The market assumes a broken housing market is a sell signal for risk assets. The tape from September 2026 says otherwise — and the mechanism is not sentiment. It is liquidity plumbing.
Redfin's latest seller-buyer inventory gap printed at 57.9%. That is the widest divergence since the agency began tracking the series in 2013. Translated into raw units: roughly 1.53 million sellers against 970,000 buyers. A 560,000-unit chasm between supply intent and effective demand. The 30-year fixed mortgage rate sits at 6.76%. Existing home sales are plumbing lows not seen since the GFC trough. And yet, in the same month this data crossed my terminal, Bitcoin printed a higher low against a flat equity tape, and the front end of the Treasury curve began pricing a policy pivot that nobody in the sell-side consensus had modeled four weeks earlier.
This is not a coincidence. This is the geometry of a liquidity regime change, transmitted through the most policy-sensitive asset class in the American economy — residential real estate — and landing in a market that still, remarkably, believes crypto is decoupled from macro.
Where code enforcement meets regulatory ambiguity, the same structural logic applies to both a mortgage underwriting model and a DeFi liquidation engine. Both are state machines. Both break the same way. The question is not whether the US housing market is breaking. It is what the breakage force-feeds into the global liquidity function over the next two quarters.
I have spent the last fourteen years watching this specific channel — the housing-to-liquidity transmission — because it is the single most reliable leading indicator for the reflexive phase of every risk cycle since 2008. In 2020, I modeled the correlation between Uniswap V2 liquidity depth and global M2 expansion and predicted a liquidity winter in late 2021. In 2024, I tracked the institutional siphon of the ETF complex as it drained altcoin depth. This cycle is different. This time the trigger is domestic, boring, and enormous.
Let me be precise about what the 57.9% figure actually encodes, because the headline number is being read backwards by most of the market.
Context: The Mortgage Rate Is a Policy Variable Wearing a Market Costume
A 30-year fixed mortgage at 6.76% is not a market outcome. It is a policy transmission vector with a spread attached.
Decompose it. The 10-year Treasury yields roughly 4.35% to 4.55% in the September 2026 window. The primary-secondary mortgage spread — the gap between the risk-free benchmark and the retail mortgage rate — is running about 220 to 240 basis points. That spread is elevated relative to its pre-2022 norm of 130 to 160 basis points, and the residual is a function of MBS duration risk and balance-sheet capacity constraints at the dealer level. But the dominant term is the Treasury yield. And the Treasury yield is the market's discounted expectation of the entire path of Fed policy plus a term premium.
The arithmetic is brutal in its simplicity. A 6.76% mortgage rate against a CPI print running roughly 2.5% to 3.0% implies a real mortgage cost of approximately 3.5% to 4.2%. That is a historically restrictive real rate applied to the single largest liability most American households will ever carry. The housing market is not responding to a cycle. It is responding to a regime.
Here is what the market consistently misreads. Housing is not an asset class in the monetary policy framework. Housing is the transmission channel. Approximately 15% to 18% of US GDP is tied to residential real estate through direct construction, indirect supplier networks, and the induced consumption that flows from housing wealth. That is not a sector. That is a load-bearing wall.
The IMF research the source material leans on establishes the first-order effect: tighter policy compresses risk asset valuations, and housing weakness is the most reliable confirmation that the tightening actually worked. Standard stuff. Textbook. The problem is that the textbook stops at first-order effects.
The second-order effect is where the trade lives. Housing weakness is not merely an output of restrictive policy. It is an input into the policy reaction function. A collapsing housing market does not just signal that the Fed has tightened enough. It creates the political and financial conditions that force the Fed to stop tightening. The source analysis I reviewed correctly identified this reflexive loop but did not resolve it, which is the analytical gap I intend to close.
Before I get to the mechanism, I want to establish the institutional layer, because this is where the 2026 cycle diverges from every prior housing-driven liquidity event, and where most macro analysts are still running 2008 mental models on a 2026 system.
Context Part II: Why the 2008 Playbook Is the Wrong Compiler for This Market
The reflexive instinct when someone says "housing market breaking" is to reach for 2008. This is a category error, and it is costing the consensus real money.
In 2008, the housing break was a solvency crisis. Levered balance sheets — household, bank, and shadow-bank — were the transmission mechanism. Mortgage-backed securities of dubious provenance were the accelerant. The Fed's response was quantitative: it had to backstop the funding markets because the plumbing had seized. The policy transmission was blocked, not merely restrictive.
