On the X platform, an anonymous user identified as Serenity recorded a book return of 45x in 2026, with all 25 publicly disclosed stocks appreciating between 1x and 10x. Within slightly over a year of joining the platform, Serenity accumulated more than 400,000 followers and over 37,000 paid subscribers. In the financial investment sector, surpassing 30,000 subscribers places an account in legendary territory, second only to 马斯克 in terms of influence. Chinese-speaking communities have dubbed him the 'White-Haired Stock God' due to his avatar, while WallStreetBets now lists his posts as daily essentials. Major outlets like Bloomberg and Reuters cite his research, and hedge funds actively track his updates. Serenity's background includes roles as an AI research scientist, a Nature author, and a member of the RISC-V Foundation. In 2018, when Nvidia's stock traded around $6, he reportedly declined an offer from Nvidia's AI team head. While these claims remain unverified, his investment performance is indisputable.

Serenity first emerged on Reddit's WallStreetBets under the handle AleaBito on January 11, 2022. The forum is known for emotional trading calls and poker-style options, yet AleaBito's early posts utilized humorous analogies.

However, in early 2022, a detailed analysis of AXTI broke this pattern, setting target prices between $15 and $150 without memes. The post meticulously mapped the supply chain from Nvidia's H100 clusters to InP substrates, noting AXTI controlled 25% to 35% of global InP production capacity. It further dissected material sources, geopolitical risks, and patent barriers, arguing that the AI optical communication industry relies entirely on InP substrates. At the time, AXT Inc. held a market value of roughly $200 million with minimal institutional coverage. Due to the technical depth clashing with the forum's casual tone, moderators permanently banned the account for promoting speculation. Subsequently, $AXTI surged from $12 to over $70, yielding a profit margin exceeding 1,000%.

In July 2025, Serenity migrated to X, publishing deep-dive analyses of the AI supply chain complete with diagrams, materials science papers, and capacity maps. Data compiled by Woofun AI shows that Serenity repeatedly emphasized that early problem identification invites skepticism. His methodology, termed the 'Perilla Leaf Theory' in Chinese communities, posits that while customers focus on tuna belly in sushi, the perilla leaf from small Izu farms ensures business continuity. Similarly, in the AI industry, GPUs and large models are the tuna belly, whereas InP substrates, CPO lasers, and high-purity phosphorus are the critical perilla leaves. Serenity probes deeper than GPU procurement, asking how tens of thousands of GPUs communicate and what fails first. His conclusion points to copper wiring interconnections, which hit physical limits regarding power consumption and bandwidth as cluster sizes expanded.

The attenuation of high-frequency electrical signals in copper wires caused severe heat dissipation issues, shifting the bottleneck from computing power to data transmission. This necessitated CPO (Opto-Electronic Co-Packaging) technology, integrating optical components and chips on the same substrate to reduce transmission distances from meters to millimeters. Woofun AI notes that Serenity identified five key bottlenecks: nanoscale alignment components, external continuous-wave laser sources, MBE equipment for compound semiconductors, 6N purity phosphorus materials, and SOI substrates. For each, he located globally scarce suppliers. Instead of relying solely on financial reports, he analyzed hyperscaler capital expenditures, data center expansion paces, and physical bandwidth limits. He cross-referenced materials science papers, patents, capacity plans, supplier certifications, and export control policies. Before publication, he fed his logic into multiple AI models to detect flaws and valuation discrepancies.

Over recent years, Serenity's core assets have focused on physical bottlenecks in AI infrastructure, including InP substrates, CPO lasers, optical transceivers, silicon photonics, and edge hardware. Representative holdings include $AXTI, $SIVE, and $AAOI. $AXTI was identified two years in advance, before ChatGPT, when the market fixated on GPUs and training chips. Serenity likened its role to the Strait of Hormuz, controlling 20% of global oil flow. $SIVE (Sivers Semiconductors), a Swedish firm supplying external continuous-wave lasers for CPO, is now a top holding. Serenity built his position when its market value was around $150 million, anticipating acquisition by $AVGO or $MRVL for $200 million to $300 million to secure the CPO supply chain. $AAOI (Applied Optoelectronics), a full-stack optical transceiver player, was bought at a $660 million market cap and has since risen more than seven times.

Another example is $Raspberry Pi, often viewed as an educational single-board computer with little institutional interest upon its London Stock Exchange listing. Serenity observed AI startups using it for intelligent control systems, predicting a 58% annual revenue growth rate versus the market consensus of 14%. The actual growth matched his prediction, driving the stock up 44% on the first day and 27% the next after earnings.

However, errors occur; in 2026, a misjudgment on Japanese packaging equipment manufacturer $TOWA caused a 20% drop. Serenity acknowledged on social media that short-term errors stemmed from one-time accounting factors and initial cost savings, though he maintained the underlying logic for the second half of the year remained intact. His latest holding is $XFAB, the only high-volume SiC foundry in Europe and the US capable of silicon photonics manufacturing. With a market value of $1.28 billion and a price-to-book ratio of 1.29, it aligns with his pattern of small-cap, under-covered assets at physical bottlenecks.

Most of Serenity's holdings rely on the premise that AI data centers will increasingly depend on optical interconnections and CPO architectures. While his portfolio spans different companies, the fundamental logic remains consistent. If data center strategies shift or CPO progress stalls, these assets face risk. Conversely, their small size and limited liquidity hinder large institutional management, allowing Serenity to identify them early. Woofun AI analysis suggests that while the market debates model benchmarks, Serenity focuses on lasers, substrates, thermal losses, and molecular purity. He asks at what scale copper wiring will fail and who controls the critical physical infrastructure. This logic rests on physical principles rather than product decisions or sentiment. From a banned WSB user to an expert followed by 400,000, Serenity demonstrates that in an era of information overload, the scarce resource is the ability to identify unavoidable physical bottlenecks ahead of the crowd.