JTBD RE Pulse Week 5/18/26
Real rates just went negative.
Services did it — not tariffs.
April CPI hit 3.8% — above the Fed's policy rate for the first time in two years. Everyone braced for tariff pass-through. The goods side stayed flat. Services delivered the punch: airline fares +20.7% y/y, shelter, lodging, electricity. Kevin Warsh takes the chair as Powell's term ends today. Trump is in Beijing — the first US presidential visit to China in nearly a decade — wrapping a two-day summit that picks up exactly where last month's petrodollar story left off.
For Massachusetts real estate investors, brokers, and operators who want the macro context behind local deal math.
- April CPI doubled — but tariffs didn't fire. Core jumped +0.4% m/m. The acceleration came from services: airline fares +2.8% m/m (+20.7% y/y), lodging +2.4%, shelter +0.6%. Goods less food and energy: 0.0%. New vehicles fell, apparel decelerated, household furnishings dropped. The categories everyone was watching for tariff pass-through underperformed.
- The Fed's real rate is now negative. CPI at 3.81% y/y blew past the 3.50–3.75% policy band. Negative real rates are stimulative — they don't fix inflation, they feed it. CME FedWatch puts June 17 at 99.8% no change; September 16 prices 18.6% odds of a HIKE versus ~0% odds of a cut. The market isn't waiting on cuts — it's pricing the possibility of further tightening.
- Kevin Warsh confirmed 54–45 on May 13. Takes the chair as Powell's term ends today. Powell stays as Governor through 2028. Warsh inherits a services-led inflation regime on Day One. His first FOMC is June 17.
- Trump in Beijing — picking up where the petrodollar story left off. First US presidential visit to China in nearly a decade. China partially rolled back rare earth export restrictions (gallium, germanium, antimony, graphite) in exchange for cleared Nvidia H200 chip sales. 200 Boeing jets agreed. Three-year "strategic stability" framework — but Al Jazeera headline this morning: "China, US disagree on what they agreed on." The truce is fragile. US debt: $39T (115% of GDP, an all-time high).
- Massachusetts is splitting in two. Supply policy keeps working (ADUs, MBTA zoning, Healey's $25K downpayment expansion). Demand is bleeding (Takeda 247 in Cambridge, Valneva 10–15% global cuts, Boston down to one tower crane in Q1, affordable-housing linkage dollars dried up from $59M to $2.1M). The May 5 rent control deadline passed with no Legislative action — the path to the November ballot is now wide open.
A market priced on feeling — and a summit in Beijing
Sentiment is doing the lifting
Markets are not driven by GDP — they are driven by what Keynes called animal spirits, and what Bob Farrell built a career codifying: fear and greed in sequence. Right now the sequence reads greed. Household balance sheets carry 72% equity exposure — higher than the late-1990s tech peak — at the same time the index itself is the most concentrated it has ever been, with the top ten names representing more than 40% of the S&P 500 (versus 28% at the 2000 peak). The bubble isn't in the technology; the technology is real. The bubble is in the price and the positioning.
Global money supply, by the Kobeissi Letter / StreetStats tracker, just hit a record $121.9 trillion — up $17.1 trillion (+16%) in two years and +$27 trillion (+28%) from the 2022 low. (Other methodologies place the figure between $98T and $123T depending on which central banks are included and how exchange rates are handled; the directional point — that global liquidity is expanding faster than global GDP — holds across all of them.) US M2 sits at a record $22.7 trillion, growing +4.6% y/y. When you compound that with concentration this extreme and household allocations this stretched, what you have is a market whose floor is sentiment — not earnings, not fundamentals, not rates.
The bond market is starting to flinch
While equity sentiment runs hot, the bond market is positioning the other way. TIPS — Treasury Inflation-Protected Securities — pulled in roughly $900 million in April, the largest monthly inflow since December 2021 and only the second positive month in the last five. Global inflation-linked bonds are up +2% year-to-date, outperforming all 24 major fixed-income indexes Bloomberg tracks. European inflation-linked bond funds had their strongest March in three years.
That is what positioning looks like when investors expect the Iran war's inflation to be sticky, regardless of where oil prints next. When equity allocations sit at a record while bond investors quietly rotate into inflation protection, the two markets are telling different stories about the next twelve months. The bond market, historically, is the one that gets it right.
