SANFILIPPO JOHN B & SON INC (JBSS) stock intrinsic value fair value estimate & owner earnings analysis

John B. Sanfilippo & Son buys, shells, roasts and packages peanuts, pecans, cashews, walnuts and almonds and makes trail mixes, nut butters and snack bars. Most of its $1.18 billion of fiscal 2026 sales (year ended June 2026) are private-brand products for large retailers, plus its Fisher, Orchard Valley Harvest, Squirrel Brand and Southern Style Nuts brands. It is a processor that earns a fairly thin margin between what it pays for nuts and what retailers pay for finished goods: gross margin has been 18-20% and operating profit $85-90 million in every year since fiscal 2021. How we measured earnings. Reported cash flow swings sharply because the company buys whole crops of nuts each season, so inventory moves by $50-60 million from one year to the next. We take out those inventory and other working-capital swings, charge stock compensation as a cost, and separate replacement spending from the roughly $90 million capacity expansion running from fiscal 2025 into early fiscal 2027 (mostly new bar and nut lines). On the formula alone, owner earnings were only $3.12 and $1.14 per share in fiscal 2025 and 2026, because the expansion spending is counted as upkeep. Counting replacement spending at about depreciation, the business generated about $5.46 per share in fiscal 2025 and $6.38 in fiscal 2026 before interest. Across fiscal 2020-2026 the normalized figure ranged from $4.77 to $6.38, with a median of $5.46. We use that median as the earnings level and do not use the stronger 2026 year. Fiscal 2017-2019 ran nearer $3.70-4.00, when the company was smaller and nut-cost pass-through was weaker. Fiscal 2026 normalized earnings rose about 17%, helped by pricing catching up with nut costs. Our base case still assumes no growth in the first year, because the company expects about $48 million of capital spending in fiscal 2027, roughly $20 million above replacement level. What drives it. The business depends on a few customers: Walmart was 41% of fiscal 2026 sales and the top five customers 67%. Those contracts are bid, so a lost program can remove a large block of volume at once. In fiscal 2026, total pounds sold fell 2.5%. Private-brand bar volume was down because of softness at a mass merchandiser, and branded volume fell 10.8% after Orchard Valley Harvest lost distribution at a major non-food customer. Revenue grew only because of price increases tied to higher tree-nut costs. Snack-nut and bar volumes are falling across the industry as shelf prices rise, and the filing names GLP-1 weight-loss drugs as a further pressure on snack demand. The founding Sanfilippo and Valentine families control the company through Class A shares, which carry ten votes each and elect three-quarters of the board. They pay out cash as regular and special dividends: $0.90 regular plus $1.60 special per share for fiscal 2026, and $0.95 plus $1.05 declared in July 2026. Base case ($82.35 per share). The business keeps earning about $5.46 per share before interest. Fiscal 2027 is lower (about $3.76) because the expansion spending finishes that year. After that, earnings stay flat with no growth credited. We give the new capacity no value in the base case because bar volume is currently falling and no signed program for it is disclosed. Valued at a long-bond-like rate, the business is worth about $1.05 billion, or $89.32 per share. It owes about $82 million: $33.6 million on the revolving credit facility, $44 million on the Wells Fargo equipment loan and a $5.6 million sale-leaseback obligation, less $1.1 million of cash. That leaves $82.35 per share for shareholders. The debt costs roughly 5-6% at floating rates, so we make no adjustment for its cost. The supplemental retirement plan payments are already charged in cash flow and are not subtracted again. Bear case ($63.18 per share). We use the company's own fiscal 2017-2019 record, when nut costs and retailer pricing moved against it and earnings stayed near $3.70-4.00 per share for three years, and add a loss of part of a top-five customer program. In this case, earnings fall to $3.90 per share for fiscal 2028-2030 and settle at $4.30 after that. The business is then worth about $826 million, or $70.15 per share, less $6.97 of net debt. Interest of about $3 million a year is easily covered at that level, so the debt does not threaten the bear case. Bull case ($94.31 per share). The capacity under construction is in service in fiscal 2027 and gets filled. In this case it adds about $0.25, $0.50 and then $0.76 per share a year, roughly a 10% after-tax return on the $90 million spent. We credit nothing beyond that, because the company's long record is of flat volume rather than steady compounding. The business is worth about $1.19 billion, or $101.28 per share, less net debt. Margin of safety price: $52. We are fairly confident in the earnings level. Operating profit has held at $85-90 million for six years through tree-nut price swings, a pandemic and an acquisition, and normalized earnings have stayed between $4.77 and $6.38 since fiscal 2020. The worst plausible event, which none of the cases fully models, is Walmart moving a large private-brand program to a competitor. That could take a quarter or more of earnings in a single bid cycle. The business would survive it: debt is under one year of cash generation, and the plants, nut inventories and receivables are hard assets that would cover the debt several times over. The things we are most likely to be wrong about are customer concentration, falling snack volumes and the return on the new capacity. If all three went the wrong way, the shares would be worth roughly $55-60. At $52, the shares carry about $2-4 a year in regular and special dividends. If the business only does what it has done in five of the last six years, the shares should be worth about $80 or more within three to four years. That implies roughly a 20% annual return including dividends, and the bear case value also sits above that price. The margin-of-safety price is not set higher because volumes are falling now and a single customer can change the outcome. Previous estimates were made on a different basis and are not comparable. There is no going-concern question, and the company sells almost entirely in the United States.