← Research/5/20/2026
Nefarious Trading
Deep Dive · discord.gg/nfrs
Vol. 01 · No. 12
April 19, 2026
OUST $24.26 ▲ HSAI $25.75 ▲ AEVA $15.60 ▲ INVZ $1.18 ▼ LIDR $1.80 ▼ MVIS $0.66 ▼ ARBE $0.67 ▼ LAZRQ $0.06 BANKRUPT OUST $24.26 ▲ HSAI $25.75 ▲ AEVA $15.60 ▲ INVZ $1.18 ▼ LIDR $1.80 ▼ MVIS $0.66 ▼ ARBE $0.67 ▼ LAZRQ $0.06 BANKRUPT
$OUST · Deep Dive · Small-Cap Physical AI

Ouster just hit GAAP profitability while its biggest competitor went bankrupt.

52% revenue growth. 60% Q4 gross margins. $211M cash. Zero debt. Analyst target at $39 — 62% upside from $24.26. The market still thinks this is a lidar company. It's not.

Price
$24.26
Mkt Cap $1.52B
FY25 Revenue
$169M
+52% YoY
Analyst Target
$39.25
+62% upside
Cash / Debt
$211M / $0
Fortress balance sheet
§ 01 — Core Investment Thesis

Everyone else died or sold. Ouster is the last one standing — and it just turned profitable.

Ouster is a picks-and-shovels bet on Physical AI — robotics, smart cities, industrial automation — not autonomous cars. Luminar filed Chapter 11 in December 2025 and sold its lidar business to MicroVision for $33M. Ouster just posted 12 straight quarters of product growth, 60% Q4 gross margin, GAAP net income of $3.98M, $211M cash, zero debt. Trades at 7.7x sales while compounding 30-50% with software bookings doubling YoY. The whole Western lidar industry consolidated around them, and the market hasn't priced it in.

The TL;DR: Only profitable-trending, debt-free, digitally-architected lidar pure-play in the West. Diversified away from automotive years ago, absorbed Velodyne's customer base, and bought Stereolabs to own the AI vision stack. I'm buying.
→ Thesis 01
Silicon, not mechanics
Digital lidar scales with Moore's Law. Doubling channels costs nothing. Analog and MEMS competitors pay for every extra laser.
→ Thesis 02
Software is the margin
Gemini and BlueCity bookings doubled in 2025. Deployed at 1,200+ sites. 80%+ gross margin SaaS attached to every sensor.
→ Thesis 03
The field cleared itself
Luminar bankrupt. Velodyne absorbed. Cepton acquired. Quanergy dead. Ouster is the last Western pure-play standing.
§ 02 — The Business

Four verticals. Two chips. One unified AI perception stack.

Ouster makes digital lidar sensors and AI perception software. Revenue splits across four verticals: automotive, industrial, robotics, smart infrastructure. In 2025, smart infrastructure and robotics drove growth — not automotive.

The Product Stack

Hardware: The OS series (spinning sensors, 360° field of view) dominates industrial and robotics, powered by the proprietary L3 chip — a backside-illuminated SPAD detector counting individual photons. The DF series (solid-state, no moving parts) is the automotive play, powered by the automotive-grade Chronos chip. Post-Stereolabs (February 2026), they also sell ZED AI vision cameras with native NVIDIA Isaac Sim integration.

Software: Gemini handles 3D perception — classifying vehicles, pedestrians, anomalies for security and logistics. BlueCity is the smart city / traffic signal solution. Combined, deployed across 1,200+ sites globally. This is where the 80%+ gross margins come from.

IP licensing: 170+ granted patents. In 2025 they collected ~$23M in primarily one-time royalties after enforcing patents against Hesai — and the U.S. Federal Circuit dismissed Hesai's invalidity appeal, preserving the North American moat.