In 2026, the housing break is a price and affordability crisis, not a solvency crisis. Household balance sheets are, on aggregate, in materially better shape. The mortgage market is dominated by fixed-rate, low-coupon originations issued during the 2020-2021 window when the 30-year was below 3.5%. Roughly 60% to 70% of outstanding mortgages are locked at sub-4% coupons. That means the marginal borrower — the one transacting at 6.76% — is a small fraction of the stock. The stock is insulated. The flow is what is breaking.
This distinction matters enormously for the policy reaction. In 2008, the Fed had to act because the financial system was insolvent. In 2026, the Fed can afford to wait because the financial system is solvent and the pain is concentrated in the flow of new transactions and in the regional economies most exposed to that flow. The Fed's tolerance for housing adjustment is structurally higher than 2008 because the systemic risk vector is different.
But — and this is the crux — higher tolerance does not mean infinite tolerance. And this is where the 57.9% gap becomes a policy forcing function rather than a market data point.
Let me now get into the actual mechanism, because this is the part the consensus is missing and the part that determines the Bitcoin trade.
Core: The Housing-to-Liquidity Transmission, Decomposed
The causal chain runs like this. I will build it as a state machine, because that is how it actually executes.
State 1: Affordability collapse compresses transaction volume.
At 6.76%, the monthly payment on a median-priced home has roughly doubled relative to the 2021 coupon environment. Buyer pool shrinks. The 970,000 active buyers against 1.53 million sellers is the empirical fingerprint of this compression. Transaction volume falls.
State 2: Transaction volume collapse transmits to residential investment.
Residential fixed investment is the GDP line item that responds directly to transaction volume. Housing starts, building permits, and construction employment all lag transaction volume by roughly three to nine months. When sellers outnumber buyers by 57.9%, the signal to builders is unambiguous: do not build. New supply contracts. Construction payrolls are the next domino.
State 3: Construction contraction feeds back into employment and consumption.
Construction is roughly 4% to 5% of total nonfarm payrolls, but it punches above its weight because construction wages are cyclically sensitive and because the multiplier into local services is high. This is where the source material's analysis of the Sun Belt — Nashville, Miami, Houston — becomes critical. These metros ran 30% to 50% price appreciation during the 2020-2022 migration boom. They are now absorbing the reversal. Nashville's seller-buyer ratio is running at 139% over. That is not an adjustment. That is a repricing event.
State 4: Consumption contraction and the OER feedback.
This is the subtle part. US households hold roughly 25% to 30% of their assets in housing equity. 60% of households own homes. When housing prices stall or fall, the wealth effect operates in reverse. Consumption softens with a six-to-twelve month lag. Simultaneously — and this is the channel most analysts underweight — the CPI Owners' Equivalent Rent component lags actual housing prices by twelve to eighteen months. The source data shows seller-market price appreciation at 5.5% year over year versus buyer-market at 1.6%. As buyer markets become a larger share of the national index, the weighted OER trajectory bends down.
State 5: Inflation deceleration forces the policy pivot.
This is the terminus of the chain, and it is where the Bitcoin trade originates. If housing is driving both the demand contraction and the disinflation via OER, the Fed's dual mandate is being satisfied in the direction of easing without an explicit recession being necessary. That is the soft-landing-to-easing configuration. And easing into a system where the real policy rate is still positive is the most historically reliable liquidity injection mechanism there is.
Now — I want to stress-test this chain, because I do not publish a causal argument without a quantitative robustness check. That was the discipline I developed in 2017 auditing ICO token emission schedules, and it has never left me.
Here is the counterfactual. What if housing weakness decelerates inflation too slowly to trigger a pivot? The answer lies in the OER weight. OER is roughly 25% to 30% of core CPI. If the buyer-market share of the housing index expands from roughly 40% of metros to 60% or 70% over the next four quarters, the weighted OER deceleration could shave 40 to 70 basis points off core CPI on a year-over-year basis. That is not a marginal variable. That is a policy pivot encoded in the data.
Let me quantify the regional divergence, because this is where the institutional flow differentiation becomes visible, and where the AI economy re-enters the picture.
