Joe Chiang 江効儒 (@jojoali930) profile picture

FB : https://nitter.gq/t.co/l5j4IrSZ2f Blog : https://nitter.gq/t.co/Uq6XPq9w5V VFX lead / filmmaker / writer / investor / crypto

Vancouver
Joined April 2009
Photo by Joe Chiang 江効儒 (@jojoali930)
Joe Chiang 江効儒 retweeted
Mornings. I’ve been revisiting my investment thesis on $AAOI following yesterday's announcement. (LONG WRITE-UP HERE) And honestly, my view of management’s capital allocation is changing. Let’s look at the bigger picture. 1. The latest dilution On November 17, AAOI announced a new $550M at-the-market (ATM) equity offering. This comes shortly after completing its previous $600M ATM programme, which generated approximately $588M in net proceeds. To be clear, the new $550M is an authorization. They don’t have to sell all the shares immediately. But the timing raises questions. At the November 16 closing price of $45.16, a fully utilized $550M programme would represent approximately 12.2M additional shares. That’s roughly 13.5% more shares compared with the illustrative existing share count in the prospectus. And this is where I think shareholders need to pay attention. 2. The expansion is enormous AAOI isn’t just buying a few machines and expanding production. They’re building an entirely larger manufacturing operation to capture the demand for high-speed optical connectivity. Some important numbers: • $928.8M in outstanding equipment and facility construction commitments as of September 30. • $200M new credit facility announced October 30. • Up to $194.4M in annual raw material purchases under a new supply agreement. • Up to $40M in required supplier prepayments. • $102.25M spent acquiring three buildings at Hightower Business Park in Houston. That’s a substantial amount of capital being committed. And it explains why management wants access to additional financing. 3. What about buying the land instead of leasing? Interestingly, I actually like this decision. AAOI initially leased the three Houston buildings with purchase options and subsequently exercised those options. I see the strategic logic. Manufacturing facilities are long-term infrastructure assets. Owning them provides greater control over expansion, modifications and future occupancy costs. And unlike rent, the purchase creates an asset on the balance sheet. But buying the buildings doesn’t automatically create additional production capacity. The return depends on what they install inside them, how quickly they can bring the facilities online and how much profitable output they generate. The property isn’t the problem. The total capital allocation strategy is what deserves scrutiny. 4. Management: The good and the bad I don’t want to overlook what management has achieved. Q3 2026: • Revenue: $269.9M, up 127% YoY. • Datacenter revenue: $155.6M. • GAAP net income: $6.3M. • Six consecutive quarters of record revenue. That’s real operational progress. But there are some warning signs. Gross margins haven’t expanded alongside revenue. Q3 GAAP gross margin was 28.2%. The company has also acknowledged construction and power-related delays in its US expansion. And despite the tremendous revenue growth, operating cash flow remains under pressure. I want to see the business become increasingly self-financing. Not increasingly dependent on issuing shares. 5. So why am I becoming more critical of management? Because I think there’s a difference between investing ahead of demand and committing so much capital that shareholders continuously have to finance the execution. Management has raised substantial equity capital this year. They’ve secured manufacturing facilities, ordered equipment, entered into major supply agreements and expanded their capacity ambitions. I understand the strategy. The AI infrastructure opportunity is enormous, and being underprepared when customers need capacity could be far more expensive than expanding early. But what happens if construction takes longer than expected? What happens if demand shifts, customer orders are delayed or margins remain compressed? Shareholders absorb the dilution while management still has to deliver on the investments. That’s the risk. 6. What I want to see next Five things matter more to me now: > Revenue growth translating into sustainable operating profitability. > Gross margin expansion as manufacturing scales. > Positive and improving operating cash flow. > Evidence that new capacity is backed by actual customer demand. > Earnings per share growing faster than the share count. The last point is particularly important. A company can double revenue and still deliver disappointing shareholder returns if the share count keeps increasing. My conclusion: I’m not abandoning the AAOI thesis because of another ATM. The bottlenec and demand for photonics are there. I still think they can benefit. The company operates in an