Good news, we found the bottom.

Good news. We found the bottom.

That was a lie. I wanted to see how it felt to type it. It felt terrible, and I am not going to do that to you.

Here is the truth instead. I have written some version of "relief is coming" in this newsletter nine times since January. I was wrong nine times. Not about the core concept, which I still believe. About the timing, which I guess I keep underestimating.

So this week I will do the opposite.

The 10-year Treasury touched 5.16% Thursday, a 19 year high. The 30-year Treasury hit its highest level since June 2004. For context on how fast this moved: the 10-year started this year around 4.15%. Mortgage rates started at 6.16% and briefly touched 5.98% in February. If you locked in February, you probably didn’t realize the lotto you were hitting.

Today - I am not going to tell you when rates come down. Instead I will tell you:

· exactly what has to happen in order for rates to come down,

· and which specific dates between now and the end of 2027 are the ones that actually matter.

Then you can decide for yourself how likely any of it is.

What Actually Pushed Yields Over 5%

Three things stacked on top of each other and none of them are about housing.

Oil is the first one. Brent is sitting around $105 after the East-West Crude Oil Pipeline shut down on September 11th. Between Hormuz and Bab al-Mandab, we are looking at disruption across a meaningful share of global shipping. Every dollar oil goes up shows up in inflation data about six to eight weeks later, and inflation data is what the bond market prices off of.

The second is the deficit. The government is issuing an enormous amount of debt into a market that is already full. When supply of bonds goes up and demand does not, prices fall and yields rise. That’s the whole story. Treasury Secretary Bessent has tried to calm this down with buyback operations and it has not worked.

The third is the Fed. They hiked on September 16th to 3.75% to 4%, and committee members have signaled at least one more before year end. Warsh went out of his way to emphasize inflation risk.

Here is what makes this different from every other rate scare we have covered this year. In previous selloffs, mortgage spreads absorbed some of the damage. Spreads were the unsung hero of this entire year. And they are continuing to fight the good fight.

About That 2007 Comparison You are going to see this everywhere over the next few weeks so let me get ahead of it.

The last time the 10-year sat firmly above 5% was 2007. Folks have been trying to draw a parallel here since the beginning of the year. Fortune ran a piece in August asking why the bond market is acting like it did before the Great Recession.

I want to be careful here because this comparison is both useful and misleading, and most people are only telling you one half of it.

The useful half: yields at this level do drag on the real economy. Homebuilder stocks are getting destroyed. DHI is down 23% over the past year and LEN is down 43%. Housing starts came in at 1.24

million, down 12.4% month over month. Existing home sales printed 3.98 million annualized in August. JPMorgan's card charge-off rate is running at 3.33%. There is real stress in real places.

The misleading half: what happened in 2008 was not caused by the yield level. It was caused by what was sitting underneath the mortgages. Loans got pooled, the pools got sliced into tranches, the worst tranches got repackaged into new securities, and then new securities got written on top of those. By 2007 institutions were holding paper that was four or five steps removed from a stated income ARM in Stockton, carrying a AAA rating, with no practical way to look through to the actual loans. When borrowers started defaulting, nobody could price anything because nobody could see what they actually owned.

That is what collapsed. Not the interest rate. In fact, after the 10-year crossed 5% in 2007, the S&P gained for four more months.

Today the agency mortgage bond market is one layer deep. A loan goes into a Fannie or Freddie pool, the pool gets guaranteed, an investor buys the pool. The credit risk sits with the GSE and the loans underneath are fully documented. There is no second or third structure stacked on top of it. Lending standards since 2010 are tight to the point of being annoying, and I say that as the person who has to explain to you why we need a letter of explanation for a $400 deposit. Homeowners have record equity. There is no wave of negative amortization loans about to reset.

What Has to Happen for Rates to Come Down

Four things could break this open, and I want to be specific about each one because vague optimism is worthless.

One -The War Ends.

This is the big one but sadly the one no one can put a date on. Plus what’s happened to the cost of oil doesn’t unwind in the supply chain quickly. Barclays has said that if the Strait reopens tomorrow and stays open, we are looking at $88 in 2027.

