Every attending with a loan balance and a positive monthly surplus eventually faces the same question: send the extra dollars at the debt, or put them in the market? The question generates more heat than almost any other in physician finance, and most of the heat comes from people arguing about the wrong comparison. Done properly, the decision has three layers: a hierarchy of moves that must come first, a clean apples-to-apples comparison of returns after tax, and an honest accounting of the psychological term that the spreadsheet cannot hold. This article works through all three, in order, with 2026 numbers.
Extra principal is a guaranteed return — almost nothing else is
Start with the reframing that settles half the argument. An extra payment against a loan charging 6.8 percent does not "feel like" a return — it is one. Every dollar of principal you retire stops accruing 6.8 percent, with certainty, for the remaining life of the loan. No default risk, no volatility, no sequence risk. If someone offered you a bond guaranteed to yield 6.8 percent after tax, you would buy it without hesitation; your own loan is that bond, available at face value, every month.
The market side of the comparison must be labeled with equal honesty. Broad equity index returns have historically run in the range of 8 to 10 percent nominal over long horizons — but that figure is an expected return, delivered with drawdowns of 30 to 50 percent along the way and no guarantee that the next twenty years repeat the last hundred. Bonds offer known yields but currently compete directly with mid-single-digit loan rates. The honest comparison is never "8 percent beats 6.8 percent." It is "an uncertain 8 percent, taxed, versus a guaranteed 6.8 percent, untaxed" — and framed that way, the loan is a far more serious competitor than the raw numbers suggest.
Key insight
A useful mental test: would you borrow money at your loan's rate today in order to invest it in the market? Choosing to invest surplus cash instead of prepaying the loan is economically the same decision. Physicians who would never borrow at 7 percent to buy index funds, yet happily invest while carrying 7 percent debt, are running the identical trade under two different labels.
The hierarchy comes first: match, high-interest debt, and the PSLF exception
The payoff-versus-invest debate only applies to genuinely discretionary surplus. Three moves outrank both sides of the debate, and one of them is a hard stop.
First, the . Contributing enough to capture a full employer match earns an instant 50 to 100 percent return before any market movement. No loan rate competes with that. The 2026 employee deferral limit is $24,500 under IRS Notice 2025-67; at minimum, contribute to the full match before a single extra dollar goes to the loan.
Second, genuinely high-interest debt. Credit card balances and anything else in double digits are settled questions — no expected market return justifies carrying 22 percent debt. Clear it first.
Third — and this one is a hard stop — the exception. If you are on a Public Service Loan Forgiveness track, extra principal payments are not a conservative choice. They are a pure loss. PSLF forgives whatever balance remains after 120 qualifying payments, so every extra dollar of principal you pay simply reduces the amount forgiven, dollar for dollar. The return on extra payments is not low; it is zero, minus the return those dollars could have earned anywhere else. As of July 2026, the landscape makes this exception more relevant, not less: the Repayment Assistance Plan (RAP) went live July 1, 2026 as a PSLF-qualifying plan with payments of 1 to 10 percent of AGI by income band, IBR remains available with the hardship test removed, SAVE is terminated, and PAYE and ICR are closed to new enrollment. If forgiveness is plausibly on your path, run the PSLF decision framework before this article's question even applies to you — and if you are weighing forgiveness routes that end in a taxable discharge instead, the tax bomb analysis is the companion read.
Important
The most expensive version of "being responsible with debt" is a physician at a nonprofit hospital, sixty payments into a PSLF track, throwing $3,000 extra at the loans every month. Over five years that is $180,000 of payments with a guaranteed return of exactly nothing — the forgiven balance shrinks by the same amount. Before optimizing the payoff-versus-invest split, verify which side of the forgiveness line you are on. It is the single most consequential fork in this entire decision.
The refinanced private loan is where the real decision lives
For physicians who have already left the federal system — refinanced to private loans, typically after ruling out forgiveness — the decision becomes a clean rate comparison. As of July 2026, aggregated marketplace data shows fixed refinance APRs running from roughly 3.6 percent at the top of the credit range to about 10.4 percent, with most physician-profile borrowers landing in the middle single digits depending on term length.
(One permanent caution for anyone still holding federal loans: refinancing to private forfeits forgiveness, income-driven options, and federal protections irreversibly. The timing question is covered in refinancing during residency.)
The rate on your paper, after tax, against labeled expected returns, produces a decision lean rather than a universal answer:
| Your fixed loan rate | The honest comparison | The lean |
|---|---|---|
| Below ~4.5% | Guaranteed 4.5% vs. taxed expected equity returns near 7–8% | Invest lean; pay the minimum and let the term run |
| ~4.5% to ~6.5% | Guaranteed mid-single digits vs. taxed, risky 7–8% | Genuine toss-up; split (below) |
| Above ~6.5% | Guaranteed return matching or beating taxed expected equity returns | Payoff lean — or refinance the rate down first, then re-run |
Note what the table does not say: it does not say the market "wins" below 4.5 percent. It says the expected value favors investing, while the guaranteed outcome still belongs to the loan. Both statements are true, and which one governs is partly the psychology section below.
The tax asymmetry quietly favors the loan at attending income
Here is the layer most comparisons skip. The two returns in this decision are taxed completely differently, and at physician income the asymmetry runs one way.
Interest saved by prepayment is an after-tax return. The student loan interest deduction (IRC §221) caps at $2,500 and phases out, for 2026, between $85,000 and $100,000 of for single filers and $175,000 to $205,000 married filing jointly, per Rev. Proc. 2025-32. Virtually every attending is above the ceiling, so the deduction is zero and the loan's stated rate is the true, untaxed return on prepayment. (Residents, note: below the phase-out you may still claim up to $2,500, which modestly cheapens carrying the debt during training.)
