The Latticework Letter — Issue #009

Last issue the story was that the September hike had died because the economy cracked first. This week the bond market sent a follow-up: the hike may be dead, but the long end doesn't care.

Monday the 30-Year Treasury touched 5.31% — its highest level since July 2007. Not on hike fears. On supply. Deficits, a heavy long-dated auction calendar, AI-related corporate issuance crowding into the same duration bucket, and five straight years of inflation above target. The 10-Year followed it up, testing roughly 4.75% Tuesday, a twenty-month high. For two days, the curve steepened for reasons that had nothing to do with the Fed.

Then two things arrived on Wednesday. First, Treasury announced it will at least double the size of its long-end liquidity-support buybacks, from $2 billion to at least $4 billion per operation, effective September 9 and running through the November 4 refunding. Second, the July FOMC minutes landed. The 30-Year fell 7.8 basis points to 5.207%; the 10-Year eased to 4.65%.

The relief lasted about a day. By Thursday's close the whole trade had unwound — the 10-Year back up more than five basis points to 4.704%, the 30-Year to 5.248%. A direct Treasury intervention aimed squarely at the long end bought roughly twenty-four hours. That is the single most useful thing that happened this week, and it wasn't the announcement. It was how quickly the market took the gift back.

The minutes were the more interesting document. The July hold was a 9–3 vote — Hammack, Kashkari, and Logan each dissenting for a quarter-point increase, with the funds rate at 3.50%–3.75% — and "many" officials on record that tightening would be needed if inflation didn't come down. That is a more hawkish committee than the September odds imply. Markets currently put a September hike at roughly 34%, down from about 60% three weeks ago. The gap between what the committee wrote in late July and what the market believes today is entirely explained by data that arrived after the meeting.

That data has not improved. Preliminary August consumer sentiment came in at 51.0 against a 54.5 consensus, down from 55.2 — a reading below the first percentile of the survey's history and below the level at the start of all six recessions in the series. July retail sales fell 0.6% month-over-month, the largest drop in more than a year, with the year-over-year figure decelerating to 5.0% from 6.7% in June. Weekly claims stayed benign — 206,000 initial, below the 210,000 consensus — but continuing claims ticked up to 1.799 million. Firms aren't firing. They aren't hiring either.

And the constraint is still oil. Brent pushed above $92 Wednesday and above $94 Thursday, a three-week high, after the UAE severed financial and economic ties with Iran and the peace talks stayed deadlocked. Hormuz traffic remains functionally closed: three vessels transited on Sunday, against a five-day average of twelve, against roughly 130 a day before February.

So the landlord's picture this week is a specific and uncomfortable one. Your discount rate is being set by a bond market worried about supply, not by a Fed worried about growth. Your tenant's revenue is being set by a consumer at a sentiment reading of 51. Neither of those improves if the Fed holds in September. Warsh gets his first Jackson Hole keynote next Friday, August 28, nineteen days before the FOMC decision — that's the next thing that moves this.

While the rest of retail net lease trades in the sixes and sevens, one category is still clearing in the low fives — and this week it produced the sharpest fight in the sector.

The category: 800-to-1,000-square-foot double-lane drive-thru coffee boxes. Dutch Bros properties are trading below 5.25% cap rates, against a historical range of roughly 5.00%–5.75%. 7 Brew — founded in 2017, past 500 locations, backed by Blackstone capital — prices in the same neighborhood. Compare that to the broader single-tenant retail average of 6.60% in Boulder's most recent quarterly survey, or auto parts at roughly 6.65%. Coffee pads are commanding something like 135 basis points of premium over the retail average.

The underwriting case is unit economics. Dutch Bros posted record average unit volumes of $2.1 million in 2025 and plans at least 181 new locations in 2026; 7 Brew reports average annual gross sales near $2 million per franchised unit. On a small box with a low rent basis, those AUVs produce rent coverage that a casual-dining tenant cannot approach. Add fifteen-year initial terms, ground-lease structures on many sites, and hard corners chosen for drive-thru stacking, and you get the closest thing net lease has to a growth story.

Now watch what happened when the category met a bankruptcy docket. In the Salad and Go Chapter 11, Dutch Bros' affiliate signed an exclusive letter of intent for 51 Arizona and Nevada sites at $105 million, plus fourteen shuttered Texas and Oklahoma leases (the fifty-dollar line item we covered last issue). At the August 7 hearing, 7 Brew appeared as a competing bidder and objected to the process — its counsel arguing the exclusivity provision barred the debtor from talking to 7 Brew for a week before filing, that a termination fee and a $10 million minimum overbid made little sense when two bidders were known to exist, and that 7 Brew had offered to serve as stalking horse and had submitted a revised bid with more consideration and more leases. Debtor's counsel confirmed the remaining roughly 100 leases will go to a competitive process, with a bid procedures motion and a sales hearing by September 3.

Here's what that means if you own one of these pads. The premium you paid for a sub-5.25% cap rate was a bet on the operator's growth, not on the real estate. When the operator fails, the value of your lease is decided in a courtroom by how the sale process was structured — and a well-run auction with two bidders produces a materially different outcome for landlords than an exclusive LOI negotiated before the petition date. The 51 sites slated for Dutch Bros will be assumed and re-tenanted. The 100 sites headed to auction will find out in September. Same brand, same buildings, two very different recoveries — decided by process design, not by dirt.

