Interest-Only vs. Amortizing: Which Wins, and When

On Equal Terms, Interest-Only Wins

We'll start by giving the interest-only argument its strongest case. Take two loans that are identical in every respect — same $40,000,000 amount, same 3.50% rate, same 30-year amortization, no MIP on either — and change only one thing: one pays principal and interest, the other is interest-only for the full term. Hold the asset 10 years, discount each at a 7% cost of capital, and the interest-only loan is worth more.

This is the part the instinct gets right: a payment toward principal is a forced reinvestment at a known rate — an effective 3.56% in this example — so when you pay principal down, your money is only earning that 3.56%. If you can earn a higher rate elsewhere (your theoretical discount rate) than 3.56%, you're better off keeping the cash and putting it to work than retiring debt. That's what interest-only does: it nets you the spread between your discount rate and the effective note rate, here 3.44%. When your money is worth more than the loan costs, paying down principal early is the worse use of it. The two structures break even exactly at the loan rate; above it, interest-only is ahead.

If the comparison were based on equal mortgage terms, every conversation about HUD versus an interest-only agency loan would end with the IO loan winning. But the HUD loan terms are usually differentiated enough from the agency option that the real question becomes whether HUD's better terms make up for the fact that it isn't interest-only.

HUD vs. Agency

Most of our borrowers aren't choosing between a HUD and agency version of the same loan amount. Instead, they're trying to pull the most proceeds the asset will support — because they believe they can deploy that equity at a return above their all-in mortgage rate. Sophisticated borrowers weigh every point at which the deal returns equity: first, cash received at the closing table, through a higher loan amount; second, cash saved each month, through a lower rate or interest-only terms; and third, cash received at disposition or refinance, through a lower payoff balance. All three feed the question of which mortgage option leaves the most equity in hand — a HUD 223(f) or an agency execution.

HUD and agency give different answers at those three cash-flow moments — the settlement table, the monthly debt service, and the payoff at exit — and on an NPV basis the sum of the differences usually runs in HUD's favor. Take a property worth $50,000,000 and size each mortgage program to its maximum LTV. A HUD 223(f) on a cash-out execution tops out at 80% LTV, so the HUD loan comes in at $40,000,000. The agencies run lighter at their highest-leverage tier — particularly with full-term interest-only — so the Fannie Mae loan sizes to 75% LTV, or $37,500,000 in this example. Imagine the HUD interest rate is 3.50% plus the 0.25% annual MIP, and the agency interest rate is 3.95% — 45 basis points higher than HUD — a conservative spread we see at high leverage against HUD's pricing. (HUD's rate is lower because of the Ginnie Mae guarantee to the mortgage-backed-security investor.) HUD amortizes over 35 years; the agency loan is interest-only for the full 10-year term. The example hold period is 10 years, then sell or refinance.

Note that the example rate used here is lower than today's market rate on purpose. Which underwriting limit binds depends on the rate environment: when rates are low, debt service is light relative to value, so the loan is capped by LTV — the case below. When rates are higher, coverage binds first, and the loan is capped by DSCR — the case we turn to next.

On the payment line the agency loan wins clearly. It costs about $50,000 less per month at the start, and across the full 10-year hold the HUD loan runs roughly $5,900,000 more in total debt service. Taken alone, that's a decisive cash-flow advantage for interest-only.

Two things HUD wins clearly, however. First, the higher HUD 80% LTV leverage funded $2,500,000 more at closing — $40M against $37.5M — which is $2,500,000 less equity left in the deal. Second, by the end of the hold period the HUD loan has amortized down to about $33,000,000 while the agency loan still owes its full $37,500,000. At payoff the HUD borrower retires $4,478,000 less debt.

The way to settle it is to value each loan the way the borrower actually experiences it: the proceeds received at closing, the debt service paid each month, and the balance retired at exit. Discount each of those streams back at your own cost of capital, and you get the net present value of the financing — what each loan is worth, today, in a single number. We'll use 7% as the discount rate, a reasonable stand-in for what a dollar of the sponsor's capital earns elsewhere. The project is the same building either way, so the loan with the higher NPV is the one that leaves the borrower wealthier.

Both numbers come out positive because, at a 7% opportunity cost, both loans are cheap money compared to paying all cash — you repay them in dollars you value less than the ones you received at closing. The HUD loan is simply worth more of it, by about $463,000 in present-value terms. The idea that the amortizing loan costs $5,900,000 more in payments is true, and it loses to the $2,500,000 more it advanced at closing and the $4,478,000 less it owes at exit.

Why the HUD Loan Wins

Three things sit in the HUD loan's favor here, and it's worth separating them so the result isn't read as a universal one.

