CODI — Compass Diversified Holdings · 2026-09-06 · Verdict: WATCH · Conviction 2
Price $11.67 (screen row, universe_under2b.csv, screen dated 2026-09-05) · Mkt cap $878.0M · EV $2.33B on the screen, ~$2.89B once ~$503M of preferred is added · EV/EBIT NM (209x trailing) · FCF yield -5.8% trailing · Net debt $(1,505)M plus $503M preferred · ADV $9.1M Sources read: 10-K 2026-02-27 (Items 1, 1A, 7), 10-Q 2026-08-10 (Item 2), DEF 14A 2026-04-10, 8-Ks 2026-06-11, 2026-06-24, 2026-07-01, 2026-07-13, 2026-07-16, 2026-08-10 (incl. EX-99.1 Q2 earnings release), Form 4s (trailing 12m).
1. What the business actually does
CODI is a Delaware statutory trust that owns controlling stakes in eight middle-market operating companies and is run by an external manager, Compass Group Management (10-K, Item 1). Five are branded consumer: 5.11 tactical apparel, BOA fit systems, PrimaLoft insulation, The Honey Pot feminine care, Velocity Outdoor archery. Three are industrial: Altor Solutions protective foam, Arnold magnetics, Rimports home fragrance (10-Q, Item 2). CODI funds each subsidiary with intercompany loans, $704.1M outstanding at 6/30/26, and the parent lives on the interest and principal on those loans (10-Q, Item 2). Ownership runs 76.5% to 93.2% fully diluted (10-K, Item 1).
2. Why it is mispriced — the edge case
The honest answer is that the obvious edge has already been arbitraged. The fraud at Lugano voided 2022 through Q3 2025 financials, the distribution was suspended in May 2025, and the shares were dumped (10-K, Item 7). But the screen row shows the stock up 84% in six months, 12-1 momentum of +47.7%, and only 8.2% off its 52-week high. The disgusted seller has already sold. What remains is a structural argument, not a forced-seller one: the manager's fee is falling, the balance sheet is shrinking, and a sum of the parts on subsidiary EBITDA is well above the equity price. Six insiders bought 209,046 shares for $2.47M with zero sales, the largest being Chairman Larry Enterline at $12.42 on 2026-08-13 and director Glenn Richter at $13.14 on 2026-08-14, all above today's price (Form 4s, 12m). That is real but small: 0.28% of the market cap, from people whose manager is paid on assets rather than on the share price.
3. Unit economics and growth
Q2 2026 subsidiary Adjusted EBITDA was $91.5M, up 12.6%, with Branded Consumer up 24.2% and Industrial down 12.8% (Q2 press release). BOA revenue grew 22.1% at a 33.9% segment margin, PrimaLoft 19.7%, The Honey Pot 17.0%; 5.11 revenue fell 3.8% but margin rose from 7.4% to 9.7% (10-Q, Item 2). The industrial side is the problem: Altor revenue fell 21.2% and its segment margin went from 8.7% to 1.6% on white-goods and cold-chain weakness plus a 21% rise in its main raw-material index. Altor and Velocity were both out of compliance with their intercompany covenants at 6/30/26 and received waivers, and PrimaLoft took a $20.5M goodwill impairment in the 2026 annual test despite growing (10-Q, Item 2). Group cash conversion is thin: 1H26 operating cash flow was $53.6M against $11.3M of capex, but $40.9M of that came from a working capital release (10-Q, Item 2).
4. Balance sheet and capital allocation
Total debt was $1,592.3M at 6/30/26 against $87.4M of cash, down from $1,890.7M at year end after $282M of Sterno food-service proceeds went to the term loan (Q2 press release). The stack is $1,029.4M of 5.25% 2029 notes, $308.8M of 5.00% 2032 notes and $252.3M of term loan at 7.13% (10-Q, Item 2). Sitting between that debt and the common are roughly 20.1M preferred shares, about $503M of liquidation preference, costing $19.4M of distributions in 1H26 (10-K Item 1 issuance and ATM tables; 10-Q, Item 2). The screen's $2.33B EV omits that entirely. Covenant leverage was 4.8x at 6/30/26 from 5.3x at 3/31/26 (Q2 press release), but that ratio is struck on subsidiary EBITDA and ignores the holding company's costs. The August 6 Sixth Amendment extended maturities to 2028 but cut the revolver from $100M to $54M and adds a $4M fee if the term loan is not repaid by 12/31/26 (8-K 2026-08-10). Common distributions remain suspended; there is no buyback.
