The Latticework Letter — Issue #014

Last week the Fed hiked and the 10-Year crossed 5%. This week the bond market decided that was an opening bid. After a quiet Monday around 4.95–4.97% — the week’s one moment of relief — Tuesday delivered the real move: the 10-Year jumped 17 basis points in a single session to close at 5.12%, its highest close since 2007, touched 5.15% midweek, and hovers near 5.20% Friday morning. The 30-Year reached 5.44%, a level last seen in 2004. The proximate causes were almost insultingly ordinary: S&P Global’s September flash survey showed business activity accelerating at its fastest pace since July 2021, oil stayed loud (more below), and futures walked the odds of an October hike from 55% a week ago to roughly 75%. Sixteen of eighteen dots said another hike was coming; the market has stopped arguing.

Oil, meanwhile, ran a full round trip inside five days. Saudi Arabia began restoring the damaged East-West pipeline early in the week and Brent slid from last week’s $107.50 to briefly under $100 — five straight down days — before Iranian officials warned the conflict could spread “to the Indian Ocean” and the price snapped back to roughly $105 by Thursday. Here is the part your gas-station tenant already knows: the pump never took the round trip. AAA’s national average hit $4.47 Wednesday, up from $4.37 a week ago and $4.10 a month ago — the retail price follows crude up the staircase and down the escalator, and the customer walking into the store has eleven fewer cents per gallon, per fill-up, for the 55%-margin food that pays the rent.

Now put the two markets that price your asset side by side, because this week they finished diverging. The public market has repriced: the average REIT is down about 8% over three months, Realty Income down 14%, Agree Realty down 16%, both scraping 52-week lows. The private market has not: Boulder’s Q2 average asking cap rate moved two basis points, to 6.82%. Those Q2 asking caps were set when the 10-Year traded between 4.20% and 4.70% and settled near 4.40%. Against Tuesday’s 5.12% close, the average net-lease deal’s spread over the risk-free rate has compressed from roughly 240 basis points to about 170 — with cap rates standing still. Public shareholders have already eaten the repricing. Private sellers are still quoting a world that ended in June. The lag we keep writing about — two to four quarters, on the way up in 2022–23 — is no longer a forecast; it is the single most important number on your term sheet, and it is this issue’s recurring theme. The sector below is what the people who move first are doing about it.

The most instructive market in net lease right now is the one that isn’t transacting. Newmark’s national single-tenant report, published Monday, contains a pair of numbers that should not be able to coexist: Q2 net-lease volume hit $13.8 billion, the largest second quarter since 2022, with single-tenant volume up 24.1% year-over-year — and yet sale-leasebacks collapsed to 6.7% of net-lease transactions, against a 13.1% average since 2019. Corporate owner-user divestment in the first half was $1.3 billion, down 67.6% from a year earlier. Newmark’s Ching-Ting Wang supplied the diagnosis: fewer deals inked, not smaller deals. The market is growing while its historic feedstock — the operating company monetizing its own real estate — has gone on strike. Translate the strike into plain English: middle-market CFOs looked at the cap rate a buyer would pay in a 5% Treasury world, compared it to what their building was “worth” on the June appraisal, and hung up the phone. The bid-ask in net lease is now a macro statistic.

So who is transacting? Follow the capital, because it is behaving with remarkable consistency. On September 8, CBRE Investment Management paid Cerberus $1.6 billion for Tenet Equity — a platform built almost entirely on middle-market sale-leasebacks. Two days later, J.P. Morgan closed its first dedicated U.S. net-lease fund at $1.1 billion, more than double its $500 million target, aimed at supply-chain-critical industrial and outdoor storage sale-leasebacks, with LACERA alone writing $200 million and more than half the money coming from investors new to the firm. The institutions cannot buy the deals — there aren’t enough — so they are buying the machines that make the deals, and pre-funding the vintage they expect when the strike breaks. Newmark’s sector split tells you where they think it breaks: industrial is now 61.5% of net-lease volume (from a third), office has fallen to 13.2%, and retail sits stable at about a quarter.

