On August 14, 2026, I published an analysis titled "Decoding a Signal the Market Has Ignored: Mortgage Rates May Have Peaked." The core call was this: Treasury Secretary Scott Bessent's intervention in the yen exchange rate showed that the Treasury had made pushing down long-term rates a priority, and its next step would very likely be to act directly on long-dated U.S. Treasuries—bringing down the 10-year Treasury yield and mortgage rates.
Just five days later, on August 19, the U.S. Treasury announced buybacks targeting 10-year and 30-year Treasuries, raising the size to $4 billion per operation. On the news, yields on the 10-year and 30-year Treasuries and mortgage rates fell sharply. The move mirrors what the Treasury did under Janet Yellen in 2023.
A call written in advance and the event that followed, tied together by a single thread
This is not just a victory lap. I want to take the call apart: what mechanism it relied on, how the chain of evidence fits together, what historical precedent exists, whether my other public forecasts over the past few years hold up, and—most importantly—how readers can use the same method to judge the next policy signal on their own. At the end, I list three questions I plan to test publicly next.
A Verifiable Forecast: Reconstructing the Timeline
A forecast is only worth something if it was written down publicly before the event, with a clear direction and time window.
The problem with much market commentary is that it is explained after the fact: when prices rise, the good news was "priced in"; when they fall, the bad news was "fully absorbed." Statements like that are always right and always useless. There is only one standard for judging whether an analyst truly understands a market: ex ante, public, and falsifiable.
The timeline here is unambiguous:
- August 14 (Friday): I publicly released a video arguing the Treasury would act on long-dated Treasuries to bring down mortgage rates and the 10-year yield.
- August 19 (Wednesday): The U.S. Treasury announced a buyback program for 10-year and 30-year Treasuries, with the size increased to $4 billion per operation.
- Market reaction: Yields on the 10-year and 30-year Treasuries and mortgage rates dropped sharply.
Five days elapsed between the call and the event. I should be candid: the speed itself involved some luck. What I called was the direction and the tool, not the exact date. But the direction (pushing down long-term rates), the tool (acting directly on long-dated Treasuries), and the assets affected (the 10-year Treasury and mortgage rates) all matched what actually happened. That is what this review is really meant to show.
The call was about direction and tools, not the exact date
If you haven't read the original analysis, start with Bessent Steps In to Rescue the Yen—Why It Means U.S. Mortgage Rates May Have Peaked. This article is its sequel.
The Mechanism: Why a Yen Intervention Pointed to a Long-Bond Buyback
A Treasury Secretary willing to step in on a currency is not really concerned with the currency—he is concerned with the long-term rates behind it.
To understand the call, you first have to understand the policymaker's constraints.
Layer One: Long-Term Rates Are a Core Treasury Concern
What the Federal Reserve directly controls is the short-term policy rate, while mortgage rates mainly track the 10-year Treasury yield. In other words, a Fed rate cut does not automatically lower mortgage rates—a point I argued in detail in Why Fed Rate Cuts Won't Move Home Prices (Full Version). When short rates fall but long rates stay stubbornly high, the institution that has both the tools and the motive to manage the long end is the one that issues the debt: the Treasury.
Mechanism: The Treasury issues U.S. government debt. The higher long-term yields go, the higher the Treasury's future borrowing costs; and the higher mortgage rates go, the more frozen the housing market becomes and the bigger the drag on the economy. The Treasury therefore has a built-in incentive to push long-term rates down.
As the issuer of debt, the Treasury has a built-in motive to push long-term rates down
Layer Two: The Yen Intervention Was a Signal, Not an Isolated Event
Japan is a major holder of U.S. Treasuries. When the yen is under pressure, Japan may sell U.S. Treasuries to raise dollars and stabilize its currency, which puts selling pressure on long-dated U.S. bonds and pushes yields higher. By intervening on the yen, Bessent was essentially cutting off one transmission channel that drives long-term Treasury yields up.
What this means: A Treasury Secretary willing to spend political capital on another country's currency is signaling very low tolerance for high long-term rates. Following that logic, the most direct next tool is to act in the U.S. Treasury market itself.
Layer Three: Buybacks Are the Most Direct Long-End Tool
A Treasury buyback means the Treasury uses cash to repurchase bonds it has already issued. Buying back long-dated securities such as the 10-year and 30-year adds demand at the long end and reduces the supply in circulation. Less supply and more demand mean higher prices and lower yields. A size of $4 billion per operation shows this is not a token gesture but a clear policy signal.
Buying back long bonds shrinks supply in circulation, and yields fall with it
From Yellen to Bessent: The Same Policy Toolkit
The people changed, but the constraints did not—and neither did the toolkit.
In my August 14 video, I specifically pointed to what Janet Yellen did in 2023, because it was the key reference point for predicting what the Treasury would do.