Trump in Beijing — tariffs, rare earths, and a $39 trillion balance sheet
Trump and Xi met in Beijing on May 14–15 — the first US presidential visit to China in nearly a decade, delayed by the US–Israel war on Iran. The summit wrapped this morning with a partial truce: China rolled back export restrictions on four rare earths (gallium, germanium, antimony, graphite) in exchange for the US clearing sales of Nvidia's H200 AI chips to several Chinese tech firms. Trump said Beijing agreed to order 200 Boeing jets. Both leaders described a three-year "strategic stability" framework. But the Al Jazeera headline this morning captures the catch: "China, US disagree on what they agreed on." The truce is real. It is also fragile.
Three numbers tell you why this matters to a Massachusetts mortgage:
- $39 trillion. US national debt is now 115% of GDP — an all-time high. In a single week of early May 2026, Treasury sold $723 billion of securities across 10 auctions ($483B of that in T-bills rolling maturing T-bills), per Wolf Street. Every weak auction widens yields; every wide yield widens the 30-year mortgage. The petrodollar loosening we traced last month is one half of this story. The debt math is the other half.
- 31.6%. The Penn Wharton-modeled average effective US tariff on Chinese imports. That cost appears to be absorbed in retailer and importer margins for now — which is the most plausible explanation for why April CPI showed apparel decelerating and household furnishings falling. Margin absorption isn't directly measurable from CPI; it's an inference. But if it cracks before truce details solidify, the tariff inflation we did not see in April lands in May or June.
- 44 million tonnes. China's known rare-earth reserves — more than half the global total. The US holds 1.9M tonnes (under 5% of China's). China dominates processing even more decisively than mining. The H200-for-rare-earths swap announced this weekend is a one-cycle deal, not a structural fix. Anything that disrupts rare-earth supply disrupts US AI buildout — which is the same buildout currently driving electricity inflation to +6.1% y/y (see §02.2).
And step back from the three bullet points to see the bigger picture. Twenty-five years ago the US sold $729B of goods abroad while China sold $266B — about a third of US exports. Today China sells $3.59T versus the US's $1.9T, and 145 economies trade more with China than with the US (up from 30). China runs a $1T+ trade surplus; the US runs the largest deficit in the world. China holds 23 times the rare-earth reserves the US does and dominates the processing capacity. China invested $290B in green energy last year; the US, $97B. The position the US is negotiating from in Beijing is materially weaker than it was a generation ago — and the petrodollar loosening from Issue 2, the services-led CPI in §02, and the Trump-Xi truce are three time slices of the same trend. The US dollar's role as the global anchor is being renegotiated from a position of contracting leverage, and the negotiation is showing up in your mortgage quote.
Last month we said the physical shock at Hormuz would land in the data months after the headlines calmed — that the gap between "ceasefire announced" and "the data actually moves" was wider than most people assumed. Status check: Iran declared the strait open April 17, but commercial traffic has not normalized. On May 13, only 11 vessels transited — 4 inbound, 7 outbound. The IEA estimates Hormuz crude and fuel flows dropped roughly 4 million barrels per day in March and April. As of May 8, more than 1,550 commercial vessels and 22,500 mariners are still stranded in or around the strait. Brent traded at $107.82/barrel on May 14 — down modestly from the early-May highs, but $41+ above year-ago. That elevated crude is the upstream source of the services-side inflation in §02. The shock didn't end; it migrated into the data, exactly on the timeline we traced.
April was supposed to be the tariff print. It wasn't.
Core doubled. The acceleration came from services.
Last month we said May 12 would be "the first real tariff test." Here is the verdict.
The April CPI report (released May 12) was the most consequential print of the year. Headline rose +0.6% m/m and +3.8% y/y — the worst annual reading since May 2023. But the headline isn't the story. The story is what drove the acceleration.
Core CPI doubled from +0.2% to +0.4% m/m — the largest one-month core print since January 2025. And the tariff watchlist underperformed: new vehicles −0.2%, used vehicles flat, apparel decelerated to +0.6% (from +1.0% in March and +1.3% in February), household furnishings −0.5%, recreation commodities +0.1%. Goods less food and energy: 0.0% — completely flat. The categories everyone was bracing for didn't fire.
Services did the work. Services-less-energy rose +0.5% m/m — the hottest reading since January 2025. Airline fares +2.8% (+20.7% y/y, highest since February 2023). Lodging away from home +2.4%; hotels and motels specifically +2.8%. Shelter +0.6% (largest since January 2024 — with a methodology caveat: this print included a one-off catch-up for the 2025 government shutdown's missed rent surveys). The contribution math: of the headline +0.6%, services-less-energy added 0.301 percentage points, energy added 0.273. Services contributed more than energy — but the BLS press line led with energy.
