FAT Brands has to sell. Here's what the AI premium does (and doesn't) buy a multi-brand QSR.

It’s a Monday morning, the kind where the docket alerts started landing in my inbox over the weekend and by Sunday night I had four PDFs open and one cold flat white I had given up on. The FAT Brands case file has gone from “watchlist” to “live auction in two weeks,” and the part of my desk that usually has the spring conference agenda on it instead has the bid procedures order from the Southern District printed and marked up in red. The auction is coming. The brands — Fatburger, Johnny Rockets, Round Table Pizza, Twin Peaks, Fazoli’s — are going to clear. The question is who buys, at what level, and on what thesis.

The contrarian read — and this is INTERPRETATION, marked as such — is that the thing depressing the bid book is not the comp trend, not the franchisee attrition, not the macro. It is the tech debt. Five chains, five POS estates, no unified loyalty, no standardized digital menu surface, no shared data spine. The financial buyer looks at that and sees an integration tax. The strategic buyer looks at the same thing and sees a $200–400M synergy line that only shows up if you can credibly underwrite a multi-year standardization onto a single modern stack. That is the bid-depressing factor. That is also where the deal gets won.

Let me walk it.

The filing, and the timeline that matters

FAT Brands filed Chapter 11 on January 26, 2026. The proximate cause was a debt stack the company had been carrying since the Twin Peaks SPAC unwind and a series of franchisee disputes that the Restaurant Dive coverage of the bankruptcy impact walks through cleanly. The court push in mid-March — the March 17 status conference, the March 20 bid-procedures order — converted what had been an open-ended restructuring into a hard auction track. The TheStreet writeup confirming the chains are formally up for sale is the cleanest public articulation of what’s on the block: Fatburger, Johnny Rockets, Round Table Pizza, Twin Peaks, Fazoli’s. Five chains, five different unit economics, one auction.

Bidding opens in April. That is the part of the calendar that matters. Indications are due, then qualified bidders, then the auction itself before mid-quarter. The court wants this priced and closed before the back-half operating season, which means the diligence window is short and the buyer who wins is the buyer who has already done the work. Not the buyer who shows up at the auction with a clean balance sheet and a SaaS thesis.

I’ll come back to who that buyer is. First let me walk the parts.

Sum-of-parts: what each chain is, on a clean basis

The honest version of this exercise is that nobody outside the data room has the unit-level P&L. What we have is segment-level disclosure from the pre-petition filings, the public franchise disclosure documents, and what operators in the space have been saying about each concept. The numbers below are interpretation; the directional ranking is not.

Fatburger. The legacy concept, ~150 units globally, mostly franchised. International unit economics are stronger than domestic; the Middle East and Asia books carry the brand. Interpretation: clean EV in the $80–110M zone on a royalty-stream basis, assuming a buyer who can stabilize the domestic comps without re-litigating the franchise relationships. The brand has equity. The brand also has a fifteen-year-old POS estate at the franchisee level that nobody has standardized.

Johnny Rockets. Smaller domestic footprint than people remember, larger international, a 50s-diner concept that has spent a decade looking for a second act. The asset is the IP and the international royalty book. Interpretation: $40–60M on a strip-and-license basis. A strategic buyer who already owns a diner concept might pay more for the IP than a financial buyer would pay for the cash flow.

Round Table Pizza. This is the underrated asset in the file. ~400 units, West Coast concentration, franchisee base that has held together better than the parent. Pizza is a category that has been re-rated on AI optionality more cleanly than burgers — the throughput math on a make-line is genuinely improvable with a forecasting layer and a kitchen-display refresh. Interpretation: $120–160M, with upside if a buyer can plug it into a modern digital ordering stack. The Pizza-category buyer pool is real.

Twin Peaks. The trouble child. ~115 units, the polished-casual concept that anchored the SPAC story two years ago and that the market has since marked down hard. The unit economics at the company-owned book are not the unit economics at the franchised book; the gap is wide enough that the diligence reads as two different concepts. Interpretation: $150–220M, with the high end requiring a buyer who is willing to separate the company-owned and franchised books and run them differently. Most financial buyers are not. A strategic operator might be.

Fazoli’s. Italian quick-service, ~210 units, the cleanest stand-alone business in the file. The comps have been the most resilient in the portfolio and the franchisee relationships are the least contentious. Interpretation: $130–170M as a stand-alone, and probably higher if the auction allows piecemeal bidding on Fazoli’s by an Italian-category specialist.

Add the ranges: $520–720M of brand-level value. Call it ~$620M at the midpoint. That is the number that should be in front of any bidder before tech-debt adjustments. It is also the number that, if the parent could sell each chain to a different optimal buyer in a clean process, would clear above $700M without much argument.

The auction is not that clean process.

The tech-debt discount — and why it’s the bid-killer, not the comp line

Here is the part the press coverage has not put a number on, and where the contrarian read lives.

A multi-brand QSR holdco, in 2026, is a tech estate. Five chains means five POS vendors at the franchisee level, five loyalty programs (in name; the real number is zero working ones), five digital menu boards from five integrators, five online-ordering stacks bolted on at different times by different VPs, and a corporate data warehouse that, charitably described, is a quarterly ETL job into a BI tool that the new CFO hasn’t fully audited.

The financial-buyer underwriting on this is unforgiving. The cost to integrate is real. The lift-and-shift onto a unified modern stack — call it Toast at the franchisee level, Olo at the digital-ordering layer, a third-party CDP for the loyalty spine — is a 24-to-36-month program with seven-figure costs per brand per year before any of it shows up as savings. A pure financial sponsor models that as a cost line, applies a discount rate, and arrives at a bid that is 15–25% below the brand-level sum-of-parts. That is the discount. That is what is depressing the book.

