ADMA Biologics [ADMA] — Trade Construction & Risk

As of 2026-07-29 · spot $8.79 (2026-07-28 close) · framework v1.7.0

This memo issues no position verdict. It scores Criteria and sizes what a position would look like if the book chose to take one. The book decides.

1. What the analysis produced

The two mandatory outputs disagree in sign, and that is not a contradiction — it is the point of running both. Over 12 months the name can re-rate toward a plasma-peer EBIT multiple on a guidance-met print (+18.9%). Over five years today's price still requires growth the company is not guiding to. valuation.md: they measure different things over different horizons. The ownership test is the implied path, and it fails.

2. If the book took a position — construction

2.1 Vehicle: common equity only. Options are not investable.

The chain was pulled, as the Liquidity Criteria requires. Full surface, Alpaca, 2026-07-29:

Representative quotes, 2027-01-15 (the natural 6-month expiry):

Verdict: no defined-risk structure is fillable at a sensible price. The deepest single strike on the entire five-expiry surface carries 2,232 contracts, open interest is scattered across odd strikes, and every two-legged structure would cross 85–130% of combined spread. This is the HCA lesson in criteria.md: "a vehicle that cannot be filled is not a vehicle." Options: FAIL.

2.2 Equity liquidity: ample

At 10% of ADV a $2m position exits in one session. Equity: PASS. Size is constrained by conviction and volatility, not by liquidity.

2.3 Sizing, if taken

Inverse-volatility sizing is the framework's active protection (criteria.md, Downside Criteria interim control). At 55.4% realised 252-day vol, ADMA sits in the top volatility tier and sizes down automatically.

Correlation to SPY is 0.20 (β 0.88) — the risk here is almost entirely idiosyncratic and single-name. It will not be diversified away by anything else in a book; it will be diversified into.

Any position would be a top-tier-vol, BINDING-fail name. On a long-only absolute-return book that fails Quality and Valuation, the size is zero. No entry, exit, or stop is proposed, because proposing one for a name that fails two BINDING Criteria would be the type-discipline failure criteria.md warns about.

3. Downside Criteria (MEASURED — logged, scored, blocks nothing)

Volatility is not the risk. Permanent impairment is.

3.1 The permanent-loss case, with its named cause

Named cause: erosion of ASCENIV's price premium, by exactly the mechanism that destroyed BIVIGAM's economics inside the last twelve months.

This is not hypothetical extrapolation. It is the observed behaviour of the same company's other IVIG in the same industry reset:

• BIVIGAM revenue fell −54.0% year over year in Q1 2026, which ADMA attributes to "lower volume as a result of competitive pressures in the standard immune globulin industry."
• The company describes the cause as "elevated raw material plasma supply, increased competitive PDT inventories across the distribution network and aggressive discounting and rebating in standard IG."
• ADMA's defence is that ASCENIV is insulated because it is a late-line therapy with patented high-titer pooling. That defence held in Q1 2026 (+27.7%). It is a product-differentiation argument, and product differentiation in plasma-derived IG has a poor record of surviving sustained oversupply.
• ADMA has 73% of revenue with two distributors. In an oversupplied market, two buyers controlling three-quarters of your volume have the pricing power, not you.

If ASCENIV's premium compresses toward the standard-IG level, ADMA's 40.6% EBIT margin converges on the listed plasma peer set — EBS 13.5%, Kamada 14.4%, Grifols 16.5%.

3.2 Scenarios

Probabilities are estimates and are logged to be Brier-scored (ledger_scorer.py). They are not evidence and nothing in this memo is gated on them. Per criteria.md, the sensitivity that matters was run on the exit multiple, not on these probabilities.

Probability-weighted: $7.61 (−13.4%). Recorded for scoring, not used as a decision input — criteria.md retired the expected-return hurdle and warns explicitly against maximising over a known-biased estimate.

3.3 Is there a going-concern case? No — and it must be argued explicitly, so here it is.

Adjusted EBITDA would have to fall ~72% from guidance before the 2.50x leverage covenant binds. The permanent-loss case here is a valuation impairment, not a solvency one. Flagged as required and dismissed on the numbers.

3.4 The capital-allocation risk that is live

The $500m repurchase authorisation is open-ended and the ASR has ~5.0m shares still to be delivered by August 2026. The March 2026 tranche was funded with $125m of revolver debt at $16.58/share, five days after guidance that was cut 12–17% ten weeks later — a mark-to-market loss of roughly 47%. The risk is not that the buyback is imprudent in isolation; it is that the same forecasting process that mis-sized it is still the one producing FY2026 guidance. Any further debt-funded repurchase before the FY2026 guidance is re-established by two clean prints should be treated as a negative signal, not a positive one.

4. What would change this analysis

The analysis is falsifiable and the falsifying events are dated.

5. Ledger entry

Logged to trade_recommendations.jsonl as ANALYSIS_ONLY — no position recommended, carrying the bear case, its named cause and its probability for Brier scoring, per the Downside Criteria promotion path in criteria.md.

ticker            ADMA
date              2026-07-29
action            NONE (analysis only; memo issues no verdict)
spot              8.79
target_12m        10.45   (+18.9%)
implied_path      required 13.3% CAGR vs 6.8% guided -> margin -6.5pp
binding_fails     Quality (accruals + non-reinvestment), Valuation
bear_case         ASCENIV price premium erodes to plasma-peer level; $2.88 (-67%)
bear_cause        premium erosion in an oversupplied IG market, as already
                  observed in BIVIGAM (-54% YoY) within the same 12 months
bear_probability  0.15
vehicle           equity only; options chain pulled and FAILS liquidity