• HUD is cheaper per dollar. The all-in HUD rate of 3.75% — 3.50% plus the 0.25% MIP — sits below the agency rate of 3.95%. Cheaper debt is worth more in present-value terms when your own capital earns more than the loan costs.
• HUD funded more. On the same $50,000,000 asset, the 223(f) reached 80% LTV against the agency's 75%, so it advanced $2,500,000 more. That extra money arrives at closing, when a dollar is worth the most, and it carries the lightest discount of anything in the stream.
• The principal comes back. The HUD loan amortizes, and that paydown reappears at exit as a $4,478,000 lower payoff. The forced reinvestment earns the loan rate, so it nets against the higher payments rather than vanishing.

What's notable about this pairing is how little the discount rate matters to the outcome. The HUD advantage barely moves — it sits in a $400,000 to $615,000 band whether you discount at 4% or 10%, easing somewhat as the rate rises. There is no break-even rate at all: no opportunity cost, high or low, flips the answer.

Near the loans' own cost the NPVs approach zero — at 4%, just above the agency's all-in rate, its NPV is already slightly negative while HUD's is still positive — but the HUD loan stays ahead of the agency at every rate. You don't have to win the argument about what discount rate to use, because every rate lands in the same place.

For the analysts: there is no discount rate at which the two loans tie. The HUD loan's cash flows carry a higher present value at every rate, so the difference between them never reverses and has no internal rate of return to solve for. This is exactly the case that breaks an IRR comparison — the incremental IRR comes back undefined — and where two plain NPVs do the job cleanly.

When the Constraint Is DSCR, Not LTV

The example above is capped by LTV, which is what binds when rates are low. Raise rates and debt service climbs until coverage becomes the tighter limit instead. Every loan is sized to the lower of the two — what the value supports (LTV) and what the income supports (DSCR) — and HUD also has a more aggressive DSCR limit too. A 223(f) sizes to a 1.15 DSCR; the agencies hold to 1.25. Take the same $50,000,000 asset with $2,500,000 of NOI (a 5% cap), now financed at a higher 5.00% for HUD and 5.45% for the agency, where DSCR is what binds.

HUD supports $4,973,000 more loan — roughly 17% more — before pricing even enters, and the advantage carries through to NPV exactly as it did under the LTV cap. At a 7% discount rate the DSCR-constrained HUD loan is worth $3,627,620 against the agency's $2,855,264, an edge of $772,000. That's wider than the LTV case, because the 1.25 coverage requirement squeezes the agency harder than the 75% LTV cap did, and it holds across any discount rate a sponsor would actually use, growing as the rate rises. Whether value or income is the binding constraint, HUD is the larger loan and the better NPV.

The comparison is generous to the agency on one more point worth naming. We've granted it full-term interest-only, but Fannie and Freddie don't hand that out automatically at their highest-leverage tiers. A max-leverage quote frequently carries partial interest-only, with principal payments beginning in year 3, 5, or 7 of a 10-year term. The full-term IO that makes interest-only attractive in the first place isn't a given at the leverage a borrower is usually reaching for, and where it's cut short, the agency loan's position only weakens further.

When Interest-Only Is the Right Call

The equal-terms result from the start of this piece is the thing to hold onto here: interest-only genuinely is the better structure when the terms are actually equal, or when the reason to choose it has nothing to do with leverage. The cases below are where that edge is real.

• The agency matches the proceeds. If you're not pushing maximum leverage — the agency loan funds what you need at the same size — then HUD's sizing advantage never comes into play, and on equal proceeds the lower-payment interest-only structure is worth more.
• Short holds. Over a two- or three-year hold, an amortizing loan recovers very little principal, so the recapture-at-exit benefit is small and the cash-flow drag dominates. The shorter the hold, the more interest-only wins. Prepayment terms enter here too and can change the answer — our analysis of HUD prepayment versus yield maintenance walks through how.

What the worked example shows is not quite as simple as "amortizing is better." It's that the amortization on a HUD loan rarely costs what it appears to, and when HUD also sizes higher and prices lower — the max-leverage case — the larger amortizing loan can be worth more outright. Reach for interest-only when the cash it frees has a better home than the loan rate, not on the payment line alone.

What the Comparison Leaves Out

Valuing each loan against the same property means the building, its income, and its sale price are identical on both sides of the comparison — so they cancel out, leaving only the loan cash flows. That clean cancellation is what isolates the financing decision, and it's also where the comparison stops. There are three things it deliberately doesn't price.

• A single rate prices return, not refinance risk. The interest-only loan reaches maturity owing nearly its full balance, to be refinanced into whatever cap rate and interest rate environment exists then. The amortizing loan has been deleveraging into that same maturity the entire hold. Discounting at one rate folds this into a return figure rather than showing it. The exposure sits with interest-only, and it's a reason to favor amortizing that the NPV gap understates rather than captures.
• It answers the loan penciling question, not the deal penciling question. The comparison tells you which loan is the better financing, holding the property fixed. It doesn't tell you whether the deal itself works — whether the extra leverage actually lifts your equity return given the property's own yield, what cash-on-cash each loan throws off, or how the sale clears. Those need a full levered-equity model with real NOI, a cap rate, and a sale price — the analysis that tells you whether to do the deal, not just which loan is the better financing.