Wendy's had the loudest week. U.S. President Pete Suerken departs August 31 and the role is being eliminated; the company is reviving the COO position, last held six years ago by current CEO Bob Wright. The filing language — the company "is evaluating restructuring and reorganization efforts" — is the part that matters to landlords. Meanwhile Trian is reportedly assembling a consortium including BlueFive Capital and Flynn Group, one of the system's longest-tenured franchisees, for a take-private bid; no formal offer has been made and the timing could slip. The stock jumped roughly 5% Wednesday on the combination. Six consecutive quarters of declining same-store sales, a president out, a COO seat being rebuilt, and an activist consortium with a major franchisee inside it — if you hold Wendy's paper, the relevant question is no longer the brand's credit but whether your specific unit survives whatever plan the new owner writes.

Walmart's Q2 landed Thursday: revenue $187.9 billion, up 5.9%, adjusted EPS of $0.81 against a $0.74 estimate, and U.S. comparable sales of 2.6% against expectations near 3.5%. Guidance went up — net sales +4.0%–5.0% in constant currency, adjusted operating income +7.0%–8.5% — and the shares closed down 9.2% at $103.84 anyway. More on that number below.

Home Depot on Tuesday was the week's quiet positive: sales $47.9 billion, up 5.7%, comps +1.7% (U.S. +1.3%, the best since late 2022), adjusted EPS $4.92, guidance reaffirmed. Management described the housing market as "frozen" — homeowners sitting on 6.5% mortgages renovating instead of moving, average ticket up 2.8% to $92.50. That is the maintenance economy, and it is a real thing to underwrite. Target Wednesday posted comps of +3.8% against a 2.4% consensus on $26.5 billion of revenue, with digital comps +8.7%, and raised full-year guidance. Lowe's Thursday: EPS $4.33 against $4.24, revenue $23.96 billion, essentially in line.

Advance Auto Parts' quarter is the one worth reading twice. Net income rose to $55 million from $15 million, gross margin expanded more than 250 basis points, and the company raised full-year adjusted EPS guidance to $2.60–$3.30. Sales and comparable sales both declined. The Pro channel grew low single digits; DIY shrank as household budgets tightened. The profit improvement came substantially from closing stores — the company operated 4,311 locations as of July 18, down sharply from its footprint two years ago. Earnings quality that comes from shrinking the footprint is good news for the equity and neutral-to-bad news for whoever owns the boxes still on the list.

On the calendar: America's Car-Mart's covenant waiver period expires September 7, extendable to September 21 or November 6 under conditions; the company must keep $7 million of liquidity every Friday and $5 million otherwise while the strategic review runs. Two and a half weeks out.

That is Walmart's U.S. comparable sales growth for the quarter — the slowest since 2020.

Take a second with it. Walmart is the tenant everyone in this business calls recession-proof, the anchor whose presence makes an outparcel financeable, the credit that lets a shopping center sponsor tell a lender the traffic is guaranteed. In a quarter when reported inflation ran in the low threes, it grew U.S. comps 2.6% — and roughly 0.8 points of that was dragged down by price caps on certain drugs in its health and wellness business, which is a policy effect rather than a demand one. Even backing that out, you are looking at nominal growth roughly in line with inflation. Real unit volume at the most defensive retailer in America was close to flat.

There is a second number sitting next to it that explains the market's reaction better than the comp does. Walmart said it is eligible for approximately $2.9 billion in tariff refunds, and that it intends to spend that money lowering prices for customers starting in the third quarter — while also expecting just over $2 billion of incremental tariff cost going forward. So: a multi-billion-dollar windfall, immediately committed to price investment, against a comparable ongoing cost. That is not a company that thinks its customer can absorb anything. That is a company using a one-time refund as a shield.

For landlords the read-through is direct. If Walmart is spending a $2.9 billion refund to hold traffic, the tenants two credit notches below Walmart — the franchised QSR, the regional grocer, the discount apparel box — do not have a refund to spend. They have a rent obligation. Underwrite accordingly.

Munger's most-quoted line is also his most under-applied: "Show me the incentive and I will show you the outcome." He meant it as a warning about how rarely we look past the stated reasons people give for their behavior to the structure that actually pays them.

Apply it to the week. In the Salad and Go docket, the debtor's professionals negotiated an exclusive agreement — with a termination fee and a $10 million overbid hurdle — with one bidder, at a moment when a second bidder had already put a letter of intent in front of them. Everyone involved acted rationally within their own incentive structure. The debtor's advisors are paid to deliver certainty of closing. The buyer is paid to avoid an auction. Nobody in that room is paid to maximize the recovery on your particular lease. The landlord — the party whose asset is being disposed of — is not at the table, has no seat to lose, and finds out the result from a docket entry.

The same lens explains the rest of the week. Advance Auto's margin improved because closing stores is what the incentive structure rewards, and no incentive anywhere in that company attaches to what happens to a landlord after the store goes dark. Walmart is committing a tariff refund to price rather than margin because market share compounds and a one-time refund doesn't. Trian is assembling a franchisee-inclusive consortium for Wendy's because the franchisees are the ones who know which units can be saved — and, not incidentally, which cannot. Three unrelated stories, one mechanism.

The practical version for an owner of net-lease real estate is a question to ask before you close, not after: who at this table is paid based on what happens to my building in year eight? Usually the honest answer is nobody. The broker is paid at closing. The lender is paid on the loan. The tenant's operator is paid on this quarter's EBITDA, and the tenant's landlord-side counterparty may be a private-equity sponsor who will have exited before your first renewal option. Once you see that clearly, the standard advice stops sounding conservative and starts sounding like the only defensible position: buy the corner, not the credit; structure the escalations you actually want, because nobody will hand them to you; and read the lease's assignment and assumption provisions as if the tenant will one day be in Chapter 11, because a meaningful share of them will.

Munger's point was never that people are dishonest. It's that they're responsive. The incentive was always visible — you just have to be the kind of investor who looks for it before the docket makes it obvious.

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.