And then there is the specialist who never stopped. On Tuesday, Getty Realty closed a $260.9 million sale-leaseback with Refuel Operating Company — 41 large-format convenience stores across South Carolina, North Carolina, Texas, and Mississippi, averaging 5,000 square feet on 2.5 acres, structured as four unitary net leases (one per state), 20-year initial terms, bumps every five years. Refuel — a First Reserve portfolio company running about 250 stores — becomes Getty’s third-largest tenant at 7.7% of rent. Study the machine, because this is what disciplined capital looks like mid-repricing: Getty has invested $455.2 million year-to-date at a 7.1% average initial yield while selling stabilized assets at a 5.7% cap — selling into yesterday’s pricing, buying at today’s, pocketing the 140-basis-point difference — funded leverage-neutral with forward equity, a new term loan, and those dispositions, with a $125 million pipeline at 7.8% behind it. And note what the unitary structure buys: one lease per state, all-or-nothing, so no tenant can hand back the three worst boxes and keep the rest. In a week when the average deal’s Treasury spread compressed to 170 basis points, Getty printed 200-plus over — with five-year bumps — because it writes checks when other buyers can’t and sellers with a use for capital can’t wait for the appraisal to catch up. The lesson for your own book runs in both directions. If you are a buyer: the only sellers at today’s clearing price are operators who need the money for something better — which is exactly the adverse-selection moment when credit work matters most, because desperation and opportunity wear the same cap rate. If you are a seller: the 6.7% number is your competition telling you they won’t sell here. The strike ends one of two ways — the 10-Year comes back, or the appraisals do. Sixteen Fed dots have a view on which.

America’s Car-Mart has now been extended in increments five times — September 7 to the 11th, to the 18th, to the 24th, and, in an 8-K filed Thursday, to October 1. The market graded the fifth extension in real time: the filing again names bankruptcy as a possible outcome, and the stock fell nearly 19% Friday morning to about $1.09. The rule from last week stands — a lender granting one-week extensions is a lender close enough to a signature to keep paying for the option — but the backdrop keeps darkening. The 10-Q filed September 9 disclosed revenue down 57.3% to $145.8 million, a $69 million quarterly net loss, and substantial doubt about the company’s ability to continue as a going concern; the plaintiff-bar pile-on added the Rosen Law Firm this week to the Cruz and Smith investigations; and the annual meeting came and went Wednesday without incident. For landlords of the 90-odd operating dealerships, the posture is unchanged — the store-by-store triage happens whether the signature reads “recapitalization” or “restructuring.” Next cliff: Wednesday, October 1.

Darden reported fiscal Q1 Thursday before the open, and the read on the middle-income diner came back mixed leaning soft. Adjusted EPS of $2.05 landed in line; total sales rose 5.1% to $3.2 billion; and the company reaffirmed — did not raise — full-year guidance of $11.10–$11.35. The tell was the brand split: Olive Garden comps grew just 1.1% against expectations of 1.3–2.0%, while LongHorn ran +6.2%. The customer paying $4.47 at the pump is still going out for steak occasions and skimping on breadstick occasions — trade-down within casual dining, not out of it. The stock fell 4–6% on the print. For Four Corners holders — roughly 41% Darden by rent — the credit is fine; the reminder is that “casual dining’s treasury bond” now yields growth from exactly one of its two big brands.

Salad and Go went before the judge Monday for approval of the $143.2 million 7 Brew sale. As of Thursday we could not find a published order on the outcome — no objection blowup made the trade press, which in bankruptcy usually means the quiet kind of news — but until the docket confirms, treat closing timing as open. If your site is among the 73, your cure and assignment terms were fixed by this hearing; the next thing to watch is 7 Brew’s conversion schedule, because 73 boxes converting to drive-thru coffee is the supply thesis from #013 arriving on an accelerated timetable.