The key insight: Ouster isn't competing with robotaxis. They're selling to warehouse operators, city traffic departments, port automation, and robotics companies like Serve Robotics and May Mobility. Boring, diversified, sticky — and profitable in Q4.
§ 03 — The Supply Chain

They don't own the factories. They own everything that matters.

Ouster is fabless — same structure as NVIDIA, AMD, and Apple. They design every critical component in-house (chips, VCSELs, SPADs, optics), then outsource the capital-intensive manufacturing. This isn't a weakness. It's the entire reason the margins work.

Where Ouster Is Vertically Integrated (The Design Stack)

Everything that creates the moat is designed and owned in-house:

  • L3 chip (custom SoC): Counts 1 trillion photons per second. All signal processing runs on the silicon itself.
  • Chronos chip: Automotive-grade variant for the DF series. ASIL-B and AEC-Q100 certified.
  • VCSEL laser arrays: 128 lasers packed into an area the size of a grain of rice.
  • SPAD detector arrays: Integrated directly into the CMOS die.
  • Micro-optics: Patented optical path. Core IP.
  • Gemini & BlueCity software: Perception models trained on 4M+ real-world objects from 800+ sites.

Where They Outsource (And Why That's Smart)

Ouster's Fabless Supply Chain

Stage 01 · In-House
Design
SF HQ. Chip architecture, VCSEL/SPAD design, firmware, AI models. 170+ patents.
Stage 02 · Foundry
Silicon Fab
STMicroelectronics fabricates the L3 SPAD chip on standard CMOS. No custom fab needed.
Stage 03 · Contract Mfr
Assembly
Benchmark Electronics, Thailand. Automotive-grade (IATF 16949). Thousands of units/month.

Why This Drives Costs Down

This model is why GAAP gross margin jumped from 36% to 49% in one year, and Q4 2025 hit 60%. Direct comparison:

Analog Lidar (Luminar, legacy Velodyne)

  • Hundreds of discrete components per sensor
  • Each laser added = another physical chip
  • Manual optical alignment required
  • Mechanical parts = higher failure rates
  • Cost scales linearly with resolution
  • Capex-heavy — own fabs or custom lines

Digital Lidar (Ouster)

  • Single SoC replaces hundreds of parts
  • More channels = denser chip, same die cost
  • No manual alignment — printed in silicon
  • No moving parts in DF series
  • Cost scales with Moore's Law
  • Capex-light — foundries do the heavy lifting

What This Means For The Multiple

A fabless semiconductor business is structurally worth more than a hardware manufacturer. No factory capex. No inventory obsolescence. No billions needed to build a fab. NVIDIA trades at a premium for this exact reason.

The working capital efficiency proves it: asset turnover expanded from 0.16x in 2022 to 0.48x by end of 2025. DSO dropped from 175 days to 76 days. Cash Conversion Cycle compressed from 273 to 91 days. They're compounding efficiency while competitors bleed capex.

The real vertical integration is in design, not manufacturing. Ouster owns the technical stack — chips, optics, firmware, AI — and lets specialized partners do the metal-bending. That's why margins keep expanding while Luminar went bankrupt trying to run its own fab. Same playbook as NVIDIA. Different scale.
§ 04 — The Numbers

12 straight quarters of product growth. 60% Q4 gross margin. Not a story stock anymore.

MetricFY 2024FY 2025Δ YoY
Total Revenue$111.1M$169.4M+52%
Product Revenue~$111M$147M+32%
Sensor Units Shipped~15,00025,000++67%
Q4 2025 Revenue$30M$62M+107%
GAAP Gross Margin (FY)36%49%+1,300 bps
Non-GAAP Gross Margin (FY)42%54%+1,200 bps
Q4 GAAP Gross Margin44%60%Record
Q4 GAAP Margin ex-Royalties~41%Core business
GAAP Net Loss−$97.0M−$60.4M+$37M better
Q4 2025 GAAP Net Income+$3.98MFirst profit
Adjusted EBITDA (FY)−$41.8M−$12.4M+$29M better
Cash + Investments~$170M$211MFortress
Long-term Debt$0$0Zero
Product Book-to-Bill1.2xBacklog growing

Q4 2025 GAAP profit was inflated by ~$21M in primarily one-time IP royalties — roughly 20 percentage points of quarterly gross margin. Strip those out and core hardware margin was still 41% in Q4, up from 36% full-year 2024. Real operating leverage.