Real rates are negative — and that's stimulative
The Fed funds rate is 3.50–3.75%. CPI y/y just printed 3.81%. The math is unambiguous: the Fed's "real" rate — policy minus inflation — is now slightly negative for the first time since early 2024. A negative real rate is not restrictive. It is the opposite. It cheapens debt in real terms, encourages borrowing, and pushes investors out of cash into riskier assets. It is stimulative of inflation, not against it.
And the parts of inflation accelerating right now are the parts the Fed cannot wave off as transitory. Electricity +6.1% y/y — up 44% since January 2020, driven directly by AI data-center demand the grid was not built for. Utilities are regulated monopolies; price hikes get rubber-stamped and rarely reverse. Food at home +0.68% m/m, +2.9% y/y — the worst since August 2023. Airline fares +20.7% y/y, and airlines hedge fuel several months ahead — meaning the Iran-war fuel shock is still working through to ticket prices. The Fed has no clean reason to cut, and now a clean reason not to.
Warsh takes the chair as Powell's term ends
The Senate confirmed Kevin Warsh as Fed Chair on May 13 by a vote of 54–45 — mostly party-line, with Senator Fetterman (D-PA) crossing over. Powell's term as Chair ends today, May 15. Powell remains a Governor through 2028, denying the administration an extra board appointment. Warsh inherits a services-led inflation regime, a labor market that's still tight on the surface (April NFP +115K vs. 55K consensus, unemployment 4.3%), and a market that has fully priced out 2026 rate cuts. His first FOMC is June 17. Watch his first public communication after taking the chair — whether he frames April as an energy outlier (dovish) or a services breakout (hawkish) tells you which way the next twelve months bend.
The dashboard
| Indicator | Current Read | Signal |
|---|---|---|
| Headline CPI y/y (Apr) | 3.8% | Worst since May 2023 |
| Core CPI m/m (Apr) | +0.4% | Doubled from +0.2% |
| Services-less-energy m/m (Apr) | +0.5% | Hottest since Jan '25 |
| Goods less food & energy m/m | 0.0% | Tariffs not in the data |
| Airline fares y/y (Apr) | +20.7% | Highest since Feb 2023 |
| Electricity y/y (Apr) | +6.1% | AI-driven, +44% since 2020 |
| 30Y Fixed (Freddie PMMS, May 14) | 6.36% | −1 bp w/w, −45 bps y/y |
| ISM Manufacturing Prices Paid (Apr) | 84.6 | Highest since April 2022 · leads services |
The pipeline is open
The single most striking line in the April release is the contrast between flat goods (0.0%) and accelerating services (+0.5%). In a tariff pass-through scenario you would expect the opposite — goods spike first, services lag. The fact that the divergence is widening, not narrowing, tells you tariffs are being absorbed in retailer margins for now and the inflation pressure is coming from somewhere else: the propagating energy shock from the Iran conflict working through transportation, freight, lodging, and labor.
ISM Manufacturing Prices Paid sits at 84.6 — the highest since April 2022 — and it leads consumer-services prices by one to three months. If that pipeline holds, the May CPI print on June 10 will show another services-led acceleration, this time without the shelter-catch-up methodology assist. That is the cleanest read we'll get on whether April was an event or a trend.
Where the national numbers land in your portfolio
At today's 6.36%, a buyer putting 20% down on the Q1 2026 median single-family home in Massachusetts ($640K) pays roughly $3,190/month in P&I. At 5.0% — late 2024 rates — that same house costs $2,750. The $440/month gap is the price tag on everything traced in sections one through three. Real rates went negative this month, but mortgage rates don't fall mechanically with that — they're priced off the 10-year plus the MBS spread, and both held this week.
Market & demand
Up 3.6% y/y per Warren Group. Prices haven't broken even as rates stayed elevated.
745 in Q1 across 14 firms. May added Takeda (247 in Cambridge) and Valneva (10–15% global headcount).
Down from a peak of 19 in 2018. The construction slowdown is hitting city tax revenue.
The demand side is the weak link. Massachusetts has lost private-sector jobs over the last five years (one of only four states). Biotech — the wage anchor for high-end Cambridge and Seaport rentals — is in its second consecutive quarter of layoffs. Cambridge lab vacancy sits at roughly 25%, a record. And look at where the layoffs are concentrated: per industry reporting, the restructurings hit general and administrative functions and commercial roles — exactly the categories AI is consuming — while cell-and-gene-therapy manufacturing, AI/ML drug discovery, and CMC regulatory affairs remain in 3:1 shortage. The same AI buildout driving electricity inflation in §02.2 is rewriting Cambridge's labor market on a parallel track.