Argument: The Treasury has long faced the same structural problem—when long-term rates are too high, they raise government financing costs and suppress the housing market at the same time.
Mechanism: When a problem recurs, institutions tend to reach for tools that have already been tested. The path Yellen took in 2023 to address pressure on long-term rates gave her successor a ready-made template.
Evidence: Bessent's buyback targeting 10-year and 30-year Treasuries follows the same path as Yellen's 2023 operations.
What this means: Studying policy means looking past the person to the institution and its tools. Many people try to guess a Treasury Secretary's behavior from political labels—"hawk" or "dove." What actually has predictive power is asking: What constraints does he face? What tools did his predecessor use under the same constraints? How well did they work?
The decision-maker changed, but the constraints and the toolkit did not
This matters especially for real estate readers. Mortgage rates are not a number set by the Fed alone; they are the product of the Fed, the Treasury, foreign holders, and market expectations acting together. Understanding the Treasury's toolkit gives you one more window for spotting a rate turning point early. For more on the new Fed leadership and long-term rates, see The Fed Turns Hawkish—and That Might Save the Housing Market? Mortgage Rates and Cap Rates in the Warsh Era.
A Track Record: Four Calls You Can Check
One hit can be luck; a string of hits suggests the method works.
This is not the first time I have put a call on the record before the event. The following were all published publicly on the channel and can be checked:
Call One: A 2026 Seattle Price Decline and Mass Tech Layoffs
In 2024, I predicted that Seattle home prices would fall sharply in 2026, alongside large-scale layoffs in the tech industry. At the time, this ran against the prevailing mood—many still believed home prices in tech hubs could only go up. The quantitative analysis is in How Much Will AI and Tech Layoffs Hit Seattle and Bay Area Home Prices?.
Call Two: Market Divergence in the Seattle Region in 2026
In 2025, I predicted that the Seattle region would diverge in 2026: Seattle city would strengthen, while Bellevue and Kirkland on the Eastside would weaken. What made this counterintuitive was that the Eastside had long been considered the most stable part of Greater Seattle. The full reasoning is in Seattle Home Prices in 2026: 5 Key Predictions.
Regional divergence: the urban core strengthens while the Eastside weakens
Call Three: The Fed's Rate-Hike Path
I correctly predicted the Federal Reserve's rate-hike path on three separate occasions.
Call Four: The Impact of the ROAD to Housing Act
A year ago, I walked through the impact of the ROAD to Housing Act on the U.S. real estate market, well before the bill became a focus of market discussion. The analysis is in Trump's New Housing Bill: How the ROAD to Housing Act Will Rewrite the U.S. Housing Supply Chain.
What this means: These calls span different kinds of questions—regional prices, regional divergence, monetary policy, legislation—but they share one feature: each was made before a consensus formed. If an analytical framework works in only one domain, it may just be experience. If it works across several, it has likely captured something more fundamental.
Why the Forecasts Hold Up: Method, Not Luck
Good judgment comes from enough firsthand information, enough time to think, rigorous training, and genuine interest.
I attribute it to four things, each of which corresponds to a habit others can replicate.
1. Accumulating Firsthand Information Over Years
For the past five years, I have spent three hours every day reading English-language newspapers in the original. For example, I have read every single article by Nick Timiraos, the reporter widely known as the "Fed whisperer."
Mechanism: Policy signals usually show up first in original reporting and official statements. Once they are retold secondhand, much of the detail and tone is lost. The link between the yen intervention and long-term rates was exactly the kind of clue you only notice by reading the originals closely.
2. Writing Every Script Myself, and Giving Thinking Time to Work
I write every script myself, and each one takes at least eight hours. Writing is research: checking data, revising the line of thought, and testing whether the chain of logic holds together.
Mechanism: For a judgment to stand up to scrutiny, it has to go through a cycle of writing it down, arguing against it, and revising. Content drafted by assistants or AI can be fluent, but fluent is not the same as coherent. Only by working through the argument from scratch yourself can you find the broken link in the chain.
Writing, rebutting, and revising yourself is how you find the broken link
3. Systematic Analytical Training
I spent four years as an undergraduate in the mathematics department at Zhejiang University, then two years earning an MBA at Arizona State University, and then worked at Microsoft for 11 years. Mathematics trained me to reason rigorously; the MBA taught me how businesses and markets work; and at Microsoft—where I worked in sales, in marketing, and as a datacenter product manager—I learned both to understand the technology and to explain complex problems clearly.
Mechanism: Forecasting is fundamentally a modeling problem—identifying the variables, judging cause and effect, and estimating timing. The value of systematic training is not that it supplies answers, but that it supplies a way of breaking problems apart.
