The Latticework Letter — Issue #017

The inflation data finally caught up to the war — and the bond market, for once, didn’t care. Wednesday’s September CPI printed hot on both lines: 3.7% headline against a 3.6% consensus (up from 3.4% in each of the prior two months) and 2.6% core against 2.5% expected. Gasoline rose 9.1% on the month and 34.2% on the year and drove more than half of the monthly increase; energy overall is running +20.4% year-over-year. Then Thursday’s PPI made CPI look polite: +1.3% on the month against +0.6% expected — the largest monthly jump since March 2022 — and +6.3% year-over-year, the biggest 12-month increase since December 2022, with core up 0.7% against 0.3%. Two prints, two misses, both in the wrong direction, five days before the quiet period ahead of the October 27–28 FOMC.

And here is the week’s genuine surprise: the 10-Year fell. The path tells the story. Friday closed at 5.24%, the bond market slept through Columbus Day Monday, Tuesday eased to 5.21%, Wednesday’s CPI spiked the yield to roughly 5.30–5.33% intraday before the move faded into the close — and then Thursday, on the hottest producer-price print in four and a half years, the 10-Year dropped about 13 basis points to roughly 5.05%, a three-week low. The mechanism wasn’t inflation optimism; it was fear changing addresses. The Nasdaq fell nearly 2% Thursday morning, the semiconductor index is down more than 10% from last week’s peak and formally in correction, and money rotating out of a cracking equity tape needed somewhere to stand. For the first week in months, the oldest hedge in the book actually hedged. Fed pricing whipsawed accordingly: October 28 hike odds ran from roughly 22% before CPI to about 50% within an hour of it, then faded to the high-20s-to-low-30s by Thursday (CME had it at 33.8% after PPI, versus 29.3% before). December is where conviction lives: futures price a December hike at roughly 90%, and a Reuters survey of twelve major brokerages found eleven — Goldman, J.P. Morgan, Morgan Stanley, Barclays, Citi, and the rest — penciling exactly one 25bp hike in December to 4.00–4.25%, with BofA the lone house calling for both October and December.

The war inputs behind the inflation prints haven’t improved. Kpler’s seven-day average of crude flows through the Strait of Hormuz is now running about 43% below pre-conflict levels, and the Financial Times counts 24 tankers hit over the past two weeks — the worst stretch of the war (the previous high-water mark, in March, was 13). Brent sat near $104.50 Thursday and WTI near $94. Yet the pump keeps easing: AAA’s national average was $4.33 at Wednesday’s CPI release, a shade below last week — the second straight weekly decline. Hold both facts at once: the price of crude is a war number, the price of gasoline is a demand number, and the gap between them is margin compression working its way through every fuel-anchored tenant’s P&L in your portfolio — mostly, so far, in their favor.

Now the spread math, which this week turned almost generous. Boulder’s fresh 6.92% overall ask against Thursday’s roughly 5.05% is about 187 basis points over the risk-free rate — from 149 two weeks ago and 170 last week. Read the progression honestly: two weeks ago the spread was thin because the Treasury had spiked; last week it rebuilt because cap rates moved; this week it widened again because the Treasury leg fell 19 basis points while the quote sheet stood still. That’s the fragile kind of widening — it evaporates the day the equity tape stabilizes and the rotation reverses, and with a 90%-priced December hike the front end isn’t going anywhere. But it is also, mechanically, the best entry math a leveraged net-lease buyer has seen since summer: the REITs noticed, with Realty Income up about 3% month-to-date in October (closing Wednesday at $55.17) after a brutal −11.5% September. One week of a working hedge is not a regime. It is, however, a reminder of what the asset class is for.

The week’s most instructive credit event is also its smallest. The Picklr — the Utah-based indoor-pickleball franchisor — filed Chapter 11 (Subchapter V) in the Northern District of Texas on October 9, parent company only; the 70-plus U.S. clubs, independently owned by franchisees across 29 states, stay open, and a franchisee-led investor group has committed $4.5 million of DIP financing subject to court approval. What put the holding company in front of a bankruptcy judge wasn’t club operations. It was the paper above them: The Picklr had corporately guaranteed facility leases for roughly 40 of its franchisees’ clubs, and those guarantees sat on a parent with $10.3 million of 2025 revenue and a $4.4 million net loss. Founder and CEO Jorge Barragan said it plainly: “We signed, unfortunately, too many corporate guarantees that landed on the holding company.”