This is the same shape, structurally, that drove the NRP Florida bankruptcy when the franchisee-side leverage caught up to a tech estate nobody had refreshed in eight years. Different layer of the cap stack, same underlying read: when the operating system is fifteen years old and the financing market has finally noticed, the asset clears below the inventory value of its own brand book.

The thing the bear case gets right is that the discount is real. The thing the bear case gets wrong is who can underwrite it back into the bid.

The synergy thesis — Toast or Olo standardization, and the $200–400M number

Here is where the AI premium actually buys something concrete, and where I want to mark my work as interpretation.

If a strategic buyer — someone who already operates a multi-brand QSR platform at scale, or a private equity consortium that has already pre-baked the integration partner before bidding — enters the auction with a standardization plan that goes to a single POS vendor and a single digital-ordering vendor across all five chains, the synergy math changes shape.

I get there with three components.

The first is POS rationalization. Five legacy POS estates onto a single modern platform — and I am explicitly naming Toast here because it is the only vendor with the franchisee-side onboarding muscle to do this at speed across 800+ combined units — produces ongoing software-cost savings, eliminates a corporate IT team, and crucially makes the franchisee base bankable in a way it currently is not. Lenders will refinance a franchisee on Toast at a margin that does not exist on a 2011-vintage POS. Interpretation: $50–90M of NPV across the five brands over a five-year hold.

The second is digital ordering and loyalty. Five separate ordering surfaces consolidated to one Olo deployment, with a single guest identity layer across the holdco. This is the leg the Olo App valuation work is built on — the network effect of a unified consumer surface across multiple brands is real, and a holdco that owns five concepts is the most natural place for it to land. Interpretation: $80–150M of NPV, on the assumption that the cross-brand loyalty pull-through delivers 100–200 bps of incremental same-store sales by year three.

The third is data spine and operations AI. One CDP, one demand-forecasting layer, one labor-scheduling vendor across the entire holdco. The savings here are the boring kind — overtime reduction, food-waste reduction, marketing-spend efficiency — but they compound. Interpretation: $70–160M of NPV across the five brands over the hold.

Add them: $200–400M of synergy NPV to a buyer who can credibly underwrite the standardization. The range is wide because the execution risk is real and because the franchisee adoption rate is the variable nobody can underwrite tightly from outside the data room. But the direction of the math is not a stretch. The same shape, at smaller scale, is what shows up in any well-run multi-concept platform conversation, and it is the shape the forthcoming Q2 M&A roundup will need to track as bidders surface. Apply that same logic to a single-concept multi-unit deal — the kind of 12-unit cafe-group pricing exercise where a tech-stack standardization line is a 10–15% adjustment to multiple — and the FAT Brands case is just the multi-brand version, with the synergy number compounding across five brand books instead of one.

That $200–400M, applied against a $520–720M brand-level sum-of-parts, is the difference between a strategic bid that clears the auction comfortably and a financial bid that walks.

The bet: who wins, who pays par, who walks

Here is how I’d handicap the bidder pool, with appropriate caveats for an asset I cannot trade.

The strategic buyer wins. Not necessarily a single-name strategic — Inspire, Yum, Restaurant Brands International are all probably too big and too distracted to bid a chapter 11 multi-brand platform — but a focused mid-cap multi-concept operator with the integration partner pre-baked. Someone who walks in with the Toast deployment plan already costed, the Olo enterprise contract already pre-negotiated, and a clear answer to the franchisee base on day one about what changes and when. That buyer can underwrite the brand-level $620M plus the synergy NPV, get to a winning bid in the $700–800M zone, and still have a sensible IRR on a five-year hold. This is the bid I expect to clear.

The financial buyer pays par. A private equity sponsor without the standardization plan locked in walks in modeling the brand book at the discount — the 15–25% tech-debt haircut on the $620M — and bids somewhere in the $470–530M zone. They are not wrong on their own math. They are wrong on what the asset is worth to someone who can do the work they cannot. They lose the auction unless the strategic doesn’t show up.

The PE consortium walks. A multi-sponsor consortium needs alignment on the integration thesis, the timeline, and the operating-partner choice. Five brands, five operating committees, five sets of LP approvals — the governance overhead alone is enough to make the synergy math impossible to execute. Even if the consortium can theoretically write the check, the delivery risk on the standardization plan that justifies the synergy is too high for any one sponsor to underwrite. Consortium bids fall apart in diligence. The deck never makes it to final.

Mark this: the auction is not a referendum on whether FAT Brands’ brands are good. They are mostly fine. The auction is a referendum on whether the bidder pool contains a buyer who can credibly underwrite a multi-year tech standardization across five franchisee bases. If yes, the asset clears at a number that makes the unsecured creditors whole and produces a recovery the trade press will frame as a surprise. If no, the asset clears at a number that books a loss and the press frames the gap as a “tech-debt discount” three months after I have already named it.

The thing I will be watching as the bidder list surfaces in April — and the thing the broader M&A roundup will need to track — is whether the named bidders have a credible standardization partner attached to the bid or not. A bid with Toast and Olo on the cap table of the integration plan is a winning bid. A bid without it is a placeholder.

Two weeks until indications. The auction is the rare clean test of the contrarian read: AI is not the story that bids the multiple up. AI tech debt is the story that bids the multiple down, until a buyer arrives who can credibly retire it. The financial buyer cannot. The strategic buyer can. That is the entire trade, and it will be priced inside a month.

The wider M&A coverage tape will tell you who shows up to the auction. The bid level will tell you whether they did the work before they got there.

— Oliver writes The Bottom Line for TableTransfers. Tips: [email protected].