Management's long-term framework: 30-50% annual revenue growth, 35-40% sustained GAAP gross margins ex-royalties, opex growth held to 5-8%. Q1 2026 guidance: $45-48M (including partial Stereolabs).

§ 05 — Valuation

At 7.7x sales, the market is pricing Ouster as hardware. The software makes this a steal.

Forget P/E — net income is still negative for the full year. The right multiple is EV/Revenue. Market cap $1.52B, subtract $211M cash and zero debt = enterprise value ~$1.31B. On $169M of 2025 revenue, trailing EV/Revenue is ~7.7x. On 2026 consensus (~$237M), forward EV/Revenue compresses to ~5.5x.

Reverse DCF — What Does The Price Imply?

To justify $1.52B at a 10% discount rate and 3% terminal growth, Ouster needs to cross into positive free cash flow by ~2028 and ramp to ~$150M annual FCF by 2035. Required trajectory:

2025 (actual)
−$54M
2026E
−$30M
2027E
−$10M
2028E
+$20M
2029E
+$45M
2032E
+$100M
2035E
+$150M

Is that realistic? Grow the 2025 base at 30% (the floor of guidance) for five years = $625M by 2030. At 35% GAAP gross margin and 5-8% opex growth, that's $250M+ gross profit with real operating leverage. Software attach doing what management claims gets them there faster than the 10-year implied horizon.

Analyst consensus: 4 analysts rate OUST Strong Buy. Average 12-month target $39.25, high $50. That's 62-106% upside from $24.26.

§ 06 — Scorecard

What could go right. What could go wrong.

Bull Case

  • Field cleared. Luminar bankrupt. Velodyne absorbed. Cepton acquired. Last Western pure-play standing.
  • Software attach doubling YoY. 1,200+ sites deployed. SaaS margins structurally lift consolidated gross margin.
  • Fabless cost advantage. STMicro does silicon, Benchmark does assembly. Ouster keeps the IP and captures the margin.
  • GDPR tailwind. Lidar strips PII automatically. European smart cities can't use cameras — they have to use lidar.
  • NVIDIA integration. Native Isaac Sim / Metropolis support. Center of the Physical AI developer stack.
  • Fortress balance sheet. $211M cash, zero debt, 3.92x current ratio. Can weather any Hesai price war.
  • Patent moat validated. U.S. Federal Circuit dismissed Hesai's invalidity appeal.

Bear Case

  • Hesai is 16x bigger. $2.75B vs. $169M — and Hesai is profitable. Scale matters.
  • Q4 margins were inflated. $21M in one-time royalties drove the 60% gross margin. Core hardware closer to 41%.
  • Still burning cash. FCF negative (~$54M TTM). ATM equity offerings create dilution risk.
  • Foundry dependency. Fabless means reliance on STMicro and Benchmark. Any disruption is a supply shock.
  • High volatility. Beta ~3. 52-wk range $6.58 to $41.65. Wild swings both directions.
  • Automotive keeps slipping. DF series was supposed to drive revenue. Level 3-5 adoption remains slow.
  • Chinese price pressure. Hesai doubling to 4M units in 2026. Commoditization could squeeze even digital margins.
§ 07 — The Competitive Landscape

Every other pure-play lidar stock. Ranked by survival probability.

The honest state of the Western + Chinese lidar competitive set, April 2026. One bankrupt. One absorbed. Four micro-caps on life support. Only two viable at scale.