5. Management: what they said versus what they did
The MSA charged 2% of adjusted net assets, $74.8M in 2024 alone (10-K, Item 7). The Ninth MSA, signed 2026-07-13, cuts the base to 1.25%, caps 2027 base fees at $30M, and replaces the incentive fee with a 0.125% share alignment award and a 0.125% performance award that pays nothing in 2027 unless the dividend-adjusted share price reaches $17.25, an expected saving of $19M to $22M (8-K 2026-07-13). That is a real concession and a useful read on what the board thinks the shares are worth. Against that: the restatement showed management fees had been overpaid by $33.8M, recouped only through future fee offsets rather than a clawback (10-K, Item 7); corporate G&A went from $19.7M to $79.7M on $60.8M of investigation cost (10-K, Item 7); Grant Thornton issued adverse internal-control opinions for both 2024 and 2025 and a going-concern paragraph for 2024, and was replaced by Deloitte on 2026-07-16 (8-K 2026-07-16). Sabo retires as CEO on 12/31/26, succeeded by Zach Sawtelle (10-Q, Item 2).
6. Valuation
The tell is the company's own Adjusted Earnings, which adds back all amortization and impairment and then deducts preferred distributions: negative $11.3M in 1H26 and $1.8M for all of 2025 (10-Q and 10-K, Item 7). The common stub earns approximately nothing today. Corporate Adjusted EBITDA was negative $55.9M in 1H26, negative $91.1M in 2025 and negative $87.8M in 2024 (10-Q and 10-K, Item 7). 2026 subsidiary guidance is $320M to $365M including about $9M from Sterno, so a continuing base near $333M at the midpoint (Q2 press release).
Base: $345M subsidiary EBITDA in 2027 at 10x, close to the ~10.8x realized on the Sterno sale ($292.5M enterprise value, 10-Q Item 2), less a normalized $65M corporate drag also capitalized at 10x, less $1,400M net debt and $503M preferred, gives roughly $11.90 per share. Bear: $300M at 8x with an $80M drag that will not shrink because SEC and DOJ matters and shareholder suits stay live, and the equity is a stub worth $0 to $3. Bull: $380M by 2028 at 10.5x with a $55M drag and net debt at $1,150M after further divestitures, about $23. Weighting 50/25/25 gives roughly $12.30 against $11.67. Reverse DCF: at $11.67 the price implies subsidiary EBITDA of about $369M against a 9.5x multiple and a $65M corporate cost, which is above the top of the company's own 2026 guidance.
7. Catalysts and timeline
Ninth MSA economics begin 1/1/27. Q3 2026 results in November test the maintained guidance and the 5.75x covenant. Lugano plan confirmation would fix recoveries at 34.79% of inventory, tax and insurance proceeds (8-K 2026-06-24). Term-loan repayment before 12/31/26 avoids a $4M fee, and further divestitures and an eventual common distribution are the sequence management has laid out.
8. Risks and pre-registered kill criteria
- Subsidiary Adjusted EBITDA guidance cut below $320M, or Q3 2026 subsidiary Adjusted EBITDA below $85M.
- Covenant leverage rising above 5.25x at any quarter end, which breaches the 12/31/26 and 3/31/27 covenant steps.
- Corporate Adjusted EBITDA worse than negative $30M in any single quarter of 2027, meaning the fee cut did not reach the shareholder.
- Net insider selling above $1M, or an SEC or DOJ resolution with a monetary penalty above $50M.
9. Verdict and one-paragraph summary
WATCH, conviction 2. Compass Diversified is a genuinely improving situation, with subsidiary EBITDA up 12.6%, $298M of debt retired in six months, a manager fee cut worth $19M to $22M from 2027, and six insiders buying above today's price, but it is not a cheap stock. The common is a thin stub behind $1.59B of debt and roughly $503M of preferred that the screen's enterprise value ignores entirely, and the holding company itself burns $88M to $91M a year, so the company's own Adjusted Earnings were $1.8M in 2025 and negative $11.3M in the first half of 2026. Value the eight subsidiaries at the ~10.8x multiple CODI actually realized selling Sterno and then subtract the capitalized cost of running the holding company, and you land within a dollar of the $11.67 price, after the stock has already risen 84% in six months to within 8% of its high. The data point worth waiting for is the 2027 corporate cost run-rate once investigation expense rolls off and the new fee schedule takes effect; if that lands near $55M with Altor stabilized, the equity is worth materially more than today, and a re-rating back toward $8 would supply the margin of safety this price does not.
Research for discussion, not investment advice. Positions and sizing are the reader's decision.