MannKind Corporation [MNKD] — Trade Construction

The memo issues no position verdict. This document scores the Liquidity Criteria against the real chain and sets out what an expression would look like if the book chose to take one. It is not a recommendation and no entry has been appended to trade_recommendations.jsonl.

Spot $3.8875 (2026-07-29). Realised vol 252d 77.0%, 63d 71.3%.

1. Equity liquidity — PASS

Alpaca SIP feed, last 63 sessions:

Measure	Value
Median daily volume	4,083,013 shares
Median daily dollar volume	~$15.1m
52-week range	$2.28 – $6.21
Spot percentile in 52w range	41st

Any position this book would plausibly take is a small fraction of one day's volume. Exit is not a constraint.

Data-integrity note that matters beyond this name. The Alpaca IEX feed returns a median of 202,214 shares/day for the identical period — 5.0% of consolidated volume. Any screen or sizing tool in this pipeline reading feed=iex for liquidity is understating traded volume by roughly 20x and will reject perfectly liquid small caps. Verified by requesting both feeds for the same window on 2026-07-29.

2. The options chain — pulled, not assumed

criteria.md: "Any proposed options structure requires the actual chain pulled first — open interest and quoted size for the specific strikes and expiry." Done. Alpaca options contracts + snapshots, 2026-07-29.

Expirations available and total open interest:

Expiry	Total OI	Usable?
2026-08-07	4,480	weekly, thin
2026-08-14	2,166	weekly, thin
2026-08-21	37,606	yes — the liquid front month
2026-08-28	548	no
2026-09-04	217	no
2026-09-18	2,617	marginal
2026-10-16	841	no
2026-11-20	934	no
2027-01-15	31,891	yes — the only usable 12-month expiry
2027-02-19	233	no

Jan-15-2027 strike detail (the expiry that matches the 12-month target horizon):

Strike	Type	OI	Bid	Ask	Bid sz	Ask sz	IV	Delta
2.0	call	504	1.83	2.28	15	110	0.96	0.92
3.0	call	3,263	1.17	1.57	16	117	0.90	0.78
4.0	call	1,704	0.72	1.13	15	262	0.90	0.61
5.0	call	4,030	0.42	0.87	757	930	0.92	0.48
6.0	call	168	0.18	0.62	11	11	0.88	0.35
7.0	call	4,456	0.22	0.43	1	217	0.94	0.28
3.0	put	181	0.17	0.53	103	68	0.79	−0.22
4.0	put	256	0.72	0.88	25	1	0.74	−0.41
5.0	put	331	1.42	1.83	75	109	0.88	−0.54
7.0	put	1,361	2.81	3.58	106	210	0.79	−0.78

Aug-21-2026 strike detail (the liquid front month):

Strike	Type	OI	Bid	Ask	Bid sz	Ask sz	IV	Delta
3.0	call	3,088	0.77	1.27	625	35	1.27	0.84
4.0	call	5,732	0.37	0.43	30	187	1.14	0.52
5.0	call	5,318	0.00	0.17	0	23	—	—
6.0	call	3,335	0.00	0.22	0	15	—	—
2.0	put	3,029	0.00	0.03	0	100	—	—
4.0	put	345	0.28	0.63	309	322	1.02	−0.49

2.1 What the chain actually permits

The Aug-21-2026 $4.00 call is the single genuinely tight contract: 0.37/0.43, a 15% spread to mid, 30 × 187 quoted, 5,732 OI. Everything else on that expiry has a zero bid above the money — the $5.00 call has 5,318 OI and no bid at all. You cannot sell the $5.00 August call. A $4/$5 August call spread, which is the structure a naive screen would propose, is not executable as a spread — only the long leg is.
Jan-2027 is wide but two-sided at every listed strike. Spreads run 30–70% of mid. The $5.00 call is the best-quoted (757 × 930) but its 0.42/0.87 market is a 70% spread — expect to work the mid and accept partial fills.
The $7.00 January call has 4,456 OI and 1 contract on the bid. Open interest is not liquidity. Any structure short that strike is un-exitable at size.