4. Genuine Interest
In 2023, while Washington State was debating its ADU bill, I went to the state capitol about twice a week on average to attend hearings, organize discussions, and build support. The drive from my home to Olympia is three hours each way, and all of it was volunteer work. In 2025, I taught a zoning course in the University of Washington's real estate program, sharing hands-on experience with zoning and development—again without taking a single dollar.
Mechanism: Interest determines how much time a person is willing to invest where there is no payoff. Hearings, bill text, and zoning codes are dull but critical primary sources. People who have actually taken part in legislation and development have a more concrete intuition for how policy gets implemented.
Counterarguments and Responses
The best attitude toward forecasting is to invite scrutiny, not avoid it.
Counterargument 1: "This is survivorship bias—you wouldn't highlight the calls you got wrong."
That is a fair challenge. Anyone who publicly reviews their own forecasts can be suspected of showing only the hits.
Response: My forecasts are all published as videos with clear release dates, and anyone can go back to the channel and review the full record, including the calls that did not pan out. I also openly acknowledge that I cannot be right every time. That is exactly why I insist on putting calls on the record in advance—it means my mistakes have nowhere to hide either. The right way to test a forecaster's accuracy is not to listen to their own summary but to look at the complete public record.
Counterargument 2: "Treasury buybacks are routine; they aren't necessarily caused by the yen intervention."
This is also a professional objection. Buybacks are an existing Treasury tool and are not necessarily triggered directly by any single external event.
Response: My call was never "the yen intervention caused the buyback." It was "the yen intervention revealed the Treasury's low tolerance for high long-term rates, so the Treasury would use long-end tools more aggressively next." That is an inference about policy intent, not a claim of single-factor causation. The buyback covered the two longest maturities, the 10-year and 30-year; the size was raised to $4 billion per operation; and long-term yields and mortgage rates then fell sharply. Those facts are consistent with a Treasury actively pushing long-term rates down. Even if buybacks are an existing tool, when it is used, how forcefully, and at which maturities is precisely where policy intent shows.
The point is not single-factor causation but the policy intent behind the move
Counterargument 3: "Rates fell once. That doesn't mean they've peaked."
Response: Agreed. One buyback and one drop in rates cannot prove that rates have peaked; they only show that the Treasury's policy direction is consistent with my call. Whether rates have truly peaked will take longer to observe. That is why I say every word I've said here is on the record, and we can come back to check in three months, six months, or a year.
A Framework You Can Apply to Policy Signals Yourself
Forecasting is not guessing the outcome; it is first working out who has the motive, who has the tools, and what has been used before.
The method behind this call can be summarized in five steps that readers can apply to any policy signal:
- Identify the decision-maker who actually holds the tools. Who influences mortgage rates? Not just the Fed, but also the Treasury and foreign holders. Ask first: who can act directly on the variable you care about?
- Pin down the decision-maker's constraints. What does he care about most? Which way do his financing costs, political pressures, and market-stability concerns each point?
- Treat unusual moves as signals. When a decision-maker does something that seems outside the core of his job—like a Treasury Secretary managing the yen—ask which of his core goals it serves.
- Look for historical precedent. Under the same constraints, what tool did the predecessor use? In this case, Yellen's 2023 precedent was the key to inferring Bessent's next move.
- Write down a falsifiable call. State the direction, the tool, the affected assets, and a rough time window; publish it; and come back to check when the time is up.
For homebuyers and investors, the practical upshot of this framework is: don't watch only the Fed's rate meetings. Turning points in mortgage rates can often be read in advance from the Treasury's issuance and buyback plans and from currency pressure on major foreign holders.
Don't watch only the Fed; rate turning points show up in other windows too
Three Questions to Test Publicly Next
Find a trend before it becomes consensus, explain what you see and why you judge it that way, and say what is likely to happen next.
I don't want to simply relay the news. The long-term value this channel and this site aim to provide is to make calls before consensus forms and to lay the reasoning out for everyone to see. Next, I plan to dig into three questions:
First, AI may be driving the biggest economic bubble in U.S. history. If it bursts, will real estate collapse—or hold up even better? The key is where capital flows after the bubble bursts and how interest rates respond. Some of my earlier thinking is in AI Is Putting White-Collar Workers Out of Work, but Blue-Collar Workers Will Get Richer.
Second, as tech companies hire aggressively for Forward Deployed Engineer roles, could these workers become the next wave of buyers propping up U.S. housing? The impact on Seattle and Bay Area housing depends on where these jobs are located and how they are paid.
Third, have Seattle home prices hit bottom? Is it time to buy the dip? This is the question many readers care about most—and the one that demands the most careful answer.
Of course, I can't be right every time. But that is exactly what makes public forecasting valuable: every call written down today is on the record, and in three months, six months, or a year, we can come back and check each one.