On the calendar: OZ 2.0 state nominations close Monday, September 28 — as of this week only 13 jurisdictions covering about 1,329 of an anticipated ~6,500 tracts (roughly 20%) have filed or announced, with Texas (605 tracts) and North Carolina (202) the heavyweights and California, Louisiana, New Jersey, and Tennessee already taking the 30-day extension to October 28. Realty Income’s €528 million joint venture with KKR — KKR at 49%, seeded with 54 European properties — is expected to close around September 30, the clearest sign yet that O intends to grow with other people’s balance sheets. Boulder’s market-wide Q3 cap-rate report is due late September or early October, and it will be the first broad private-market print taken after the hike. Getty reports Q3 on October 21. Next FOMC: October 28 — futures say three-in-four.

That is the spread between the average net-lease asking cap rate — Boulder’s Q2 print of 6.82% — and Tuesday’s 5.12% close on the 10-Year Treasury. When those Q2 cap rates were being set, the 10-Year averaged near 4.40%, and the same portfolio offered roughly 240 basis points over the risk-free rate. Seventy basis points of risk premium evaporated in a quarter, and not one basis point of it shows up in asking prices.

Sit with the mechanics, because this is how private-market repricing always starts: invisibly. The seller’s broker opinion of value still says 6.82%. The buyer’s spreadsheet now discounts against 5.12%. Both are “right,” which is why volume — not price — is the first casualty: Newmark’s sale-leaseback share at half its historic average is what 170 basis points looks like in transaction data. The spread has to be rebuilt, and there are only two tools: the 10-Year falls, or cap rates rise. The futures market puts 75% odds on the Fed pushing the first tool further away in October. The last time spreads compressed like this, in 2022–23, cap rates took two to four quarters to answer — and the answer, when it came, was not optional. If you are pricing a deal today at Q2 comps, you are not buying a 170-basis-point spread; you are betting it widens back in your favor before your hold period cares. That is a rates call wearing a real-estate costume. At minimum, know that you’re making it — and price the bumps, the credit, and the re-tenanting story as if the Treasury is what it is, not what the comp sheet remembers.

Munger’s favorite instruction was to ignore what people say and watch what they do — incentives drive behavior, and behavior is the only honest disclosure. Revealed preference, the economists call it. He put it more bluntly: show me the incentive and I will show you the outcome.

Run this week’s tape through that filter and notice that every actor in net lease just told you their forecast without saying a word. The middle-market owner-user, offered today’s cap rate, declined to sell — sale-leasebacks at half their historic share is thousands of CFOs revealing they believe today’s bid is tomorrow’s regret. CBRE IM spent $1.6 billion buying a sale-leaseback platform rather than sale-leaseback assets — revealing it believes the assets get cheaper and wants the machine in place when they do. J.P. Morgan’s investors doubled the fund target for the same reason; you do not commit $1.1 billion to a strategy because you think its inventory is fairly priced today. Getty sold at 5.7 and bought at 7.1 in the same fiscal year — the cleanest arbitrage of belief in the sector. Even Silver Point, extending Car-Mart six days at a time, is revealing a precise estimate: the transaction on the table is worth more than the liquidation, but only barely, and only this week. The one actor whose behavior hasn’t changed is the private seller quoting June’s cap rate — and a quote without a trade is not a price; it is a wish. Ask of every number you see this quarter: did money actually move at that level? The 6.82% moved almost no money. The 5.12% clears a trillion dollars a day. When the talkers and the doers disagree, Munger’s rule tells you which market is lying — and it is never the one with the volume.

The Latticework Letter is published weekly on Friday mornings. It is independent analysis — we have no brokerage relationships, no listings to push, and no financial products to sell. Our only interest is giving you a clearer picture of the market.

Forward this to someone who owns NNN assets and is tired of getting their intelligence from people with something to sell.