Ticker Company Price Mkt Cap FY25 Rev Growth Profitable? Focus Status
OUST Ouster $24.26 $1.52B $169M +52% Q4 yes Diversified / Physical AI Top Pick
HSAI Hesai Group $25.75 $4.0B $2.75B +32% Yes (+$430M) Chinese auto / ADAS Market leader
AEVA Aeva Technologies $15.60 $989M $18M +99% No (−$145M) FMCW 4D / NVIDIA Drive Speculative
INVZ Innoviz Technologies $1.18 $246M <$30M Slow No Automotive (BMW, Daimler) At risk
MVIS MicroVision (+ Luminar IP) $0.66 $290M Negligible No (distressed) ADAS + Luminar Iris/Halo High risk
LIDR AEye $1.80 ~$50M Negligible No Software-defined / ADAS Micro-cap
ARBE Arbe Robotics $0.67 $86M $1M +34% No 4D radar (not lidar) Pre-revenue
LAZRQ Luminar Technologies $0.06 $6M No Automotive (Volvo terminated) CH. 11 · LIQUIDATED

Private / Acquired Players

RoboSense (2498.HK): Hong Kong-listed Chinese MEMS competitor. Strong in China robotaxi/robotics. NDAA exposure limits U.S. access.

Cepton: Acquired by Magna International (MGA). Now a Tier-1 supplier unit, not a pure-play.

Seyond, Valeo, Continental, Bosch: Private, or lidar is a small division of a massive auto supplier. Theoretical long-term threats — none ships a competitive digital lidar today.

$PONY, $VLN: Pony.ai is an autonomy developer (customer-side). Valens is automotive connectivity chips (adjacent). Neither is a direct lidar competitor.

The table says it all: Outside Hesai (NDAA-constrained in North America) and Ouster, there is no other viable Western lidar pure-play. Luminar is being liquidated. INVZ, LIDR, MVIS, ARBE are all sub-$2 with market caps under $300M. AEVA is the one speculative alternative — real tech, NVIDIA partnership, but priced on a dream at $18M revenue / $989M cap.

Head-to-Head: OUST vs AEVA vs HSAI

Three real choices in 2026.

HSAI — Growth at a reasonable price. Profitable, dominant in China, ~32x forward P/E. Risk: geopolitics could lock them out of the West permanently.

AEVA — High-beta bet on FMCW and NVIDIA Drive Hyperion. Real differentiated tech. But ~55x sales on $18M revenue. 3-5x if OEM adoption accelerates, dead money if it doesn't.

OUST — Best risk-adjusted. Four verticals, software attached, fortress balance sheet, near-breakeven, 7.7x sales on 9x AEVA's revenue.

§ 08 — My Take

The market thinks Ouster is a lidar company. It's not.

Ouster is a fabless semiconductor company that happens to sell sensors. Semiconductors scale with Moore's Law — adding resolution costs nothing. Analog and MEMS competitors pay for every extra channel. That's why Hesai is doubling production and Ouster doesn't need to. Scaling problem is design-time, not factory-time.

Stereolabs changes the pitch. No longer "lidar vs. cameras vs. radar" — it's "Ouster sells you the unified sensor fusion stack on NVIDIA's robotics platform." Every robotics developer starting a project in 2026 is a potential customer.

The numbers point to sustained GAAP profitability in late 2026 / early 2027, real free cash flow by 2028, and 30-50% revenue compounding for three more years. At $24.26, the market prices the hardware — not the software, not the IP licensing, not the fabless margin structure, not the fact that every competitor died or got bought.

I'm buying OUST between $22-$26.

Target: $39.25 (+62%). Playing for the midpoint over 12-18 months. Thesis breaks if (1) software attach stalls, (2) Hesai finds a way around NDAA, or (3) ATM dilution hits meaningfully before breakeven. Adding on any pullback below $20.

"This is the one Western lidar stock that didn't die — and now it has the whole field to itself."
§ 09 — X Sentiment Tracker (April 12-19, 2026)

What the street is saying. Overwhelmingly bullish.

