Aluminium beverage cans · #3 globally · Americas (US, Brazil) ~60% / Europe ~40% of EBITDA · ~76% owned by creditor-controlled Ardagh Holdings
FCFF to the firm · equity as levered residual · leases as operating cost (EBITDAaL) · multiple cross-check · reverse DCF · valued 2026-10-14
OVERVALUEDRISK: HIGH
Price $4.27 · Base $2.32 · Price is +84% above base
Price (10/14 intraday)
$4.27
Base intrinsic value
$2.32
Bear → Takeout range
$0.41 – $5.15
Key driver
EV/EBITDA: each 1.0x turn = $1.31/sh (31% of price)
The Take — Sell the yield. The 9% dividend is paid with borrowed money, and the stock is priced for a takeout that may never come.
The equity is a thin sliver on top of a $4.2B debt stack. Net debt (incl. $325M leases) is $4,155M = 5.2x LTM EBITDA; book equity is −$750M. Market cap is $2.55B, so ~62% of EV is debt. That makes AMBP a 2.6x-levered bet on the EV multiple: every 1.0x turn of EBITDA is $1.31/share — 31% of the price. At 7.0x the equity is worth $2.21; at 8.0x $3.52. Today's quote needs 8.6x 2026E EBITDA — a Ball-like multiple for a business with half Ball's balance-sheet capacity.
The dividend is not covered and hasn't been for five years. $0.40/share = $239M/yr. Management's own 2026 cash bridge — EBITDA ~$782M less capex $240M, cash interest $220M, lease principal ~$215M, tax ~$30M, working capital — leaves ~$62M, i.e. ~26% coverage. 2021–25 cumulative FCF after leases was −$187M while $1.04B of dividends went out; net debt rose to pay them. In Nov-25 AMP issued $1.29B of 2031 secured notes partly to redeem the €250M 9% preferred — sensible (the prefs were expensive), but it moved leverage from 5.2x to 5.7x like-for-like. The low-coupon unsecured 2029s still have to be refinanced into a ~7–8% market: ~+$40M/yr of interest. Something gives — and the dividend is the only discretionary line.
Operations are fine; that's not the problem. Q2-26 adj. EBITDA $240M (+14%), Europe +36%, FY26 guide raised to $775–790M. Aluminium is a pass-through: revenue (+18%) inflates with LME and premia, margins don't — which is why we value EBITDA, not sales, and why Q2's beat (~half metal-timing, ~⅓ reverses in H2) shouldn't be capitalized. North America volumes are falling (−5% on contract resets), Brazil −15%. Normalized EBITDA ~$770M; FCFF on that at an 8.0% WACC is $2.32. Price needs ~$899M of normalized EBITDA (+15% vs guide) or a 7.0% WACC.
The bull case is a sale, not the business. Ardagh Holdings (bondholder-owned since the 2025 restructuring; ~76% of AMP) hired Evercore in August to sell. A strategic at 9.25x 2026E EBITDA prints $5.15; that's the ceiling, and minority holders may not get a control premium if AHSA sells its stake privately. Creditors-turned-owners want cash, not a 9% dividend leak to a 24% minority.
Call: Sell/avoid above ~$3.50; holders, use any sale-headline pop toward $5 to exit. Fair accumulation only below ~$2.50 (≈ base, where the dividend — even halved — is a ~8% yield). Do not short it outright: a takeout at 9–10x prints $5–6 overnight. Bearish expression with defined risk: a Jan-27 $4/$3 put spread (never financed with a short call — that is exactly the takeout risk). Holders who want to wait out the sale: write $5 covered calls, which caps you at roughly the takeout value while collecting premium.
Dividend sustainability — ~$239M paid vs ~$62M 2026 FCF after interest/leases. A cut re-rates the yield-buyer shareholder base.
HIGH
Parent / control — AHSA (creditor-owned) runs a sale process. Long: strategic takeout at 9–10x. Short: stake sold at a discount, or the process fails and parent cash needs pressure the dividend/upstreaming.
MED
Volumes & contract resets — NA −5%, Brazil −15% in Q2; can demand growth low-single-digit. Long: can-vs-glass/PET share gains, Europe +3–5%.
MED
Aluminium & tariffs — LME/premia passed through with lags (timing ±$10–15M/qtr); US Midwest premium/tariffs inflate working capital and can squeeze customer volumes.
LOW
FX — EUR/BRL translation; partly natural hedge via EUR notes.
What the "inventory" really is: a stable, low-growth can-making cash flow (~$770M EBITDA, ~$630M after leases) that is almost entirely pre-committed — to capex (~$240M), interest ($220M), leases (~$150–215M) and tax. The equity owns the residual of a residual; its value is a leverage-amplified bet on the exit multiple.