The dataset shows a bifurcation that is almost without precedent. On one side, the Sun Belt metros — Nashville, Miami, Houston — display extreme buyer markets with seller-buyer ratios at or above 130%. On the other side, San Francisco displays a seller's market, with price appreciation at the top of the distribution. The source analysis attributes this to the AI wealth effect, and I think that is directionally correct but analytically incomplete.
What San Francisco's housing market is actually providing is an independent, market-priced verification layer for the AI capital expenditure boom. This is the geometry of trust in a permissionless system applied to a permissioned one. If you want to know whether the AI capex cycle is real — whether the hundreds of billions in announced commitments are actually converting into salaries, bonuses, and equity for the workforce deploying them — you do not read the corporate press releases. You watch whether the local housing market can clear at a 6.76% mortgage rate. San Francisco can. That is the market's audit finding. It is the AI truth layer expressed as price.
This is structural decoupling within a national economy, and it has profound implications for how the Fed reads its own data.
The Fed sets one policy rate for a bifurcated economy. San Francisco is overheating on AI wealth. Nashville is deflating on affordability. The weighted national data tells the Fed that housing is weakening. But the composition is split between an autonomous, wealth-driven seller's market and a policy-sensitive, credit-driven buyer's market. The policy rate is hitting the buyer's market and missing the seller's market entirely. This is monetary policy operating with a targeting error, and the error is the AI corridor.
Now, the institutional flow differentiation. In a normal cycle, institutional capital would flee the housing-exposed sectors — homebuilders, building materials, mortgage originators, and the consumer discretionary basket tied to the wealth effect — and rotate into defensives. That rotation is visibly underway. What is less visible, and far more important for this audience, is the rotation occurring at the macro asset allocation level.
I tracked this in 2024 during the ETF approval cycle. The institutional siphon I documented then had a specific signature: capital flowed into the Bitcoin ETF complex not as a crypto bull thesis but as a duration and liquidity hedge, substitutable against long-duration Treasuries. The ETF buyers were not buying a coin. They were buying a macro expression. And that expression was: the Fed will eventually be forced to ease, and I want an asset that is maximally sensitive to that easing without the credit risk of the housing sector that is forcing the easing.
That same configuration is re-emerging now. The 57.9% gap is a signal to the institutional allocator that the forcing function for easing is intensifying. And the instrument of choice for that expression is increasingly the same. Decoding the signal within the noise of volatility requires separating the crypto-native narrative — ETF flows, halving cycles, regulatory clarity — from the macro narrative that is actually driving the institutional bid. In this cycle, the macro narrative is housing.
Let me make the correlation mechanics explicit, because this is where I have the strongest empirical conviction.
The 90-day rolling correlation between Bitcoin and the S&P 500 has historically oscillated between roughly 0.2 and 0.8. The high-correlation regimes coincide with periods where the dominant macro variable is liquidity — that is, when both assets are being priced off the same discount-rate and risk-appetite function. The low-correlation regimes coincide with periods where crypto-specific narratives dominate. Throughout 2024, I warned that the ETF complex would structurally raise the baseline correlation by giving institutional allocators a reason to co-trade the two assets through the same mandate, and that played out.
In the current window, the housing-driven easing expectation is precisely the type of macro variable that pushes the correlation to the high end of the range. If housing is forcing the pivot, and the pivot is a liquidity event, then Bitcoin and equities are trading the same variable. The divergence that retail traders are hoping for — a crypto-specific bull run decoupled from macro weakness — is structurally unlikely in this regime, because the institutional bid is macro-first.
This is the institutional liquidity siphon re-run, but with a different forcing function. In 2024, the siphon was driven by ETF mechanics. In 2026, it is driven by the anticipation of a housing-forced ease. Same structure. Different trigger.
I want to dwell on the reflexive mechanism one more time, because I think it is the most important and least-understood element of this setup.
The market's current default reading is binary and linear. Housing weak equals risk-off. Fed eases equals risk-on. Two separate trades, sequenced. That reading is wrong because the two forces operate on different timescales and the market has not priced the handoff.
Here is the actual nonlinearity. In the near term — call it the first one to two quarters — housing weakness dominates and risk assets face gravitational pull. This is the phase the IMF research describes. It is real, and it is where we are. Institutional allocators de-risk, credit conditions tighten at the margin, and the marginal consumer pulls back.