Supply policy & the affordable-housing funding cliff
The supply-side reforms are still doing real work. Healey just expanded the MassHousing downpayment program to offer up to $25,000 at 0% interest for first-time buyers who lock between April 27 and July 31 and earn up to 135% of area median income (income caps from $129,870 in Hampden County to $205,335 in eastern MA). The 0% rate saves a buyer roughly $31,000 over the life of the loan. ADUs remain a permitted use statewide (1,224 approved in the first full year). MBTA Communities Act compliance continues; the law is secure.
But the linkage funding model that paid for affordable construction is collapsing on the back of the broader development slowdown — which is itself the downstream consequence of everything traced in §02. High rates broke developer financing math, fewer projects filed, fewer linkage dollars flow to the affordable-housing fund, fewer affordable units get built, the rental market tightens at the bottom, and political pressure for rent control rises. This is the chain that links April CPI to the November 3 ballot. Boston's affordable-housing fund relied heavily on linkage payments from new commercial development. Projects filed in 2022 are expected to generate $59M+ in linkage payments. Projects filed in 2024: $3.6M. Projects filed in 2025: $2.1M. The funding source for one-third of Boston's $95M affordable housing budget is essentially gone. The Wu administration projects just $40M in additional tax revenue from new development in FY27 — the smallest figure since 2016.
Rent control: deadline passed, the path to November opened
No action by the Legislature. The path to the November 2026 ballot is now wide open.
Supporters had 124,000+ from Nov 2025; they need a small second round to place the question on the November ballot.
If passed: annual rent increases capped at CPI or 5% (whichever is lower). Applies to vacancy turnover — no resetting between tenants. Baseline: Jan 31, 2026 rents. Exempts owner-occupied buildings with four or fewer units and new construction within 10 years of completion. That carve-out is decisive for the New England three-decker math: an owner-occupied triple-decker is exempt; the same building under non-owner-occupied ownership is not. If you are underwriting MA multi-family right now, you are underwriting two parallel scenarios — and the gap between them just got more material.
Your Boston buy pool just got smaller — Compass is pre-marketing 30–40% of deal flow
If you are underwriting MA multi-family, this is a structural change in where your deals are coming from. Compass now controls 30–39.5% of unit sales across Boston and four comparable metros — and an independent analysis found Compass dual-representing 20–42% of those deals, with Boston leading every brokerage in that rate. Internal Compass slides surfaced in the now-dropped Zillow lawsuit showed off-market sales double-end at a rate 72% higher than on-market sales. On May 12, Zillow filed a new federal antitrust suit against Compass and Chicago's MRED MLS over the same pre-marketing tactics.
The investor lens: every property that closes via "private exclusive" or pre-marketing is a property you never saw on MLS, never had a chance to underwrite, and never got to bid on. As that share grows toward 40% in Boston, the public pipeline you've been working from is shrinking. The asymmetry is not theoretical — the deals routed through Compass's in-house buyer network are, by design, the ones with the cleanest comps and the least price discovery. And the macro context matters: off-market thrives in slow, rate-locked markets like the one we're in. Sellers who don't want to publicly cut prices use private exclusives as a hiding place — which is exactly why Compass's share is growing in a 6.36% rate environment, not despite it. If three or four well-priced triple-deckers came on Boston MLS this week, the question worth asking is how many also traded this week that you simply never saw.
Three defensive moves: (1) get on the pre-marketing lists at the top Boston brokerages, including Compass — the only way to see off-market inventory is to be on the inside of it; (2) build direct-to-owner sourcing (mailers, networking, expired listings) so your pipeline doesn't depend on a brokerage whose business model rewards keeping deals in-house; (3) when an MLS property does appear, recognize that it may have already failed a pre-marketing cycle — which can be useful information about pricing.
Four dates to track
Every thread in this issue — services-led inflation, negative real rates, a new Fed Chair, MA's split-screen market, a wide-open rent control ballot path — converges on four dates over the next four-and-a-half months. These are the moments when the data confirms or breaks the trajectory.
The Fed's preferred gauge. PCE weights shelter less than CPI, so it should run cooler than the +0.4% core CPI — expect 0.25–0.35% m/m with y/y ticking to ~2.9–3.1%. The market reads this against the 2% target.
The cleanest test of the services-pipeline thesis. April's shelter print included a one-off catch-up; May won't. If services accelerate again, the pipeline is real and 2026 cuts move further out. If services revert, April was an event.
The decision is not the event — the statement language and dot plot are. Watch how Warsh distinguishes between headline and services-side acceleration, and where the median 2026 dot sits relative to the current band.
The May 5 Legislative deadline passed without action. Petitioners now collect 12,429 additional signatures to place the question. If they do — and signature drives at that volume are routine — every MA multi-family asset gets repriced.
