction. The gap between the retirement date and Social Security is exactly that (SS-05), and it is when conversions are cheapest.

$30,000 of space, a 24% marginal rate now, conversion at an effective 12% in a gap year ten years later, 5% real return.

Roth now
Tax paid up front	$7,200
Into the Roth	$22,800
After ten years	$37,139

Traditional now, converted in a gap year
Into the traditional	$30,000
Tax saving invested alongside	$7,200
Traditional after ten years	$48,867
Converted at 12%	$43,003 (Roth)
Side account	$11,728 (taxable)
Total	$54,731

The conversion path delivers 47% more, and the mechanism is not clever: a 24% deduction taken now against a 12% tax paid later, by a household that genuinely will have low-income years.

Three caveats, because this is the section most likely to be over-applied. The side account is taxable and carries a drag the model ignores. The conversion window is narrower than the brackets suggest — during ACA years the subsidy cliff caps conversions far below the bracket (SS-05 §3), so the cheap years are typically after Medicare starts and before Social Security does. And if your gap-year rate is not actually lower, the whole argument reverses — which is the case for someone with a large pension or who retires straight into RMD age.

The rule this yields: favour traditional while working if you can name the years in which you will convert cheaply, and favour Roth if you cannot.

5. Two rungs the generic list gets wrong for this audience

Debt, priced against a retirement date rather than a return

The usual rule puts “high-interest debt” at rung 3 and defines high as roughly 6–7% — the rate above which a guaranteed payoff beats an expected market return. That comparison is right and incomplete for someone with a date.

A mortgage retired before the retirement date does something no investment does: it lowers required spending permanently. Cut $2,000 a month of principal and interest and you have cut $24,000 a year of spending — which at a 4% withdrawal rate is the same as adding $600,000 to the portfolio (ER-01), and it also lowers the bridge requirement in §2, the MAGI needed to fund spending (ER-05), and the guardrail cut you would have to absorb in a bad decade (TX-07).

So a 4% mortgage that a generic waterfall says to keep may still be worth retiring — not because the rate beats the market, but because the spending reduction compounds through four other decisions. The rate comparison is a portfolio question; the retirement date makes it a spending question.

The counter-argument is real and belongs here too: paying it off converts liquid assets into home equity, which is the least accessible container of all, and §2 is a page about accessibility. Do not fund a payoff out of the bridge.

Two people, two access ages

The waterfall is written for one person. A couple has two sets of accounts and two different unlock dates, and that changes which container the next dollar should go to.

• The older spouse’s plan may cover the gap the younger one’s cannot. If one is 56 at retirement and the other 49, the Rule of 55 is available on the older spouse’s most recent employer plan and useless on the younger’s ✅ (ER-04).
• A governmental 457(b) held by either spouse serves the household ✅, which can make one person’s job change far more consequential than their salary (ER-05).
• The younger spouse’s 401(k) is the longest-locked asset the household owns and is usually the wrong place for the marginal dollar in the last years before an early retirement.
• Roth contribution basis should generally be built in the younger spouse’s name if the household expects to draw before either reaches 59½, since it is reachable at any age ✅.

And the survivor case is the one nobody models. The household’s Social Security falls to the larger single benefit at the first death and the brackets halve (SS-02, TX-02 §3) — so the account mix should be able to fund a survivor’s spending without forcing large withdrawals into single-filer brackets.

6. Where this map ends

At the retirement date, and the handoff is clean.

• How much to take → ER-01 and TX-07
• From which account, in what order → TX-03
• Before 59½, ranked by the income each route creates → ER-05
• What to do in the five years before the date → TR-01

The two halves genuinely do answer different questions, and conflating them is the most common way a savings plan and a retirement plan end up incompatible. This page asks where does the dollar go in? Those pages ask what does taking it out cost?

Three Scenarios

A. 38, planning to retire at 50, everything in the 401(k). He is doing the standard waterfall correctly and building the §2 shortfall. The fix is not to save more but to redirect — taxable and Roth basis alongside the 401(k), and a ladder started at 45 (ER-02). He has twelve years, which is ample; at 48 it would not be.