attractive part of the AI infrastructure buildout. Its optical technology, manufacturing ambitions and customer demand remain interesting. But my assessment of management has changed. Previously, I viewed the equity raises as opportunistic capital allocation, taking advantage of a strong share price to finance future growth. After another $550M authorization, I find it harder to defend that explanation without asking tougher questions about financial discipline. I still understand why they’re expanding. I just want management to demonstrate that they’re building a business capable of generating attractive returns on all this invested capital. Building capacity is one thing. Building shareholder value is another. And at this stage, AAOI still has something to prove on the latter. Curious how other $AAOI shareholders see this. @JonahLupton @FinnStockinger @outliercapx @TheBigBerbowski Furthermore I'm interested in getting in touch with anyone who has some sector knowledge here. You might work as specialist in the industry. If so, please comment, i'm very interested in your perspective on demand e.g. I beleive at this point AAOI is starting to loose retail. Loose retail, and they'll practical loose their strategic funding as of now apparently. Imo, it would be beneficial for AAOI at this point to be proactive in their communication. Finally, where do I stand: Well as many, in generally, AAOI has fallen a lot during this AI drawdown. I consider my entry to be modest comparing to whats going on in the market. I'm obviously underwater at this point also. For now I'll sit tight (this is due to demand and the sector) - my position is LUCKILY one of the smallest in my portfolio (before and after this). Trust has to be re-stored here, and untill then I'll be looking for opportunities to leave and exit my position. -BP Not financial advice.
42
21
5
317
66,832
Joe Chiang 江効儒 retweeted
Everyone is understandably frustrated watching $AAOI file a $550M ATM shelf 12 days after CFO Stefan Murry said on the Q3 call that they planned to rely more on debt and internal cash flow. Management deserves the hit to their credibility for how they telegraphed this. Once you get past the communication fumble and open the Q3 10-Q, though, the numbers explain what is actually happening behind the scenes: -The Unfunded Commitments in Note 18: The headline Q3 release showed $621.5M in cash and a new $200M bank loan ($821.5M total liquidity). What most people missed is Note 18 of the 10-Q, which shows $AAOI is carrying $928.8M in binding equipment and facility purchase commitments for Sugar Land Phase 2 and Taiwan. They were already over $100M short of their committed CapEx bill before funding a single dollar of new working capital. -Funding a $5B+ LTA Without Customer Prepayments: Unlike $LITE, which received upfront customer prepayments to build capacity,$AAOI's first $1B to $1.5B 1.6T contract included zero upfront capital from the hyperscaler. On the Q3 call, Thompson Lin stated a second LTA is in the final stages, weeks away, and at least 5x larger than the first ($5B to $7.5B). Scaling toward $471M/month in revenue by late 2027 means $AAOI has to prepay 1.6T DSP suppliers and front hundreds of millions in inventory and receivables on its own balance sheet while waiting 60 to 90 days for hyperscaler invoices to clear. -Passing the Counterparty Audit & Selling Into Strength: No Tier 1 hyperscaler risk committee is signing off on a $5B+ contract with a $4B market cap supplier unless that supplier can prove it has the liquidity to survive the working capital ramp. Filing the ATM right after the 10-Q gives $AAOI ~$1.37B in total liquidity capacity to check that box for the customer. It also puts the legal shelf in place now so Raymond James and Needham can quietly sell shares into the massive volume spike after the 5x LTA hits the tape (just like they executed the prior $600M ATM at a $105.36 average price), rather than killing the rally by filing for dilution the day after the announcement. -The Dilution Runway Through Late 2027: With $1.37B in combined cash, debt, and ATM capacity, the entire $928.8M CapEx commitment is covered with over $440M left for working capital, and Q4 Non-GAAP net income is guided to jump to $46M to $60M in a single quarter. Unless they land yet another massive LTA on top of this upcoming one that requires expanding capacity beyond 1.2M units/month, this should be their last ATM through the end of 2027. The ATM creates near-term overhead supply and management handled the messaging poorly. Even so, factoring in 10% to 12% dilution and a 50% execution haircut to their late 2027 run rate, $AAOI is trading around 10x buffered 2027 earnings with a $5B+ contract weeks away. Selling in the low $40s right before Q4 earnings inflect misses what this balance sheet is being built to support. I’m holding.
