Ouster [OUST]
as of 2026-07-29 · spot $31.30 (SIP close) · no position verdict is issued here
This document sizes and constructs hypothetically, conditional on a book deciding to own the name. The memo issues no Long / Short / Watchlist / Avoid.
The brief instructs that Alpaca's IEX feed reports roughly 5% of true volume and that ADV should be scaled by ~20x. In this environment the SIP feed is entitled and was pulled directly, so no scaling was necessary — and the ~20x rule of thumb would have been materially wrong. Measured ratios, 20-day average share volume:
| IEX ADV | SIP ADV | actual ratio | 20x rule would have given | |
|---|---|---|---|---|
| OUST | 119,282 | 4,286,292 | 35.9x | 2,385,640 (−44%) |
| ONDS | 1,971,682 | 110,682,196 | 56.1x | 39,433,640 (−64%) |
| ASTS | 412,152 | 15,707,091 | 38.1x | 8,243,040 (−48%) |
Recorded for the calibration file: the 20x IEX multiplier understated true volume by 44–64% on these three names. Use SIP where entitled; the multiplier is a fallback, not a substitute.
| Metric | Value (SIP) |
|---|---|
| ADV, 20-day | 4,286,292 shares · $181.4m/day |
| ADV, 60-day | 5,188,003 shares · $207.7m/day |
| 252-day realised volatility | 104.9% |
| 52-week range | $16.63 – $62.52; spot at 50.1% of the high |
| 50-day / 200-day MA | $41.67 / $28.55 |
At $181m of daily dollar volume a $10m position is 5.5% of one day's volume — trivially exitable. PASS.
The brief's expectation was that chains on names this size would be uninvestable. On OUST that is false. Full active chain, expirations 2026-09-01 to 2027-07-01 (354 contracts):
| Expiry | Contracts | Total OI | Max OI at one strike |
|---|---|---|---|
| 2026-09-04 | 52 | 57 | 10 |
| 2026-09-18 | 52 | 1,834 | 588 |
| 2026-11-20 | 98 | 13,741 | 1,366 |
| 2027-01-15 | 96 | 32,862 | 6,231 |
| 2027-02-19 | 56 | 2,743 | 1,112 |
Selected Jan-2027 quotes (live snapshot, 2026-07-29):
| Contract | Strike | OI | Bid | Ask | Bid size | Ask size | IV | Delta |
|---|---|---|---|---|---|---|---|---|
| Call | 25.0 | 2,222 | 12.32 | 14.02 | 917 | 588 | 1.32 | 0.76 |
| Call | 30.0 | 1,856 | 10.26 | 12.22 | 861 | 644 | 1.32 | 0.69 |
| Call | 35.0 | 1,296 | 8.76 | 10.91 | 1,023 | 697 | 1.33 | 0.63 |
| Call | 40.0 | 1,474 | 7.42 | 8.64 | 917 | 554 | 1.27 | 0.56 |
| Call | 50.0 | 1,543 | 5.49 | 7.22 | 1,010 | 769 | 1.31 | 0.47 |
| Put | 30.0 | 508 | 8.57 | 9.56 | 824 | 531 | 1.23 | −0.32 |
| Put | 25.0 | 405 | 5.56 | 6.48 | 1,087 | 399 | 1.21 | −0.24 |
| Put | 20.0 | 860 | 3.14 | 4.15 | 1,227 | 854 | 1.23 | −0.17 |
Fillable — but expensive, and the reason is specific. Quoted sizes are in the hundreds of contracts and open interest at the round strikes is 1,000–6,000, so a defined-risk structure of a few hundred contracts is executable. This is a different world from the HCA precedent in this project's record (maximum 18 contracts across an entire chain). However:
Vehicle conclusion: common stock. No options structure clears the cost hurdle at 1.22x IV/RV and 22% spreads, on a name whose 12-month implied path already fails the Valuation Criteria. If a book wanted convexity the least-bad expression is a Jan-2027 $25/$40 call spread (OI 2,222 / 1,474, both quoted in size) — but its net debit of roughly $6.50–8.50 against a $15 maximum payoff needs the stock at $34+ just to break even, which is above the base-case 12-month target of $22.10. Stated so the reader can see the structure was tested and rejected on arithmetic, not skipped.