B. 55, retiring at 62, maximum contributions, no HSA. Her plan is sound but she has no dedicated healthcare fund and seven years of HDHP eligibility left. The HSA is the one rung with a deadline — nothing can be contributed after Medicare ✅ — and it is the account that funds HC-07’s $57,228 gap. Rung 2 for a reason.

C. High earner with a mega-backdoor Roth available, retiring at 55. He is filling it to the limit, which is right on tax and creates the same accessibility problem in a nicer wrapper — Roth conversions and earnings are not reachable penalty-free just because the account is Roth (ER-05). He needs the ER-04 Rule of 55 to work, which means checking the plan document before separating, not after.

💡 Pro-Tips

1. Ask the access question and the tax question about the same dollar. “What is this worth after tax?” and “when can I reach it?” have different answers.
2. Take the match first, always. No argument on this page outranks a 50–100% immediate return.
3. Fund the HSA on schedule, not opportunistically — it is the only rung with a hard deadline ✅ (HC-04).
4. Compute your bridge requirement in dollars (§2) and check the reachable share against it annually.
5. Favour traditional if you can name your cheap conversion years, Roth if you cannot (§4).
6. Do not roll a governmental 457(b) into an IRA — it is the best bridge asset in the code ✅ (ER-05).
7. Keep taxable holdings deliberately, and set cost-basis method to specific ID at the account level while you remember (TX-01).
8. Start the ladder five years before the first rung is needed, which means starting it while still working if you retire before 55 (ER-02).

⚠️ Common Pitfalls

1. Optimising the current-year tax bill only, and arriving at 50 with $515,000 unreachable (§2).
2. Believing more contributions fix an access problem. Composition fixes it; size does not.
3. Skipping the HSA because the deductible is high — and losing the eligibility permanently at Medicare ✅.
4. Rolling a 457(b) to an IRA “to consolidate” and destroying its no-penalty status ✅.
5. Assuming all Roth money is reachable. Contributions are; conversions and earnings have their own clocks ✅ (ER-05).
6. Choosing Roth reflexively without checking whether cheap conversion years are coming (§4).
7. Choosing traditional reflexively when a pension or early RMDs mean the gap-year rate is not lower.
8. Planning the mega-backdoor without checking the Rule of 55 plan document (ER-04).
9. Treating this page as a decumulation plan. It stops at the retirement date on purpose (§5).
10. Leaving the cost-basis method at FIFO and losing the lever years later (TX-01).

✅ Actionable Checklist

Once, now

• Write down your intended retirement age and the age you reach 59½
• Compute the bridge requirement in dollars (§2)
• Compute what share of your current portfolio is reachable without penalty
• Confirm whether you have a governmental 457(b) and whether you intend to keep it ✅

Each year while working

• Take the full employer match before anything else
• Fund the HSA if eligible, and note the years of eligibility remaining (HC-04)
• Fill sheltered space to the limits in FN-02
• Direct the marginal dollar to close the access gap, not only the tax gap
• Re-check the traditional-versus-Roth call against your expected gap years (§4)

Five years before the date

• Start the conversion ladder if you will retire before 59½ (ER-02)
• Verify the Rule of 55 plan document if you are separating at 55+ (ER-04)
• Move to TR-01, which owns the countdown from here

Sources & further reading (verified August 2026)

• FN-02 for every contribution limit and threshold referenced here — this page deliberately carries none of its own.
• ER-05 for the ten access routes ranked by the income each creates, and for which of them must be built during the working years.
• ER-02 for the five-year seasoning that makes a ladder work, and therefore for when it must be started.
• HC-04 for HSA mechanics and the Medicare cutoff that gives rung 2 its deadline.
• TX-02 and SS-05 for what actually sets the retirement-year tax rate, which is the input the traditional-versus-Roth comparison in §4 needs and the generic version guesses at.
• tools/fn04_worked_examples.py — the bridge arithmetic and the traditional-versus-Roth comparison, computed and checked against this article by CI. Rates, returns, and the portfolio are stated inputs.
• TX-03 and ER-05 own everything after the retirement date; TR-01 owns the five years before it.