We'll revisit each of these in Issue 4 on June 12.
If you're tracking a deal, a refinance, or a development timeline — these are the four data points that move your math.
What changes for the next deal you write
Issue 2 laid out the framework for the cycle — zero-appreciation underwriting, discount as non-negotiable, growth only if you create it. Those still hold. This section is narrower: what specifically changes in your model because of the data in this issue.
Drop the rate-cut from your 12-month model — the market is pricing hikes, not cuts
If you've been underwriting on the assumption that 6.36% becomes 5.5% by next spring, this issue's data says retire that assumption — and then go further. Negative real rates are stimulative, not restrictive — the Fed cannot cut from here without actively feeding inflation. Services inflation is propagating off ISM Prices Paid at 84.6, and the Trump-Xi truce is fragile. Live CME FedWatch as of today: June 17 — 99.8% no change, 0% cut. July 29 — 95.5% no change, 4.5% hike, 0% cut. September 16 — 80.7% no change, 18.6% HIKE, ~0% cut. The market is not waiting on cuts; it is pricing roughly 1-in-5 odds of further tightening by September. Underwrite at 6.5%+ for new deals and at 6.75%+ for refinances that don't have to happen until Q4 — and stress-test at 7% to see what survives a September hike. If a deal pencils there, it pencils anywhere. If it only pencils at 5.5%, you are not buying real estate — you are buying a rate forecast the market is actively pricing against.
Raise your OpEx growth assumption — the services inflation is in your building
Electricity is up +6.1% y/y and +44% since 2020. That trend doesn't reverse. AI data center demand is the structural driver and utilities are regulated monopolies — the price hikes get rubber-stamped. Insurance premiums (already broken), property taxes (Boston up 13%), and repair-labor costs (services inflation, +0.5% m/m in April) are all moving on the same curve. The 3% OpEx growth assumption most operators were modeling pre-pandemic is stale. Model 5%+ OpEx growth and see whether the deal still pencils. Most don't, at the prices being asked. That is itself a useful answer.
The hedge that does work: capital improvements that generate revenue against the services-inflation curve, not just reduce exposure to it. In-unit washer/dryer is the cleanest MA multifamily example — Boston-area triple-deckers with in-unit laundry command rent premiums of $100–$200/month per unit versus comparable units with basement coin-op or no laundry at all, and the install runs $2,500–$4,000 per stack. Payback inside two years, then permanent rent capture for the life of the building. At 5%+ OpEx growth, you need the rent line to keep pace; this is one of the few capex moves that lets it. Apply the same logic to dishwashers in older units, in-unit HVAC where central isn't an option, and (in heat-paid-by-owner buildings) heat pump conversions where the operating-cost savings flow to you rather than the tenant.
Underwrite MA multi-family in two scenarios — the ballot path is open
The May 5 Legislative deadline passed without action. Petitioners need 12,429 additional signatures to put the rent control question on the November 3, 2026 ballot — a routine sig-drive at that volume. Underwrite as if it passes and as if it fails. The decisive detail in the ballot language is the owner-occupied 4-unit exemption: a triple-decker or four-family that the owner lives in is permanently exempt; the same building under non-owner-occupied ownership is not.
That carve-out, combined with Healey's $25K downpayment expansion (open through July 31) and FHA's 3.5% minimum on owner-occupied 1–4 unit purchases, creates a very specific window: house-hackable triple-deckers in stable MA neighborhoods now have a structurally protected return profile that comparable non-owner-occupied buildings do not. If the ballot passes, that gap widens permanently. If it fails, you still bought a house-hackable triple-decker at 2026 prices. The asymmetry is the trade.
Fix your sourcing — 30–40% of Boston flow is moving off-market
Section 4.4 above is not a brokerage gossip item; it is a sourcing problem. If 30–39.5% of Boston unit sales clear through one brokerage with a structural double-end incentive, the public MLS pipeline you've been working from is materially smaller than it was three years ago. Two moves this month: (1) get on the pre-marketing lists at every brokerage you can — Compass, Coldwell Banker Realty, Gibson Sotheby's, Berkshire Hathaway — even if you have an agent, layer in direct broker relationships; (2) build at least one non-MLS sourcing channel (direct mail to small landlords, networking with estate attorneys, expired-listing follow-up). One out of three Boston deals is invisible to public-MLS-only buyers. Don't be that buyer.
Don't fight the data. Underwrite to it.
The deals that pencil at 6.5% with 5% OpEx growth and two rent-control scenarios are the ones still standing in twelve months.