Scenario values per share vs price
Bear: normalized EBITDA $720M, 1% growth, leases 22%, WACC +50bp · Base: $770M, 3%→2%, leases 20%, WACC 8.0% · Bull: $790M, 4.5%, leases 18%, WACC −50bp · Takeout: 9.25x 2026E EBITDA $782M less $4,155M net debt (incl. leases), no minority discount. Prob-weighted (25/45/15/15): $2.56. Even the bull FCFF only reaches the price; only the sale clears it.
Driver tornado — value/share from base $2.32
Two inputs each move value by ±30%+: normalized EBITDA (±$50M) and WACC (±50bp). That is the signature of a levered residual — small changes in enterprise assumptions swing the equity wildly. No single swing reaches the price.
Green = >15% above price, red = >15% below. Equity value ÷ EV at 7.0x / 8.0x / 8.6x (market) / 9.25x: 24% / 34% / 38% / 43%. Every $50M of EBITDA ≈ $0.67/share at 8x; every 1.0x of multiple ≈ $1.31. Our DCF EV ≈ 7.1x 2026E EBITDA.
History — adj. EBITDA vs free cash flow after leases minus dividends ($M)
Green/red bar = FCF after leases − dividends paid. Negative every year: the dividend has been debt-funded since the 2021 SPAC listing.
Normalization worksheet ($M)
Year
Revenue
Adj. EBITDA
Margin
EBIT
Lease pmts
EBITDA − leases
Capex
FCF after leases
Dividends
Coverage
2021
4,055
659
16.3%
316
48
611
679
-269
0
n/a
2022
4,689
625
13.3%
263
59
566
585
-439
251
-1.75x
2023
4,812
600
12.5%
177
78
522
368
170
263
0.65x
2024
4,908
672
13.7%
218
97
575
167
186
264
0.70x
2025
5,497
739
13.4%
271
111
628
173
165
262
0.63x
TTM Q2-26
5,991
793
13.2%
309
134
659
187
174
250
0.70x
5-yr avg
4,792
659
13.8%
249
79
580
394
-37
208
Adj. EBITDA: company (2022–25; 2021 and TTM per stockanalysis/company H1). Revenue, EBIT, capex, lease payments, FCF (OCF − capex − lease principal) and dividends (common + preferred) from stockanalysis.com (IFRS). SBC immaterial (not disclosed separately). 2021–22 capex elevated by the US/Germany/Brazil growth programme.
What normalization says. 5-yr avg adj. EBITDA is $659M (13.8% avg margin; margins are meaningless here because aluminium inflates revenue) and 5-yr avg EBIT $249M. 2026 guidance ($775–790M) is a cycle high after two years of price/mix recovery and includes favourable metal timing. We normalize at $770M — above the 5-yr average to credit the 2021–22 capacity build, below the guide to strip timing. Normalized EBITDAaL (after ~20% lease cost) ≈ $616M; economic NOPAT (maintenance capex as depreciation, 25% tax) ≈ $400M on ~$3.4B invested capital (net debt + book equity) → ROIC ~12%, used as the marginal return on growth capex. Market-implied 8.6x is consistent only with ~$899M normalized EBITDA.
Base-case FCFF ($M)
Year
EBITDA
EBITDAaL
Tax
Econ. NOPAT
Capex
FCFF
WACC
PV
2027
793
634
74
406
316
235
8.00%
217
2028
817
654
78
416
323
242
8.00%
207
2029
841
673
83
426
270
309
8.00%
246
2030
867
693
88
436
278
317
8.00%
233
2031
893
714
94
447
286
325
8.00%
221
2032
918
734
106
449
284
334
7.95%
211
2033
941
753
118
451
281
344
7.90%
201
2034
964
771
130
453
278
353
7.85%
191
2035
985
788
139
457
276
363
7.80%
182
2036
1005
804
142
466
273
378
7.75%
176
Terminal
g 2.0%, ROIC 12% → reinvest 17% of NOPAT; stable WACC 7.75%
TV 6,715
3,132
EBITDA grows 3% (yrs 1–5) fading to 2%. Leases treated as an operating cost (lease cash ~17–27% of EBITDA; liability $325M with ~2–3-yr average life is effectively recurring opex), so net debt in the bridge excludes leases. Tax 25% on EBITDAaL − D&A ($340M incl. Rexam-era intangible amortization, rolling off after 2031). Growth capex = g × NOPAT ÷ 12% ROIC; + $60M/yr 2027–28 for Spain/UK capacity. PV explicit 2,086 + PV terminal 3,132 = EV 5,219 → − net debt ex-leases 3,830 = equity 1,389. Terminal share 60%.