Inverse-volatility sizing is the framework's active protection and it does the work here without a judgement call. At 104.9% realised volatility against a book-typical 30–35% reference, the volatility-scaled weight is roughly 0.30x a normal position.
| 252-day realised volatility | 104.9% |
| Inverse-vol scalar vs a 32% reference | 0.31x |
| Indicative size on a book that would otherwise carry 4% | ~1.2% |
| Days of ADV at 1.2% of a $250m book ($3.0m) | 0.02 days — immaterial |
Liquidity does not constrain size on this name; volatility does. That is the correct ordering and it is why the Liquidity Criteria passes while the position remains small.
| Level | Value | Basis |
|---|---|---|
| Spot | $31.30 | SIP close 2026-07-29 |
| 200-day moving average | $28.55 | Spot is 9.6% above it. The last defended level in the 12-month uptrend |
| 50-day moving average | $41.67 | Spot is 24.9% below it — the shorter tape has already broken |
| July issuance price (gross, implied) | ~$55 | The company itself sold equity here 23 days ago; spot is −43% |
| Base-case 12-month target | $22.10 | OUST_Valuation.md §5 |
| Invalidation for a long | a close below $28.55 | Loses the 200-day, which is the only trend structure still intact. Below it there is no technical support until the $16.63 52-week low |
| Invalidation for the negative research view | a clean quarter of gross margin ≥46% with no royalty in it | That single print flips the Quality Criteria (OUST_Research.md §6) and removes the core objection |
Momentum note (MEASURED, timing only, blocks nothing): 12-1 momentum is +132.9%, cross-sectionally strong. But the name is −42.1% in one month and −49.9% from the 52-week high, and it is falling through a level at which the company chose to issue equity. Entering into that is buying a knife. The momentum criterion says when, not whether, and its answer is: not yet.
Named cause: ASP compression in digital lidar, with no contracted backlog to cushion it.
The mechanism is not hypothetical and the company has already written it down twice. The FY2025 10-K attributes gross-margin movement partly to "lower ASPs"; the Q1-2026 10-Q repeats it — "offset in part by lower ASPs" — in the same sentence in which gross margin fell to 42.9%. Chinese volume suppliers price into the industrial and automotive lidar channel; China rose from under 10% to 15% of Ouster's own revenue in FY2025. Ouster has no disclosed backlog ("backlog" appears zero times in the FY2025 10-K), so there is no contracted revenue to slow a demand or price shock. Add a 50%-of-accounts-payable single supplier and one customer at 42% of receivables.
| Scenario | Probability | Outcome | Cause |
|---|---|---|---|
| Bear — permanent impairment | 30% | $12.20 (−61%) | Gross margin drifts to the high-30s on ASP; growth decelerates to the low 20s; the multiple reverts to the 30th percentile of its own history (2.7x). No solvency event — $355m of net cash against a $42m burn — so the loss is a valuation impairment, not a wipeout |
| Base | 45% | $22.10 (−29%) | Multiple reverts to the most recent observation (5.4x) on ~$225m NTM revenue |
| Bull | 25% | $38.75 (+24%) | Clean gross margin recovers above 46%, software attach becomes visible in margin, multiple holds the top of its own range (9.6x) |
Going-concern case: explicitly argued and rejected. Net cash of ~$355m against TTM operating cash outflow of $42.4m is more than eight years of runway at the current rate, before the undrawn ATM and an unlimited S-3ASR. Ouster cannot be forced to raise. The permanent-loss case is a de-rating, not an insolvency.
The 30% bear probability is a judgement input, not an output. Per criteria.md the Downside Criteria is
MEASURED precisely because these estimates are not yet trusted; it is logged to be Brier-scored, and it
constrains nothing here.
Stated plainly, because the arithmetic in OUST_Valuation.md can obscure it:
Ouster is a ~32%-organic-growth, 42.5%-clean-gross-margin (rising ~4pp/yr) sensor company with no backlog, no operating profit in its history, $355m of net cash, a doubled authorised share count and an unlimited shelf, trading at the 90th percentile of its own five-year EV/sales range, 43% below the price at which it sold stock three weeks ago.
The Quality Criteria fails on a flat gross-margin level; the Valuation Criteria fails by 15.6pp; Liquidity passes; momentum says wait. Two of three BINDING Criteria fail. The book decides.