Hook
49.4%. That’s the drawdown on Serenity’s portfolio. Not a flash crash. Not a rug pull. A slow bleed from peak to trough on three tickers: AXTI, SIVE, AAOI. The chart didn’t lie—49.4% is a real loss, not a retracement. The trader with 26 years of 4502% cumulative returns now sits on a 23x multiple of his initial capital, down from 46x. He calls it profit retracement. I call it a margin call waiting to happen.
I bought the pixel, not the promise. Nine years ago, I liquidated 60% of my DeFi positions during the DAO hack because I verified the code myself. Here, I see no code—only narratives pinned to 2027 revenue inflections. Let’s dissect the order flow.
Context
The pseudonymous trader “Serenity” went viral on Crypto Twitter this week after revealing a 49.4% peak-to-trough drawdown on a concentrated portfolio of AI small-cap equities. His thesis: invest in “AI bottleneck areas”—companies providing critical infrastructure components like optical modules (AAOI), substrate materials (AXTI), and something called SIVE (a fabless semiconductor play). He believes these bottlenecks will resolve into a revenue inflection point in H2 2027, turning his current losses into future gains.
His track record is legitimate: 26 years of consistent outperformance, likely from early bets on Bitcoin, DeFi, and now AI. But the numbers don’t care about pedigree. At the peak, his capital multiplied 46x from initial investment. Now it’s at 23x. That’s a 50% loss of his portfolio’s absolute value—not a profit retracement. Every candle tells a story of fear. This candle says: “Smart money is rotating, and small caps are at the back of the bus.”
Core
Let’s get technical. AAOI (Applied Optoelectronics) makes optical transceivers for data centers. AXTI (AXT Inc.) produces gallium arsenide substrates. SIVE (SiVi) does something in the AI chip space—vague enough to be a red flag. All three trade at high revenue multiples with negative or marginal earnings. The 2027 inflection thesis assumes that the next generation of AI clusters (think NVIDIA’s Blackwell Ultra in 2026) will require these specific components at scale. But here’s the problem: order flow data from recent filings shows no material growth in customer orders for AAOI or AXTI. Their revenue is flat.
I’ve run this playbook before. In 2021, I automated NFT flips on OpenSea and learned that execution risk kills theoretical value. The same applies here. Serenity’s position sizing is execution risk on steroids. He’s all-in on a single narrative: “AI infrastructure bottlenecks will be solved by these three companies.” But technology seldom follows a straight line. In my 2022 Terra analysis, I spotted that Anchor Protocol’s yield was unsustainable because the peg relied on algorithmic minting. Here, the 2027 peg relies on algorithmic belief—no contracts, no committed customer orders, just a PowerPoint timeline.
Let’s quantify the pain. If Serenity’s initial capital was $1 (theoretical), at peak he had $46. Now he has $23. That’s a loss of $23 from peak—more than double his original stake. He calls it profit retracement, but in absolute terms, he’s lost more money than he put in. The only way this works is if the 2027 inflection actually happens and the stocks rally 3x or 4x from here. But the discount rate on that future cash flow is enormous. Every quarter of no execution pushes the inflection further right, and the stock prices will adjust accordingly.
I don't trade feelings; I trade backtests. I’ve programmed trading bots that exploit cross-chain arbitrage. The edge comes from verifiable data, not hope. For Serenity, the only verifiable data is the 49.4% drawdown. The 2027 thesis is an unproven hypothesis. In my 2024 Bitcoin ETF arbitrage, I executed 50+ trades in two weeks because the premium spread was measurable. Here, the spread is between narrative and reality—and it’s widening.
Contrarian
The crowd on CT is cheering Serenity as a diamond-handed legend. They see the 49.4% as a buying opportunity. I see the opposite: a classic trap of survivorship bias. For every successful small-cap AI investor, there are 20 who got wiped out. Serenity’s 26-year track record proves he’s a skilled trader, but past performance doesn’t guarantee future returns—especially with such concentrated risk.
My contrarian take: this portfolio is a ticking time bomb. The 2027 inflection assumes no technology substitution. But co-packaged optics (CPO) could render AAOI’s optical modules obsolete. AXTI’s gallium arsenide faces competition from larger silicon substrates. SIVE’s chip design might be replicated by in-house AI builders. The same risk exists in DeFi: code is law, until it isn’t. Smart contracts get exploited, peg mechanisms break, and liquidity vanishes when the music stops.
Furthermore, Serenity’s “profit retracement” framing masks a dangerous psychological bias: the inability to realize losses. He’s emotionally anchored to his 2027 thesis, which might cause him to hold through further declines. In my own trading, I’ve learned to cut losing positions when the thesis breaks—like when I sold my BAYC clones after recognizing the floor price divergence. Serenity is doing the opposite: he’s increasing conviction as the price drops.
The market is a discounting mechanism. Current prices for AAOI, AXTI, and SIVE already factor in a high probability of the 2027 inflection failing. If I were reading this as an options strategist, I’d buy downside puts on these names—not because I hate AI, but because the risk/reward for the next 12 months is skewed against them. Patience is a virtue, but so is capital preservation.
Takeaway
Risk isn’t a feeling—it’s a number. Serenity’s number is 49.4% and counting. The only way this trade works is if the 2027 inflection materializes with precision. One earnings miss, one technology pivot, one macro shock, and the drawdown deepens. I’ll be watching the Q2 2025 earnings calls for these companies. If I see no order growth, I short the narrative. If I see customer commitments, I might nibble. But for now, the candle tells a story of fear—and I’m not buying the fear.