Sit with the mechanics, because they are the whole lesson. Forty-odd landlords signed leases with thinly capitalized franchisee LLCs and took comfort — and probably a tighter cap rate at acquisition — from a franchisor guaranty. That guaranty felt like credit enhancement: a second obligor, a second balance sheet, a national brand behind the local operator. But the second balance sheet was never independent of the first. The franchisor’s revenue is royalties on the same clubs whose rent it guaranteed; when pickleball supply ran ahead of demand and club economics sagged, the tenant’s weakness and the guarantor’s weakness were the same weakness, arriving by two doors at once. And in Chapter 11 the guaranty converts into exactly what it always was: an unsecured claim against a $10 million-revenue holdco, standing in line, with the forward-looking promise that there will be “some network contraction” over 18–24 months. The landlords in that line don’t own credit enhancement anymore. They own litigation paper.

The market, to its credit, has been pricing this distinction all year — just look where the grading pen has already been. Boulder’s Q3 data puts corporate-backed QSR assets nearly 100 basis points inside franchisee-backed ones, across every lease-duration bucket: that’s the market saying a real corporate covenant and a franchisee-plus-wrapper are different species, not different flavors. The Crumbl file says the same thing from below — a Kentucky franchisee’s Chapter 7 earlier this month, at least 57 Crumbl closures in 2026, dozens of franchise units listed for sale — stress that lives entirely at the operator layer, where the guaranty math is thinnest. And Car-Mart (see Tenant Watch) is the scaled version of the identical question: what is a guaranty worth when the signer’s whole enterprise is the thing that’s failing? The emerging-concept version is simply the purest: a franchisor guaranty on an emerging brand is a bet that the system survives, dressed up as protection against the site failing. Those are not the same bet. They fail together.

What to do with it. First, reread every guaranty in your file and ask the only question that matters: what does the guarantor own that is not the business paying my rent? If the answer is nothing — a holdco whose assets are royalty streams from your own tenant’s siblings — you hold one credit with two signatures, and it should be priced like one credit. Second, underwrite guaranty durability, not guaranty existence: a guaranty from a parent with separable assets (real estate, other brands, investment-grade paper) is insurance; a guaranty from a single-concept franchisor is correlation. Third, on anything emerging-concept — pickleball, cookies, salads, boutique fitness — run the re-let math as the base case, because the 363 process (ask Salad and Go’s landlords, or soon True Food Kitchen’s) lets a buyer cherry-pick leases, and the guaranty you’re litigating won’t backstop the box that goes dark. Fourth, if you’re buying: the market’s 100-basis-point corporate-versus-franchisee spread is not an inefficiency to arbitrage. It’s a price on exactly this mechanism, and this week it printed a working example.

America’s Car-Mart bought another week — again. An 8-K filed Thursday extends the lender waiver to October 22, the eighth extension by our count (Sept 7 → 11 → 18 → 24 → Oct 1 → 8 → 15 → 22), agreed Wednesday, the day before the seventh expired. The language is unchanged — “significant progress towards a transaction,” verbatim for the fifth consecutive filing — and so is the forward-looking boilerplate naming bankruptcy protection and a total loss for stockholders. A deal in documentation keeps reading like a deal in documentation. Two new data points while we wait. One: Pomerantz LLP announced a securities-fraud investigation Tuesday, citing the July 2025 restatement disclosures and the September 9 Q1 print (revenue −57.3% to $145.8 million; net charge-offs 9.5% of average receivables versus 6.6% a year earlier) — ambulance-chasing follows distress, but it adds friction and discovery risk to whatever transaction is being papered. Two: the operational footprint is smaller than most landlord files assume — the company consolidated 60 dealerships during fiscal 2026, taking active locations from 154 to 94, with customers and loan books moved to surviving stores. The stock closed Thursday at $1.35. Behind the waiver stands a November 1 covenant-relief deadline, so the one-week extensions are now running into a two-week wall. The read since Issue #013 is unchanged and now urgent: a lender extending seven days at a time is a lender close to a signature, and the signature window is now October 22. If you hold any of the ~94 operating sites — or the 60 dark ones — your triage should have been done weeks ago. If a transaction prints before the 22nd, #018 takes it up.

Leslie’s had the best kind of week a debtor can have: a boring one. The case is tracking its prearranged script — the 76-store closure list is out and granular (26 stores in California alone, 8 in metro Houston, all per the omnibus lease-rejection motion), the equity is suspended, and the next gears turn on schedule: plan and disclosure statement due around October 20, final DIP hearing October 27, confirmation targeted for early January. The restructuring agreement still bars new store closures without lender consent, so the 900-plus surviving boxes stay open by covenant through the case. If you own one of the 76, your math is §502(b)(6) math and your claim clock is running; if you own one of the 900, the only date that matters this month is October 20, when the plan shows you what the reorganized credit actually looks like.

True Food Kitchen keeps moving through its sale process — the case carries about $42 million of funded debt per BankruptcyData, 34 restaurants operating, 12 closed at filing — with no stalking horse surfaced in the trade press as of Thursday. The Salad and Go precedent governs: when the bid does surface, the buyer cherry-picks leases, and lifestyle-center landlords find out which side of the line they’re on. Watch for the bid procedures order.