38
21
3
242
34,368
Joe Chiang 江効儒 retweeted
The governor-elect of Texas, Gina Hinojosa, has explicitly said that all currently permitted datacenters in Texas will be allowed to finish construction. This essentially covers all of the 2027 and likely well into 2028. She is going to pass laws that ensure that all *newly* permitted datacenters going forward for delivery in 2028/2029 will pay taxes, commit to no increase in power prices for consumers, don’t impact water availability for people, be constructed with union labor and give “Texans a seat at the table.” Sounds reasonable. She doesn’t want to lose a lot of good blue collar jobs and property taxes for small towns in rural Texas. Datacenter commentary starts at 6:30 in this interview. Also attaching her official campaign position from her website in the next post. Sidenote: After watching a bunch of her speeches and interviews, I do think she is a political talent and probably has a future in national politics. youtu.be/D-T6H3b8ZeE?is=mEWu…
1,629
2,202
1,045
17,437
18,439,303
Joe Chiang 江効儒 retweeted
trade[XYZ] and ICE, owner of the New York Stock Exchange, are joining forces to bring 24/7 perpetual futures on Hyperliquid to the United States. In just over a year, our markets on Hyperliquid have surpassed $650 billion in volume. They provide continuous price discovery for stocks, indices, and commodities, and have become trusted reference points for retail and institutions alike when traditional markets are closed. Much of the world trades them. Until now, Americans could only watch. We intend to change that. Subject to CFTC approval, trade[XYZ] and ICE plan to establish a new, dedicated U.S. market, with novel exchange and clearing infrastructure built on Hyperliquid. Planned contracts span equities, indices, commodities, FX, and crypto, with access through participating brokers and platforms. We’ll share more on products, eligibility, and timing as approvals progress. ICE brings more than 25 years of experience operating regulated markets, clearing, and managing risk. trade[XYZ] brings the product development, market design, and technology behind the most liquid and robust markets on Hyperliquid. Innovation and the protections of U.S. regulation shouldn’t be a trade-off. Together, we plan to prove it. ICE is making a strategic investment in trade[XYZ] and joins us in building towards the future of Hyperliquid. A small group of market luminaries is joining as well, beginning with Stan Druckenmiller. For more than four decades, few investors have anticipated the great turns in markets as often as Stan. We’re proud to have him with us for the next one. We’re building the U.S. market in parallel with our existing HIP-3 markets, which we’ll continue to expand and innovate on. We expect both to flourish. They each have a long road map ahead, from new asset classes to new ways to trade them. To the Hyperliquid community, whether you were here before us or came because of us, thank you. You took a chance on new markets and you made them matter. You’ll forever be a part of this story. Anything. Anytime. On ICE[XYZ].
630
1,060
803
6,277
1,563,900
Joe Chiang 江効儒 retweeted
$SKHY ABSOLUTELY CRUSGED THEIR Q3 EARNINGS • Revenue: $78.9B vs. Est. $74.8B • Operating profit: $62.8B vs. Est. $58.5B • Gross Margin: 86% vs. Est. 85% • DRAM ASP: +28% QoQ • NAND ASP: +25% QoQ • Cash: $86.3B Q4 Guidance • DRAM shipments: +25% QoQ • NAND shipments: +5% QoQ FY27 Guidance • DRAM demand: +25% YoY • NAND demand: +18% YoY Signed 5-year LTAs with 14 major customers securing over 50% of its 2027 and 2028 capacity.
59
206
36
1,872
226,704
Joe Chiang 江効儒 retweeted
NEAR is in fact booming and the data shows it
21
48
13
396
41,753
Joe Chiang 江効儒 retweeted
Replying to @ManInTheeArena
Yeah it's wild. It is most likely going to do this all the way to $100K+. Once it breaks the 0.75 then it typically rips to ATHs quickly. This is how it looked last cycle. We followed the whole ride up from the cycle low to ATHs with fib fans.
1
1
36
2,270
Joe Chiang 江効儒 retweeted
#Bitcoin This is why chartists chart, even if they don't trade often.
42
70
7
886
45,674
Joe Chiang 江効儒 retweeted
Rotation is coming… BOOKMARK IT.
