🍑MortgageInGeorgia
Mortgage Rates

The $200 Billion That Bought Georgia About Three Weeks: Rates Dipped Under 6%, and They're Back at 6.65%

In January, the White House directed Fannie Mae and Freddie Mac to buy $200 billion of mortgage bonds. Freddie Mac's 30-year fixed briefly hit 5.98% in late February — the only sub-6% print of the year — and is now 6.65%, higher than the 6.16% it averaged the week the program was announced. In August, both the MBA and Fannie Mae revised their forecasts sharply upward, to 6.7%–6.8%. Here is what the round trip costs a Georgia buyer, and why the policy could only ever move the number for weeks.

By Mortgage in Georgia Editorial··AI-assisted
This article may be AI-assisted and is published as general editorial information. Verify current rates, program rules, and lender requirements with primary sources before acting on it.
Share
Line graph of 30-year mortgage rate chart showing rates dipping under 6% then climbing back at 6.65% through 2026.

On February 26, 2026, Freddie Mac's weekly survey put the average 30-year fixed mortgage at 5.98% — the first time in three and a half years the headline number started with a five. On August 20, 2026, the same survey printed 6.65%. A year earlier it was 6.58%, which means rates are now higher than they were before any of this started, and higher than the 6.16% average the week the policy that caused the dip was announced.

If you are a Georgia buyer who paused your search in January because you were told rates were headed into the fives, that pause has cost you money. This is what happened, why it happened, and what the number on your rate lock actually depends on.

What was announced, precisely

On January 8, 2026, President Trump announced via Truth Social that Fannie Mae and Freddie Mac would purchase $200 billion of agency mortgage-backed securities. FHFA Director Bill Pulte confirmed the directive the same day, saying the two government-sponsored enterprises had "ample liquidity" to do it, and confirmed that purchases would not exceed $200 billion.

That is the whole policy. It is worth stating plainly, because most of the coverage that followed treated it as something closer to a rate cut. It was not a rate cut. It was a commitment by two large buyers to buy more of a specific kind of bond.

The 24 hours it worked

The market reaction was immediate and, for one day, dramatic. Daily rate indices had the 30-year at 6.21% on Thursday, January 8. By Friday morning, January 9, the daily average was 5.99%. Twenty-two basis points, overnight.

Here is the detail that explains everything that came after: the 10-year Treasury yield barely moved over those 18 to 20 hours. What moved was the spread between mortgage rates and the 10-year, which compressed from roughly 1.15% to roughly 0.95%.

Your mortgage rate is, functionally, two things added together. There is the 10-year Treasury yield, which reflects what the bond market thinks about growth and inflation over the next decade. And there is a spread on top of it — the extra yield investors demand for taking on mortgage risk, which widens when investors are nervous and narrows when a large, reliable buyer shows up. The January announcement was a large, reliable buyer showing up. It touched the spread. It did not touch the level.

Spreads are the small half of the equation, and they mean-revert. Levels are the big half, and they don't.

Why $200 billion was never going to be enough

Two numbers put the program in proportion.

The first is the size of the market. Total U.S. household mortgage balances stood at $13.1 trillion at the end of the second quarter of 2026, according to the New York Fed — down $74 billion on the quarter, out of $18.8 trillion in total household debt, with $505 billion of new originations in the quarter alone. Against $13.1 trillion, a $200 billion purchase program is roughly 1.5% of outstanding mortgage debt. It is a real number. It is not a market-moving number.

The second is more damaging, and it comes from the flow side rather than the stock side. Analysts at Loomis Sayles laid out the structural problem: net agency MBS supply in 2026 was expected to run around $400 billion, and the Federal Reserve's own balance-sheet runoff accounted for roughly $200 billion of that. The Fed has been letting mortgage bonds roll off its books — it is a seller, in effect. So GSE buying largely replaces the demand the Fed is withdrawing rather than adding new demand on top of it.

Loomis Sayles put the ceiling on the impact at about 25 basis points. In practice it was smaller: agency MBS spreads tightened roughly 7 basis points on the Bloomberg index at announcement, with the Z-spread moving in about 6 basis points to around 37.

The buying stalled, too

There is a second reason the effect faded, and it has nothing to do with market structure. The GSEs slowed down.

Their combined retained portfolios peaked in April 2026 — Fannie Mae at $175 billion, Freddie Mac at $142 billion — and then drifted lower, to $174 billion and $139 billion by the end of June. For context, a year earlier those portfolios held $85 billion and $97 billion, so the buying was real. But each GSE is capped at $225 billion under its senior preferred stock purchase agreement with Treasury, and BTIG analysts expect Freddie Mac in particular to be "opportunistic" — meaning the $200 billion figure is a ceiling, not a purchase schedule.

The timing lines up uncomfortably well. Rates bottomed at 5.98% on February 26, were back to 6.38% by March 26, and reached 6.53% by late May. The portfolio growth flattened out over the same stretch.

The lever that actually matters is pointing the wrong way

The Federal Reserve has held the fed funds target range at 3.50%–3.75% for all of 2026. Two things about that are commonly misread.

First, fed funds is an overnight bank lending rate. It is not your mortgage rate, and cuts to it do not pass through to 30-year fixed pricing in any reliable way. Your mortgage tracks the 10-year Treasury.