But the very mechanism that is causing this near-term pressure is the mechanism that triggers the medium-term easing. The housing market's deterioration is not a passive backdrop to Fed policy. It is an active input into the Fed's reaction function, mediated by the OER disinflation channel and the employment channel. So the same data print that is bearish on a one-quarter view is bullish on a three-to-four-quarter view. The market's error is in applying a single sign to a variable that switches sign based on horizon.
This is the silence before the algorithmic deleveraging — the moment where the market is pricing the near-term sign change and has not yet re-priced the medium-term sign change. The reflexive loop is: housing breaks, patience is assumed, the market sells risk, the Fed eventually eases, the market re-prices risk. The danger is that the market front-runs the easing too early and forces the Fed to delay, or that the market refuses to price the easing until the employment data deteriorates too far and the easing arrives into a recession rather than a soft landing.
The critical variable — and this is my central analytical contribution here, the piece the source material left unresolved — is the Fed's tolerance function.
I spent time in March 2022 modeling the tolerance function during the Terra/Luna episode. My conclusion then was that the Fed's tolerance for financial instability is a function of whether the instability is solvency-based or price-based. Solvency-based instability forces immediate intervention because it threatens the payment system. Price-based instability permits patience because it merely reallocates wealth.
The 2026 housing adjustment is price-based. That means the Fed can be patient. But patient is not the same as indifferent. The threshold at which patience converts to action is the threshold at which the price-based adjustment threatens to become solvency-based — that is, when construction employment deterioration and consumer pullback begin to generate a negative feedback loop into the banking sector's credit books.
So my tolerance-function model for this cycle has three inputs. First, the trajectory of construction and construction-adjacent employment. Second, the trajectory of core CPI, with specific weight on the OER component. Third, the trajectory of the term premium embedded in the 10-year Treasury, because that determines the mortgage rate independently of the Fed's policy rate.
The third input is the one the source material completely omits, and it is a serious blind spot. The 30-year mortgage rate is not mechanically determined by the Fed funds rate. It is determined by the 10-year Treasury yield plus the primary-secondary spread. The 10-year yield contains a term premium that reflects the market's assessment of fiscal risk — specifically, the supply of Treasuries required to fund the deficit. If fiscal deficits expand — for any reason, whether stimulus, defense, or debt service — the term premium rises, the 10-year yield rises, and the mortgage rate stays elevated even if the Fed cuts the front end.
This is the trap in the reflexive loop. Housing weakness could force a front-end cut, but if the fiscal trajectory keeps the term premium elevated, the cut does not transmit to the mortgage rate, the housing adjustment does not stabilize, and the Fed is left having eased without achieving the intended effect. That is the scenario in which the reflexivity breaks down, and it is a scenario the consensus is not pricing.
Let me build the probability map, because a causal argument without a probability distribution is journalism, not analysis.
Scenario A — Clean reflexive pivot. Probability roughly 45%. Housing weakness drives OER disinflation, core CPI falls comfortably below 2.5%, construction employment softens enough to justify but not enough to panic, and the Fed cuts into a slowdown. The term premium stays contained. Risk assets, including Bitcoin, rally on the easing. This is the path where the reflexive loop completes as designed.
Scenario B — Fiscal-constrained pivot. Probability roughly 30%. The Fed cuts the front end, but the term premium keeps the 10-year elevated, the mortgage rate doesn't fall, and housing continues to deteriorate. Risk assets initially rally on the cut, then sell off as the transmission failure becomes evident. This is the trap scenario.
Scenario C — Delayed pivot into recession. Probability roughly 20%. Core CPI remains sticky above 3%, the Fed holds too long, construction employment collapses, and the easing arrives too late. Risk assets, including Bitcoin, take a second leg down before the recovery.
Scenario D — No pivot, housing stabilizes on its own. Probability roughly 5%. Affordability improves through price declines rather than rate declines, and the market clears at lower prices without Fed intervention. Structurally possible, but politically and economically unlikely given the employment feedback.
Notice that in three of four scenarios, the eventual direction of risk assets is higher, but the path — and the volatility along the path — differs enormously. This is why the trade is not a directional bet. It is a volatility and timing bet. And it is why I have been telling institutional counterparties that the correct posture is long volatility, not long direction.
Now, the crypto-specific transmission, because this is where my focus differs from the standard macro desk.
Bitcoin's relationship to this setup is best understood through the lens of what it is not. It is not a hedge against housing weakness. The data does not support a negative correlation with housing deterioration. It is a hedge against the monetary response to housing weakness — that is, against the debasement of the currency unit in which the mortgage is denominated. Those are different exposures with different timing.