45
47
1
691
121,801
Joe Chiang 江効儒 retweeted
Aave quarterly call tldr : 1- Coinbase Tokenized Stocks on Aave V4 Coinbase Tokenized Stocks (AAPLc, AMZNc, GOOGLc, METAc, MSFTc, NVDAc, TSLAc) live on Aave V4’s Equities Hub on Base. Users deposit them as collateral and borrow USDC. Chainlink provides official oracle data. Available only to eligible non-US users (thanks @jessepollak and @base for making this happen) - $9.6M deposits - $4.1M USDC borrowed - Initial aggregate collateral cap ~$29M Steady initial steps towards equities as collateral 2- Q4 Direction More stocks expected over time, ethereum:0x40d16fc0246ad3160ccc09b8d0d3a2cd28ae6c2f as a borrowable asset, and supply/borrow caps revisited as liquidity deepens 3- V4 Architecture + Institutional / Curator Growth - @SentoraHQ Hub (largest vault curator, ~$2.28B AUM / ~23% market share) is launching its own externally curated Hub on Aave V4 on Ethereum, Starts with stablecoin-focused markets (borrowables like RLUSD, PYUSD, OUSD, yield and bluechip spokes). Isolated from other Aave Hubs - Aave DAO keeps 50% of the Hub’s protocol revenue. Sentora handles day-to-day risk parameters, oracles, and collateral selection (with some DAO visibility/delay on certain changes) - Continued push on RWA/Horizon-style products and institutional partners. Q3 also saw ether.fi Cash as a dedicated V4 hub ($311.5M deposits) and other new markets totaling ~$1.3B deposits 4- Aave App - Early traction on the retail-facing Aave App, positioned around “bringing savings back” for a broader consumer audience. - Continued growth emphasis into Q4 as a way to onboard retail users beyond pure DeFi power users. 5 Aavenomics 3.0 - @StaniKulechov mentioned Aavenomics 3.0 will include AAVE buybacks, staking, and burn mechanisms ethereum:0x7fc66500c84a76ad7e9c93437bfc5ac33e2ddae9 buybacks are coming back 6- Financials - DAO revenue ~$18.9M - Net income ~$4.0M - Expenses down significantly (~32% excluding certain items) - Protocol earnings (mainly borrow fees) solid and rising through the quarter - Deposits ~$33.6B (+41%) - active loans ~$13.2B (+32%) - ethereum:0xe1753f2e00940cc31213dd92013cf019dfe4ca1d grew ~29% to ~167M. Strong governance execution 7- what's next - Further V4 multichain and Hub/Spoke expansion (more external curators expected) - Continued tokenized equities growth + more RWA/institutional onboarding - Aave App and retail savings push - Potential buyback restart and Aavenomics 3.0 formalization - Pipeline items mentioned in related reports: additional chains/hubs, cross-chain sGHO, more equity collateral, etc. house of all credit and equities finance thanks @StaniKulechov @luigidemeo and @0xKolten for the update aave will win
10
5
1
44
2,210
Joe Chiang 江効儒 retweeted
Do you get how extremely bullish this is for $AAOI and $SIVE? Lumentum fully booked. Stopped taking new OCS orders mid-April and “lacks concrete plans for expansion”. ~72k units / $4.8-5B in 2028 alone. Clearest signal yet that optics is the next hard bottleneck after memory. Demand so strong the leader can’t take more orders and has no clear path to add capacity. Spillover is massive: $AAOI → 1.6T transceivers + in-house lasers. Capacity ramping hard while demand already exceeds supply through mid-2027. $SIVE → pure-play InP CW DFB lasers & ELS for CPO. Scaling to 100M+ lasers/year. Exactly what the AI fabric is short of. Next-gen TPUs in 2028 are deeply optical. When the leader is oversubscribed with no expansion plan, the pure-plays capture the overflow. This is just the start. We’ve been warning about this for months.
Bullish — $LITE $COHR $AAOI $IQE $SIVE Morgan Stanley: “By 2031, CPO and NPO will become the dominant technologies in the 3.2 Tbps high-speed port market, where single-lane rates reach 400 Gbps.” Optics is moving inside the scale-up domain. As we’ve been saying, electrical interconnects are already hitting distance, power and density walls. $META lab CPO data (6× fewer interruptions, 1.35M hours clean) is the reliability proof point the market needed. NPO commercialization is called out as a catalyst for $LITE / $COHR. Optics and networking remain among the best-positioned (if not the best) sectors for the coming years. Positioning accordingly.