When that scenario materializes, the Fed is working with different data. If oil drops $25 a barrel and stays there, headline inflation falls hard, the Fed stops hiking, and the term premium investors are demanding on long bonds starts to compress. I think you could see the 10-year back in the mid 4s within two quarters of a durable peace, and mortgage rates following into the low 6s.

That is not optimism. I’m just running this mechanism on oil prices and inflation in reverse from where we started.

Two - Inflation data cooperates without the war ending.

Less likely but possible. If the labor market softens enough to offset the energy shock, core inflation could keep cooling even with oil elevated. That is the scenario where the Fed holds in October rather than hiking again, and the bond market takes that as evidence that we are near the top.

Three - Spreads normalize.

I have written about this all year. Mortgage spreads are still wider than historical norms. Every basis point of spread compression is a basis point off your rate that does not require the Fed to do anything at all. This is the quiet path and it is the one most likely to deliver relief first.

Four - AI productivity shows up in the data. This is a 2027 story, not the 2026 story. Lawrence Yun said it out loud last week. If productivity gains start showing up in the numbers, that is structurally disinflationary and it pulls long rates down regardless of what happens in Washington or the Middle East.

The Two Pivot Points That Actually Matter

Let me get specific about the calendar, because there are only a couple of moments between now and the end of 2027 where the direction can actually change.

October 28th and November 3rd. The Fed's next decision lands October 28th. The midterms are November 3rd. Six days apart. It is going to be very interesting to see what Warsh does with Trump watching six days before Americans head to the polls. I’ll come back to that.

Right now the forecasting consensus is heavily tilted. The generic ballot average sits around D+8. Prediction markets put Democrats at roughly 93% to take the House and around 63% for the Senate. Trump's approval is running in the high 30s. I am not telling you how to feel about any of that, I am telling you what the bond market is looking at.

Here is why it matters for your mortgage rate.

If we get a divided government, the practical effect is legislative gridlock. Gridlock means fewer large spending packages, which means less new debt issuance, which means less supply pressure on the bond market. Bond markets have historically liked divided government for exactly this reason. Again, no personal bias here. Just stating the facts.

If Republicans hold both chambers, the market reads that as continuity on fiscal policy, which in the current environment means continued heavy issuance. That is more pressure on the long end, not less.

There is a second layer here that matters more. Whichever way it breaks, the Fed's independence becomes the story. Trump has been publicly saying rates should be at 1% while his own appointee hikes. If Warsh holds the line at the October meeting six days before a national election, the bond market gets evidence that the Fed is not going to be politically captured. That evidence is worth more to long-term inflation expectations than almost anything else that could happen this year.

If he blinks, the opposite. A Fed that looks politically steerable is a Fed the bond market charges a premium to lend to. That premium just might show up in your rate.

The second half of 2027.

This is where the war, the productivity story, and the Fed's credibility either compound or collapse. Steven Barrow at Standard Bank, who called 5% correctly, has the 10-year going to 5.2% by year end and 5.3% in Q1 2027 before topping out. If he is right, the back half of 2027 is when the relief arrives.

Put those together and my honest read is this. The rest of 2026 is a grind. Rates in the low 7s, maybe testing mid 7s if the war escalates further. The first half of 2027 depends almost entirely on Hormuz. The second half of 2027 is when the pieces could finally line up.

What This Means for Buyers and Sellers

The 7% threshold is a psychological barrier as much as a financial one. Crossing it could slow transactions considerably this fall. If you are listed right now, your buyer pool just got smaller and it is going to keep getting smaller through the holidays. You are competing with builders who have bigger margins at their disposal. Price for the market you are in, not the one you were hoping for.

For buyers, the math is ugly and I am not going to pretend otherwise. But here is what is also true. Existing home sales at 3.98 million means less competition. Homebuilders are desperate and will structure concessions that did not exist a year ago. Housing starts down 12.4% means the supply pipeline is thinning, which is a 2028 problem for buyers but good news for your appreciation opportunity IF you can lock something down now.

You are buying into the worst rate environment in 19 years with the least competition in a decade. Those two facts live in the same market at the same time.

And if the war resolves and rates land in the low 6s by late 2027, every person who bought at 7% refinances and keeps the price they negotiated in a market where nobody else showed up. That is what I’ve started calling an Owners Market.