Briefly noted: The demand side of the tape still exists — Ross Stores completed a 115-store year (51 openings in September–October, including its first Massachusetts stores) and now operates 2,377 locations with a stated path to roughly 3,600; junior-anchor demand into sub-4.5% retail vacancy is what keeps re-let math honest. Rocky Mountain Chocolate Factory did a $6.6 million sale-leaseback of its Durango HQ at a 9.45% implied yield — to a buyer tied to the interim CEO’s family; related-party sale-leasebacks at 9-handles are their own disclosure genre and worth a skeptical read. Global Net Lease reports Q3 on October 29. No rating-agency actions on major net-lease tenants surfaced this week.

On the calendar: Q3 earnings open next week — Agree October 20; Getty, Essential Properties, and NetSTREIT October 21; W.P. Carey October 27; GNL October 29; Realty Income November 2. FOMC October 27–28 (hike odds ~30% and unstable; December ~90% priced). Leslie’s plan due ~October 20; final DIP hearing October 27. OZ 2.0: the October 28 submission deadline is 13 days out — the tracker shows 2,185 tracts nominated across 20 jurisdictions, about 33% of the expected 6,575, with New York, Florida, Ohio, Illinois, and Pennsylvania all still outstanding. Car-Mart’s eighth waiver expires Thursday, October 22, with the November 1 covenant-relief deadline behind it.

That is what Alpine Income Property Trust is paying per square foot for 1.9 million square feet of industrial — 13 properties across 11 states for $117.3 million, announced Friday evening and headed for a Q4 close. All of it absolute triple-net, 100% leased to nine tenants, 10.2 years of weighted average term, no expiration before late 2031, about 19% of rent from investment-grade names. The cap rate was not disclosed (“attractive initial yield”), which is itself a sign of the times — but the basis was, and the basis is the story.

Because $63 a foot is the anti-guaranty. This issue’s Spotlight is about paper protection that evaporates exactly when you need it; this number is the opposite kind of protection — the kind that doesn’t require anyone upstream to stay solvent. At $63 a foot, generic industrial in secondary markets is being bought for a fraction of what it would cost to build today, which means the re-let scenario — the one every guaranty is supposed to insure against — more or less underwrites itself: lose the tenant, and market rent on cheap basis still clears your yield. And Alpine’s portfolio carries the quietest credit statistic of the week: the tenants have operated at these sites for an average of roughly 40 years. Four decades of a business not leaving a building tells you more about renewal probability than most corporate guaranties tell you about collection probability. Boulder’s Q3 industrial numbers frame the entry: asking caps at 7.28% (up just 3bp, the stickiest sector) while industrial supply on the market rose 11.9% — more product, firmer pricing, and now a public REIT marking the clearing level at $63. In a market obsessed with who signs the lease, the week’s most interesting buyer paid up for what can’t be un-signed: dirt, steel, and tenure.

Charlie Munger’s latticework always included one engineering idea he considered close to sacred: the backup system. Bridges get redundant load paths, planes get second engines, power plants get emergency generators — and the engineering discipline behind all of it rests on a single, easily forgotten condition: redundancy only creates safety when the failure modes are independent. A second engine that draws from the same contaminated fuel tank is not a second engine; it’s a first engine with better marketing. Engineers are trained to hunt for this — they call it common-mode failure — because the deadliest systems are the ones that look redundant and aren’t: every backup fails precisely when the primary does, which is precisely when you need it, which is precisely when you discover you only ever had one system.

Finance relearns this lesson roughly once a decade, always at the bottom. The monoline insurers of 2008 guaranteed mortgage bonds while their own solvency rested on mortgage bonds; the wrap failed the day it was called. This week’s version seats forty-odd landlords in a Texas courtroom holding franchisor guarantees from a parent whose only real asset was royalties on the very clubs whose rent it guaranteed — two signatures, one cash flow, zero redundancy. So run the engineer’s audit across your own file, and be unsentimental about it: the guaranty behind your lease (does the guarantor own anything that isn’t your tenant’s business?); the master lease that diversifies you across twenty sites (operated by one company?); the portfolio diversified across forty tenants (in one concept? one war-sensitive cost structure?); even your macro hedges (how many of your “uncorrelated” positions are all, quietly, a bet on rates?). The question is always Munger’s style of question — invert it: what does my second line of defense have in common with my first? If the answer is “everything,” you own one defense and are paying for two. And note what the week offered by contrast, in both directions: Thursday, for the first time in months, Treasuries rallied while equities broke — a backup system firing because its failure mode genuinely differs — and Alpine paid $63 a foot for buildings whose protection is physical basis and 40-year tenants rather than anyone’s promise. Real redundancy exists. It’s just never the kind you get free with a signature.

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.