Investment returns, by contrast, are taxed unless sheltered. In a taxable brokerage account — which is where this marginal dollar goes for an attending already funding retirement accounts — dividends are taxed annually and gains eventually, at 15 to 20 percent federal plus any state tax, plus the 3.8 percent net investment income tax at physician income.
Example calculation
Assumptions, stated explicitly: fixed loan rate 6.0%; expected nominal equity return 8.0% in a taxable account; combined dividend and capital-gains drag approximated at 1.0 percentage point per year for a buy-and-hold index investor at attending income (15–20% LTCG plus 3.8% NIIT, state tax excluded); student loan interest deduction unavailable at attending MAGI.
Loan prepayment, after-tax return: 6.0%, guaranteed Taxable investing, after-tax expected return: 8.0% − 1.0% ≈ 7.0%, not guaranteed
The raw gap of 2.0 points shrinks to roughly 1.0 point after tax — the entire remaining premium is compensation for bearing equity risk. Whether one uncertain point is worth 30–50% drawdown exposure is a legitimate question with two defensible answers.
The asymmetry weakens when the alternative dollar goes into a account instead — which is precisely why the hierarchy puts retirement accounts ahead of this entire debate.
The balance-sheet peace argument is legitimate — name it as a preference
Now the term the spreadsheet cannot hold. A large fraction of physicians who prepay aggressively do so because being debt-free changes how they feel — about call schedules, about a bad employer, about risk itself. That argument is legitimate, and the honest way to handle it is to name it precisely: it is a preference, not math, and preferences are allowed to win.
Two observations sharpen it. First, the preference has real option value: a physician with no payments can drop to part time, change jobs, or absorb a partnership buy-in without the debt constraining the choice. Second, revealed risk tolerance matters more than stated risk tolerance — a physician who would sell equities in a 35 percent drawdown will not actually earn the expected return in the table above, and for that physician the guaranteed 6 percent was the better investment all along. Prepaying debt is the one "conservative" allocation nobody abandons at the bottom of a bear market.
What the preference does not license is dishonesty about cost. Choosing the loan at 4 percent over an expected 7 percent, sustained for a decade on large balances, has a six-figure expected price tag. Pay it knowingly if peace is worth that to you — just pay it with open eyes.
The split protocol: automate the argument away
For everyone in the toss-up band — and for couples who disagree, which is most couples — the split protocol ends the debate operationally.
- Complete the hierarchy: full employer match captured, high-interest debt gone, PSLF status verified with certainty.
- Define the monthly surplus after fixed costs and baseline retirement contributions.
- Split it by rate: 50/50 as the default; tilt to 70/30 toward the loan above roughly 6.5 percent, 70/30 toward investing below roughly 4.5 percent.
- Automate both sides on the first of the month — the extra principal payment (flagged to principal, not next month's payment) and the automatic investment into your planned allocation, per the asset allocation framework.
- On payoff, redirect the entire loan-side payment into the investment side automatically, so the freed cash flow never dissolves into lifestyle.
Quick takeaway
The split is not a compromise that satisfies nobody — it is the recognition that both sides of the argument are right about something. The market side is right about expected value; the payoff side is right about certainty and behavior. A physician who runs a 50/50 split for eight years finishes with no debt and a six-figure taxable account, and never once had to win the argument.
Common questions
Should I pay off a 5 percent loan before funding a backdoor Roth IRA?
Fund the Roth first. The comparison is not this year's 5 percent against this year's market return — it is a permanent, irreplaceable slice of tax-free compounding space against one year of guaranteed 5 percent. Annual contribution room expires if unused; the loan will still accept extra payments next year.
Do extra loan payments count as the "bond portion" of my portfolio?
Functionally, prepaying a 6 percent loan resembles buying a 6 percent bond, and some physicians reasonably hold a more equity-heavy portfolio while directing surplus at debt. But the loan "bond" cannot be sold in an emergency and pays its return by disappearing. Treat the parallel as a reason not to hold large bond allocations while carrying higher-rate debt, not as a license to skip real, liquid fixed income forever.
What about my spouse's loans versus mine — which first?
Rate first, tax treatment second, psychology last: highest after-tax rate gets the extra dollars regardless of whose name is on the note, unless one loan carries a forgiveness path — in which case that loan gets minimums only, whoever holds it.
Does the answer change in residency?
Mostly it dissolves. On a resident income the match, a , and PSLF-qualifying minimum payments typically consume the entire surplus before the payoff-versus-invest question arises — and below the MAGI phase-out the interest deduction modestly favors carrying the debt. The real residency decision is the forgiveness fork, not the prepayment split.
What to do next
- Verify your forgiveness status first — pull your loan inventory and qualifying payment count; ten minutes of reading determines whether extra payments are a strategy or a donation.
- List every loan with its current rate, and mark which rates are fixed and which variable.
- Confirm you are capturing the full employer match and clearing any double-digit debt; those outrank everything below them.
- Compute your own after-tax comparison using the calculation above with your actual rate and state taxes.
- Set the split — 50/50 or rate-tilted — and automate both payments for the first of next month.
- Recheck once a year: rates, balances, and the forgiveness landscape have all moved annually since 2020, and a protocol set in 2026 deserves a 2027 review.
The argument between payoff and investing has employed a great deal of internet bandwidth that a two-line automation could have retired; the protocol above works with or without us. This is education, not individualized financial advice.