In the near term, as Scenario A through C unfold, Bitcoin will likely trade with the high-beta duration complex — correlated with the Nasdaq more than with gold, because the marginal institutional holder is a macro allocator expressing a duration view, not a gold-equivalent store-of-value buyer. This is the correlation regime I flagged in 2024, and it remains intact.
The Ordinals and inscription wave is relevant here for a reason most macro analysts miss. The fee revenue generated by inscription activity increased Bitcoin's miner revenue baseline and, by extension, its security budget robustness. A network with a healthier security model is a network that can sustain institutional holding through a drawdown without the existential concern that a weak security budget would introduce. This is a second-order effect, but it matters for the durability of institutional positioning. Without the inscription fee revenue, the security model is a genuine long-term worry, and that worry would cap institutional conviction precisely when the macro setup is most compelling.
So the crypto-specific fundamentals and the macro setup are aligned in a way that is not coincidental. The institutional allocator wants Bitcoin precisely because the macro setup demands a duration-expressing, non-credit asset, and Bitcoin's security model has become robust enough to hold that mandate through a cycle.
I want to bring in the fourth signature element — the AI truth layer — because the 2026 setup is not separable from the AI economy, and I said earlier that the San Francisco housing market is the market's audit of the AI capex cycle. Let me extend that.
I spent three months in 2026 building a behavioral analytics tool to distinguish human from bot transaction activity in an AI-agent payment protocol. The finding that stuck with me was not the bot prevalence figure. It was that the synthetic activity was concentrated in exactly the periods when the project's public metrics were under the most scrutiny. The synthetic flow was a narrative-support mechanism. That is the pattern to watch for in the AI capex verification problem.
Apply it to the AI-corridor housing thesis. The claim that AI wealth is real is currently verified by San Francisco housing prices. But if the AI capex cycle is partly synthetic — if some of the announced commitment is narrative rather than deployed capital — then the San Francisco housing market is either being sustained by a genuine wealth concentration (which is my base case, given the density of the AI workforce there) or it is the last domino standing before the verification arrives. The housing data is the market's audit, but audits can be gamed, and I have seen it done.
The reason this matters for the macro-Bitcoin thesis is that the AI capex cycle is the marginal growth support in the US economy right now. If it is genuine, Scenario A becomes more probable because the growth cushion allows the Fed to ease without a recession. If it is partly synthetic, Scenario C becomes more probable because the growth cushion disappears precisely when housing is contracting. The San Francisco housing data is therefore a leading indicator not just for regional real estate but for the probability distribution across my four scenarios. It is the single most informative non-official data point in the current cycle, and it is almost entirely absent from the consensus macro models.
This is the new insight I would put in front of any institutional desk: the San Francisco versus Sun Belt housing bifurcation is a real-time, market-priced audit of the AI capex cycle, and it is cross-validated by Bitcoin's correlation regime. If the bifurcation narrows — if San Francisco prices begin to soften — the AI capex cycle is being repriced, the growth cushion is disappearing, and the four-scenario probability map shifts sharply toward Scenarios B and C. Watch the bifurcation, not the national aggregates.
Now let me address the positioning implication, because I do not write analysis without operational conclusions.
The dominant error in current positioning is treating this as a housing-sector trade. Homebuilders, building materials, mortgage originators — these are the obvious expressions and they are crowded. The less-crowded and structurally superior expression is the duration and liquidity trade that the housing weakness is forcing. That is long-duration Treasuries, long duration-equity exposure, long volatility, and long Bitcoin as the highest-beta expression of the eventual liquidity injection. The housing sector is the cause. The duration complex is the effect. Trade the effect, not the cause.
The second implication is that the correct timing framework is not price-based but data-based. The source material correctly identifies the housing data as a leading indicator, but leading indicators only have value if you know the lag structure. My working mapping: transaction volume leads residential investment by three to six months; residential investment leads construction employment by six to nine months; construction employment leads the Fed's employment assessment by one to two quarters; and OER disinflation lags housing prices by twelve to eighteen months. The implication is that the housing print we are seeing now is pricing the Fed's reaction in the window twelve to eighteen months forward. That is the investment horizon. Anything shorter is noise. Anything longer is a different trade.