Second — and this is the part most buyers have backwards — the risk on Fed policy right now skews toward tightening, not easing. The July 28–29 FOMC meeting held rates for the fifth consecutive meeting on a 9-3 vote, and all three dissents — Hammack, Kashkari and Logan — favored a hike. The minutes released on August 19 showed officials discussing the need to raise rates if inflation does not cool. Inflation has now run above the 2% target for more than five years.

If you are waiting on the Fed to rescue your purchase price, you are waiting on an institution whose internal debate is currently about going the other direction.

The forecasters just capitulated

This is the most important section of this article, and it is new as of the past week.

If you read a mortgage rate forecast in July, you saw numbers in the mid-6s: the Mortgage Bankers Association at about 6.5%, Fannie Mae at about 6.4%. Those numbers are stale. Both were revised sharply upward in August.

On August 21, the MBA raised its forecast to 6.7% — not just through the fourth quarter of 2026, but through all of 2027 — up from 6.5%. The driver was specifically its Treasury call: the MBA now sees the 10-year at 4.7% by the end of 2027, up from 4.5%. It also cut its refinance outlook.

Fannie Mae's August ESR housing forecast moved further: 6.7% in the third quarter of 2026, 6.8% in the fourth, and roughly 6.8% holding into the first half of 2027. That works out to a 6.5% average for full-year 2026 and 6.7% for 2027. It is the steepest one-month upward revision Fannie Mae has made this year.

Two independent forecasters revising the same direction in the same week is about as clear a signal as this market produces. And practitioners got there first — in a May 21 interview, a division president at Security First Financial said: "We thought we'd probably be solidly in the fives towards the end of this year. I think that's probably not reasonable anymore," pointing to 2027 for meaningful relief. An inflation scare and an oil price spike tied to U.S.–Iran conflict drove the bond volatility behind the reversal.

What this costs a Georgia buyer right now

Take a $400,000 Atlanta-area purchase with 20% down — a $320,000 loan on a 30-year fixed. Principal and interest only; no taxes, insurance or HOA.

| Rate | What it represents | Monthly P&I | | --- | --- | --- | | 5.98% | The Feb. 26, 2026 low | ~$1,914 | | 6.65% | Freddie Mac survey, week ended Aug. 20 | ~$2,054 | | 6.77% | A typical Georgia retail quote, mid-August | ~$2,080 | | 6.80% | Fannie Mae's Q4 2026 forecast | ~$2,086 |

The round trip from the policy low to today is about $140 a month. That is roughly $1,680 a year, and roughly $50,400 over the full life of the loan. That is the price of the three weeks.

And it was three weeks, at most. In Freddie Mac's weekly survey, exactly one week in 2026 printed below 6.00% — the 5.98% reading on February 26. The week before was 6.01% and the week after was 6.00%. Sam Khater, Freddie Mac's chief economist, called it "the first time in three and a half years" the 30-year dropped into the 5% range. The window when the survey sat at or under about 6% ran roughly February 19 to March 5. If you needed to close a Georgia purchase inside that band, you needed a contract already in hand.

The 30 basis points you actually control

Here is the part worth acting on. In the second half of August, Georgia retail quotes from major sources spanned a wide band on essentially the same days: NerdWallet showed 6.47% rate / 6.48% APR for Georgia on August 22; Zillow showed 6.625% on August 18; Rocket Mortgage showed 6.75% on August 12; Bankrate's Georgia average APR ran about 6.746% over the trailing week.

There is no single "Georgia rate." There is a lender-to-lender range of roughly 30 basis points, and on our $320,000 loan that range is worth about $31 a month. Put differently: shopping three or four lenders on the same afternoon recovers about a fifth of everything the policy round trip took away — and it is the only variable in this entire article that you control.

First-time buyers should also check the [Georgia Dream program](https://dca.georgia.gov/affordable-housing/home-ownership/georgia-dream-mortgage-products/georgia-dream-lenders/current) through the Department of Community Affairs. Its pricing is set programmatically and moves independently of the retail market, so it is worth a separate look rather than an assumption that it tracks the national average.

What to do with this

Price the house at today's payment, not a promised one. That is the whole lesson of the past eight months, and it generalizes: policy headlines move the spread, and spreads move for weeks. The 10-year Treasury and inflation set the level, and levels move over years.

Treat any future dip as a refinance option you may or may not get — never as a purchase timeline. If you believe relief is coming, the practical way to express that belief is in loan structure, not in waiting: a low-cost or lender-credit structure keeps the refinance option cheap, because you have not sunk thousands of dollars into buying down a rate you intend to abandon. Paying points to buy down a rate you plan to refinance out of in eighteen months is the most common way buyers pay twice for the same optimism.

Meanwhile, the housing you are shopping for is the housing that exists now, and every month you wait is a month of rent or a month of someone else's amortization.

Rates cited here are national survey averages and lender advertised rates as of August 20–22, 2026. Your individual quote will vary by credit score, down payment, loan amount, county loan limits, occupancy and points. Retail comparison pages update daily; verify current pricing directly with lenders before making a decision.

Sources

Share
Mortgage in Georgia is an editorial site. Verify current rate quotes, underwriting standards, and program eligibility directly with lenders and official program sources before acting on this article.

Related

Related Reading

🍑

Georgia AI

Typically replies instantly