2.2 IV context

Jan-2027 IV clusters at 0.79–0.96; realised 252d vol is 0.77 and 63d is 0.71. Implied sits above realised across the surface — MannKind options are not cheap. Long premium is paying a volatility risk premium of roughly 15–25 vol points on a name whose next binary (Q2 results) is inside two weeks. This argues for spreads over outright calls, and the spread markets argue back.

3. If the book wanted the exposure

Ranked by executability against the chain above, not by payoff aesthetics.

Option A — common stock (the default, and probably correct here)

The equity is deeply liquid ($15.1m/day); the options are not. Given a 12-month target of $4.55 (+17%) and a bear at $3.16 (−19%), the payoff is roughly symmetric and does not obviously reward paying a 15–25 vol point premium.
Sizing must be inverse-volatility. Per criteria.md, inverse-vol sizing is the active protection for a MEASURED Downside Criteria, and at 77% realised vol MannKind is in the top volatility tier. It is sized down automatically and materially.
Invalidation: a close below $3.16 on Tyvaso DPI / Tresmi news (see §4).

Option B — Jan-15-2027 $4.00 / $7.00 call spread (the only defined-risk 12-month structure that exists)

Buy $4.00 call at ask 1.13 (OI 1,704, 15 × 262); sell $7.00 call at bid 0.22 (OI 4,456, 1 × 217).
Net debit $0.91, max value $3.00, max profit $2.09 — 2.3 : 1.
Breakeven $4.91, which is +26% to spot and above the base-case target of $4.55. The structure only pays properly in the bull case.
Execution warning: the short leg has one contract on the bid. This is realistically a 20–30 contract trade worked at mid, not a size expression. Working both legs at mid gives roughly $0.60 debit (mid $4.00 call 0.925 − mid $7.00 call 0.325), a 4.0 : 1 structure, but there is no guarantee of that fill.

Option C — Aug-21-2026 $4.00 calls, outright, as an event trade into Q2 results

0.37 / 0.43, 5,732 OI, tight. 23 days to expiry. Delta 0.52, IV 1.14 against 63d realised 0.71.
Paying 114 vol for a 71-vol name into a print. This is an expensive lottery ticket, stated as such. It is included only because it is the one contract that will actually fill at size.

What is explicitly not available

No Jan-2027 put spread at size — the $3.00 put has 181 OI, the $4.00 put has 1 contract on the ask. A protective collar or a long-dated put hedge cannot be constructed on this chain.
No LEAPS beyond Feb 2027 (233 total OI). There is no 2028 expiry.

4. Risk management


Volatility tier	Highest (252d realised 77.0%) — inverse-vol sizing applies, position size scales down accordingly
Invalidation level	$3.16 (the bear target) on Tyvaso-DPI / Tresmi news, not on a vol move
Hard structural invalidation	Tyvaso DPI royalty revenue printing flat or negative year-on-year in any quarter. It is 36% of revenue at ~100% incremental margin and has gone +42% → +25% → +9%. The next step is zero.
Second invalidation	Furoscix net revenue printing below ~$15.5m in Q2 2026 — that would mean the Q1 sequential decline was the run rate, not a payer reset, and the entire long argument is one product that has stopped growing
Financing risk, not default risk	Blackstone is a 2030 bullet with a single $40.0m minimum-liquidity covenant against $133.9m of resources. There is no 2026 maturity wall. The risk is equity issuance into weakness: the 23 July 2026 placement raised $50m at a 0.6% discount to market to fund a $45m bill. Assume that repeats on any shortfall.
Correlation note for the book	MNKD's dominant economic exposure is to United Therapeutics' Tyvaso franchise (64% of revenue), not to diabetes. Any existing UTHR or PAH exposure in the book is correlated with this position through the same underlying product, in the same direction on volume and the opposite direction on Tresmi substitution.

5. Liquidity Criteria — PASS

Equity liquidity is ample. An options chain exists, was pulled, and one 12-month defined-risk structure is constructible (Option B) — at ~25 contracts, not at size. The framework's default assumption that a defined-risk spread is always available fails here for puts entirely, and the naive August $4/$5 call spread that a screen would propose is not executable because the $5.00 August call has no bid. Recorded so the pipeline learns it.