The third implication, and the one I would emphasize most strongly, is the volatility structure. In all four of my scenarios, the path to the eventual re-pricing runs through at least one significant drawdown. The reflexive loop does not execute in a straight line. The correct posture is to be positioned for the eventual easing while hedged against the near-term pressure. That is a convexity trade, not a directional trade. I have watched too many macro funds get the direction right and the path wrong and blow up on the path. The 2020 DeFi liquidity trap taught me that the derivative of the correlation is more important than the correlation itself. That lesson applies here.
Contrarian: The Consensus Has the Sign Backwards on the Wrong Variable
Here is the contrarian angle, stated precisely, because it is easy to state it imprecisely and lose the edge.
The consensus believes the risk is that housing weakness causes a recession, and a recession is bearish for risk assets including Bitcoin. That is a coherent thesis and it is the near-term truth.
The contrarian reframe is that the housing weakness is not primarily a recession risk. It is primarily a policy-forcing function, and the policy it is forcing is easier, not tighter. The consensus is running a 2008 mental model — where housing weakness meant solvency crisis and systemic risk — on a 2026 system that is credit-solvent and flow-constrained. The result is that the consensus systematically underweights the ease-forcing pressure of the housing data and overweight the recession signal.
Where code enforcement meets regulatory ambiguity — and the mortgage market is precisely a code-enforced system operating in regulatory ambiguity — the failure mode is not the collapse of the system. It is the system doing exactly what it was designed to do, which is transmit policy, and the observers misreading the transmission as a malfunction.
The 57.9% gap is not a malfunction. It is the mortgage market doing its job with brutal efficiency. The Fed tightened. The transmission channel transmitted. The output is what mechanical efficiency looks like when the input is restrictive policy. The market is treating the symptom as the disease.
The second contrarian point is about Bitcoin specifically. The consensus crypto-native view is that Bitcoin will decouple from macro weakness and rally on its own narrative — ETF flows, halving cycle, institutional adoption. I have argued consistently since 2024 that this view is structurally wrong in the current regime because the institutional bid is macro-first. The contrarian implication is that Bitcoin's near-term correlation with equities will stay high, and the crypto-native narrative will not save it from a macro drawdown. The decoupling thesis is a retail hope, not an institutional reality. The geometry of trust in a permissionless system does not immunize the price of the asset in that system from the discount rate applied to the system's fiat counterparties.
The third contrarian point, and the one I have the least consensus agreement on, is that the AI corridor is the key variable and the market is not watching it. Everyone is watching the national housing data, the mortgage rate, and the Fed dots. Almost nobody is watching the San Francisco versus Sun Belt spread as an audit of the AI capex cycle. But that spread is the single most informative non-official signal in the entire macro landscape right now, because it is the only market-priced verification of the growth support that determines whether the ease arrives into a soft landing or a recession. The consensus is watching the dependent variable and ignoring the independent one.
Takeaway: Position for the Handoff, Not the Print
The US housing market is not a signal that risk assets should be sold. It is a signal that the forcing function for the next liquidity injection is intensifying, and the market has not yet re-priced the handoff from the near-term recession pressure to the medium-term easing impulse.
The variable to watch is not the mortgage rate in isolation. It is the mortgage rate decomposed into its policy component and its term-premium component, because that decomposition tells you whether the reflexive loop is executing cleanly or being trapped by fiscal constraint. The variable to watch is not the national housing aggregate. It is the San Francisco versus Sun Belt bifurcation, because that bifurcation is the market-priced audit of the AI capex cycle that determines the probability distribution across my four scenarios. And the variable to watch is not Bitcoin's price in isolation. It is Bitcoin's 90-day rolling correlation with the Nasdaq, because that correlation regime tells you whether the institutional bid is macro-first or narrative-first at any given moment.
Three years ago I published a report arguing that crypto was becoming a secondary asset class driven by institutional capital flows. That thesis has fully matured. Bitcoin is a duration expression, a liquidity expression, and increasingly a policy-anticipation expression. The housing market is the policy variable it is currently anticipating. When the market closes the gap between the 57.9% and the eventual Fed response, the repricing will be sharp, and the current correlation regime will have told you which way it was going several weeks before the price did.
The silence before the algorithmic deleveraging is not a warning to exit. It is a window to position. The question is not whether the housing market breaks. It is whether you are positioned for the liquidity that the breakage forces, and on what horizon.