11
21
1
174
31,836
Joe Chiang 江効儒 retweeted
I know $NBIS price action looks bleak right now. But let me break down a quote I found from Jensen at GTC Berlin 2026. "Startup companies like Nebius is doing incredibly well." Startup companies like Nebius - Nebius is an early stage startup worthy of the CEO of a 5 trillion dollar company mentioning publicly. Early because they got so much freakin' growth ahead of them. Btw, I did not find any mention of CoreWeave or familiar neocloud. Doing incredibly well - Coming from Jensen Huang and his expectation for "ok", to "well" to "incredibly well" - it means it's so fucking exceptional S tier. That's all I need to know. Nebius will take care of you.
31
18
1
480
33,430
Joe Chiang 江効儒 retweeted
Meta $META just dropped its Q3 earnings. Meta reported $64.6 billion in revenue, beating expectations of $63.4 billion, with revenue growing 26% year over year. However, EPS came in at just $3.83 versus the $6.80 analysts were expecting, a massive 44% miss but what really caught my attention is the company's AI spending. Meta is reportedly guiding for $135 billion to $145 billion in capital expenditures for 2026, while projecting Q4 revenue of $73 billion to $76 billion. We're talking about a company potentially spending nearly $140 billion in a SINGLE year, largely to expand its AI infrastructure, including GPUs, data centers, networking equipment, and the power required to support all of it. And this is precisely why I'm still extremely bullish on the broader AI infrastructure trade. Meta is continuing to generate strong revenue growth while committing enormous amounts of capital to building out compute capacity. The biggest question now is whether all of this spending will eventually translate into meaningful earnings growth, especially as depreciation and infrastructure costs continue climbing. For now, the demand for AI infrastructure certainly doesn't appear to be slowing down.
5
14
2
98
30,927
Joe Chiang 江効儒 retweeted
TSMC $TSM Q3’26 EARNINGS HIGHLIGHTS 🔹 Revenue: $46.72B (Est. $44.93B) 🟢; +41.2% YoY 🔹 EPS: $4.60 (Est. $4.48) 🟢; per ADR 🔹 Gross Margin: 66.4% (Est. 66.3%) 🟡 🔹 Oper Margin: 58.7% (Est. 57.7%) 🟢 🔹 Net Income: NT$763.06B (Est. NT$737.9B) 🟢 🔹 Net Profit Margin: 51.1% Revenue by Platform: 🔹 HPC: 68%; +22% QoQ 🔹 Smartphone: 21%; +10% QoQ 🔹 IoT: 4%; +5% QoQ 🔹 Automotive: 4%; +23% QoQ 🔹 DCE: 1%; -12% QoQ 🔹 Others: 2%; -10% QoQ Technology Mix (Wafer Revenue): 🔹 2nm: 8% 🔹 3nm: 33% 🔹 5nm: 28% 🔹 7nm: 10% 🔹 7nm and below: 79% Other Q3 Metrics: 🔹 Wafer Shipments: 4,476 Kpcs; +9.6% YoY 🔹 North America: 79% of revenue 🔹 ROE: 45.5% 🔹 OpEx: 7.8% of revenue 🔹 Inventory Days: 82; -5 days QoQ 🔹 CapEx: $13.84B; $40.64B YTD 🔹 Free Cash Flow: NT$476.56B Key Updates: 🔸 Gross margin was 1.3 ppts lower than 2Q26, primarily due to the margin dilution from the initial ramp-up of N2 technology. 🔸 Non-operating items were a gain of NT$34.59B, vs NT$95.83B in 2Q26, which included disposal and mark-to-market gains on VIS shares. 🔸 Diluted EPS NT$29.43 (+68.7% YoY); revenue NT$1,494.24B (+50.9% YoY)
40
190
55
1,123
149,989
Joe Chiang 江効儒 retweeted
Two leaders building private compute for crypto. I read both whitepapers, token schedules and fee models so you don't have to. $ZAMA vs $ARX: → Who burns tokens, who doesn't → When the unlock walls hit → Where the upside is Full breakdown 👇
5
2
3
20
6,237
Joe Chiang 江効儒 retweeted
⊢ $ZAMA rn > 92% of circulating supply is staked > only $16M worth is liquid > $116M already shielded on @zama ∎
19
17
6
127
9,360
Joe Chiang 江効儒 retweeted
So far, USDT.D is playing out well. We saw the deviation followed by a retest, suggesting $BTC may have put in a higher low at 82.2K. Any sweeps into next week are worth watching. As long as dominance stays below 6.7%, a break below 6.56% will mean BTC legs upto 90-95K.