Sentiment analysis of high-engagement $OUST posts on X over the past week. Weighted toward posts with highest bookmark counts — bookmarks signal lasting value, not just reaction.

88/100
Strongly Bullish
POSITIVE 88%
9%
3%
Positive (88%) — conviction buyers, thematic bulls Neutral (9%) — charts, partner updates Cautious (3%) — competitor / dilution questions

Top Engagement Posts This Week

@BlackPantherCap · Deep Dive 17+ bookmarks · 1000s of views
"$OUST is building the eyes every AI machine needs."
Comprehensive thesis framing Ouster as the foundational sensing and perception platform for Physical AI — not just a lidar company. Drove the weekend engagement spike. Widely quoted and praised.
@MMMTwealth · Repost 112+ bookmarks · 20k+ views
"$1.5B company, ~$900M projected revenue in 4 years, only ~0.2% TAM penetration."
March prediction reposted amid the recent buzz. Calls the 5+ year story "extremely strong" at ~5.8x sales for 30-50% growth. Reinforces long-term conviction.
@TheRonnieVShow · Technical 39 bookmarks · 25k+ views
"Bull trigger setup. Targets $30 → $35 → $43."
Weekly MACD golden cross called out. Chart-focused post describing a breakout setup above key moving averages heading into Q1 earnings (May 5).
@YoYInvestor · Position Initiated 45 bookmarks · high views
"Ouster is what allows these robots to see. Long $OUST."
Added to concentrated portfolio. Cited StereoLabs deal, clean balance sheet, and growth metrics. Called the @BlackPantherCap thesis "the most robust and comprehensive on $OUST."
@EndicottInvests · Client Tracker 22-30 bookmarks · 4-8k views
"Komatsu multimillion deal, John Deere, Motional, Serve Robotics / Uber Eats. Zero debt."
Lists real-world traction and partner wins. Community-building tone. Notes $LAZR bankruptcy and $HSAI Chinese exposure as competitive clearing.
@itschrisray · Thematic 13 bookmarks · 1k+ views
"$OUST is a 1 of 1 leader for an industry about to take off. Like $RKLB and $ONDS."
Frames Ouster as the lidar equivalent of Rocket Lab (space) and Ondas (drones) — the dominant small-cap pure-play in an emerging sector.

Recurring Bull Themes in High-Engagement Posts

  • StereoLabs as game-changer — unifies digital lidar + stereo cameras + AI compute + fusion software into one platform. Doubled TAM. Only integrated offering on the market.
  • "Eyes for every AI machine" narrative tied to Jensen Huang's robotics comments. Applications span warehouses, ports, construction, drones, robotaxis, humanoids, agriculture, defense.
  • SAFE LiDAR Act and broader NDAA/DOT restrictions on Chinese lidar seen as structurally favoring U.S. players like Ouster. Policy-driven moat in defense and critical infrastructure.
  • Honor robot half-marathon story (Chinese firm using Ouster lidar) sparked "even Chinese manufacturers prefer $OUST over Hesai" reactions.
  • Execution track record — 12 consecutive quarters of product growth; 48% sensor shipment growth in 2025; ~$211M cash, zero debt.
  • Valuation attractive at ~5-6x NTM sales for a high-growth name with only ~0.2% long-term TAM penetration.

What This Tells Me

Sentiment is catalyst-driven, not momentum-chasing. The @BlackPantherCap deep dive acted as a unifying resource the community saved and shared — that's the kind of activity that builds a holder base, not a pump crowd. Bear commentary is minimal and factual (competitor questions, not "scam" rhetoric). Cross-tagging with $RKLB, $ASTS, $ONDS suggests the autonomy/defense/space community is adopting $OUST as a core holding.

The risk of bullish sentiment: When everyone agrees, positioning gets crowded and any disappointment gets punished. Q1 earnings are May 5. If the company beats and guides in line with the 30-50% framework, this continues. If core hardware margins disappoint ex-royalties, the stock sees a sharp correction regardless of how good the thesis is.