It is extremely important that... if we are following the 2023 fractal, $BTC now shows acceptance back within the range on the LTFs after deviating above the highs. Right now, USDT.D has also deviated above its highs, similar to what we saw in 2023, and is now pushing back up for a retest. If this retest forms a lower high, it would strongly suggest USDT.D is continuing its macro downtrend. That would support the idea that BTC is forming a higher low here before expanding towards 90K. What we do not want to see is USDT.D reclaiming those highs and holding above them with no sign of rejection. If that happens, it would suggest BTC is likely headed back towards the low 70Ks. These are the key levels to watch now in determining whether this low is truly in... or not.
49
45
4
912
93,564
Joe Chiang 江効儒 retweeted
SpaceX vs. $ASTS reminds me of Apple vs. Spotify. At the time, Apple was already a tech giant with gigantic distribution, the iPhone, iTunes, and the App Store. And Spotify was a nobody, just a Swedish startup with a big vision and every incentive to make music streaming work. You can imagine early backers asking: “What happens to our investment if Apple cuts you off the App Store?” Never bet against Apple, right? Then Apple launched Apple Music in 2015. A real thesis breaker. How the hell does a startup compete with that? Had Spotify been publicly traded at that point, it’s easy to imagine early investors chickening out, turning into doomers overnight, and telling everyone it was going to zero. But Spotify fought back. It accused Apple of blocking app updates and using App Store rules to disadvantage competing music services. In 2019, it filed an antitrust complaint with the European Commission and turned the dispute into a very public PR battle. Through all of this, Spotify kept building, improving its product, and growing its subscriber base. Today, it has 300M paying subscribers and is the world’s largest music streaming subscription service. Apple today is an even bigger giant than a decade before, and about 50x larger than Spotify. But in music streaming, Spotify won. Yesterday, the market reacted as if SpaceX buying spectrum was a thesis killer for $ASTS. Add to that the fear that rocket companies could completely deprioritize launching AST’s satellites, and you can almost hear the same question: “How the hell do you compete with that?” In fairness, depending on competitors, Blue Origin included, for launches is a real disadvantage. But Spotify also depended on infrastructure controlled by a giant competing directly against it, and that didn’t prevent it from winning. I’m not saying AST will or won’t make it. It carries real risks and also offers a major opportunity, as many early-stage businesses have before it. How to weigh those two is up to each investor. All I’m doing is rejecting the binary thinking that there can be only one winner in satellite connectivity, that you must own the rockets to build a constellation, and that the larger company must always win in every application layer it sets its sights on. History shows it is a little more nuanced than that.
15
26
3
312
20,242
Joe Chiang 江効儒 retweeted
so you’re telling me $ZRO did 91.6% of the volume across these 6 protocols in 30 days.. while Chainlink, Wormhole, Hyperlane, Socket and Axelar shared the remaining 8.4%? and $ZRO is still roughly 75% below its all-time high? damn.
38
31
10
347
51,433
Joe Chiang 江効儒 retweeted
$QNT is to The Clearing House what $ZBCN is to FedNow. Quant got picked to move tokenized deposits across RTP and CHIPS. The bank-to-bank rails. Zebec built payroll on ACH and FedNow. The rails that reach your paycheck. Same bet. Programmable money. Different floor of the same building. Quant's customer is the banking system. Zebec's customer is your employer. Quant builds the highway between banks. Zebec is the truck that delivers to your door. $QNT: $2.2B $ZBCN: $215M Not rivals. Adjacent. Position accordingly NFA.
8
33
171
4,284