每周编辑精选 Weekly Editor's Picks (0530-0605)

The information flow is too fast, making it easy for in-depth analysis articles to be overwhelmed by hot topics. This "Weekly Editor's Pick" column sifts through the vast amount of news to find valuable content, helping you filter out the noise, retain insights, and gain inspiration.

Macro Landscape

30-Year Treasury Yield Breaks 5% Again, the Era of "Everything is Cheap" Comes to an End

The three pillars that supported low inflation and low interest rates in the US over the past 50 years—cheap capital, cheap labor, and cheap energy—are simultaneously crumbling. There are also several "slow variables": escalating government debt, intensifying geopolitical friction, and the spread of populism.

The combined effect of these risks is that lenders demand higher risk premiums before they are willing to lend out their money—especially for longer durations. This directly pushes up long-end interest rates, i.e., the 30-year Treasury yield.

The US is bidding farewell to the low-interest-rate era and entering a new phase with more persistent and diversified inflationary pressures. The trajectory of AI will be the biggest unknown determining future inflation trends.

After the "Midas Touch" Hit IBM, President Trump's Next Target Emerges

Over the past year, publicly listed companies that Trump has named and praised are showing increasingly clear overlaps with his holdings, government industrial policies, and federal funding flows.

Perhaps the most memorable instance was when Trump turned the White House South Lawn into a Tesla product launch event. In front of media cameras, he sat in a Model S, calling Tesla a "great product" and the Cybertruck having the "coolest design."

This was followed by a string of companies like Dell, Intel, Micron, Nvidia, IBM, Apple, and Thermo Fisher gradually making it onto his public commendation list.

Some companies saw significant stock price movements after being named; for others, Trump's accounts had already established positions before the praise; and some companies simultaneously received government contracts, subsidies, export licenses, or other policy support.

The next batch most likely to receive a call-out from Trump are companies the government has already invested in: MP Materials (MP), Lithium Americas (LAC), IonQ (IONQ), Rigetti (RGTI), D-Wave (QBTS), and others. According to The Wall Street Journal, companies including IonQ (IONQ), Rigetti (RGTI), and D-Wave (QBTS) are discussing securing at least $10 million in grant support through "government equity or quasi-equity arrangements." Quantum Computing (QUBT) and Atom Computing are also being discussed within a similar framework. These quantum computing sectors are still in their very early stages, but their uniqueness lies in the fact that they almost naturally belong to the intersection of national security and basic scientific research.

President's Q1 Holdings Disclosure: Is Trump's Money Accelerating Towards AI Infrastructure?

In the first quarter, the largest sell-offs in Trump-related accounts were concentrated in Microsoft, Amazon, and Meta. The money flowed into: semiconductors, AI hardware, enterprise software, consumer electronics, broad-based indices, as well as some bonds and preferred stocks.

Directly copying his trades isn't very meaningful for three reasons: the disclosure is lagging; the disclosed amounts are just ranges; and the relevant accounts might be independently managed by third-party institutions, so outsiders don't know whether each trade was an active decision, portfolio rebalancing, or model-based allocation.

Its true value lies in indicating directional changes. Three structural clues are worth noting: AI trades are moving from models and applications to infrastructure; semiconductors are no longer just one company (Nvidia); and the AI transformation of enterprise software might be the most underestimated aspect.

Also recommended: "Four Valuation Anchors, One Musk Premium: The Real Disagreement in the SpaceX IPO".

Investment & Entrepreneurship

Crypto is Dead, Long Live Perps

Over the past decade, the core capability of the crypto space was asset issuance, but native assets are heading towards a slow death. What's sucking up liquidity and attention are old-world assets: US stocks, US Treasuries, gold, crude oil, indices... "Crypto is dead" refers to the era of relying on the constant expansion of native assets coming to an end.

The CEX-ification of on-chain Perps + trust transfer after 10.11 + volatility in macro assets like gold and crude oil + the explosion of US stock trading = the rise of Hyperliquid.

If you're doing US stock Perps, the platform only needs to create a contract pool around the price. Liquidity can be provided by ecosystem partners, and users trade price exposure without directly holding the underlying equity. It bypasses the heaviest part and captures the part with the most trading demand. This is what makes Perps both fascinating and insidious.

Perps don't create new assets, they create new casinos. They don't provide ownership, but they do provide risk exposure. Their goal is not to reconstruct the financial world, but to turn all assets into a "price" that can be traded 24/7, creating unprecedented liquidity and price discovery efficiency.

Today, the most successful currency in crypto is the US dollar, the most successful asset is Bitcoin, the most successful application is trading, and the "most anticipated new growth" comes from US stocks.

Bitwise CIO: Crypto Has Become a Contrarian Investment, Three Logics to Understand the Current Market

The crypto market will continue to face pressure in the short term. The tug-of-war over the CLARITY Act continues, SpaceX is about to IPO, Anthropic has filed its prospectus, and AI themes keep dominating financial headlines.

Increasing crypto holdings now will likely lead to a poor experience, but the essence of contrarian investing is precisely to position in areas no one is watching, making counter-intuitive decisions against the trend—anchoring to fundamentals and value to identify quality assets, with long-term returns expected to be substantial.

After MSTR Broke Its "Never Sell BTC" Promise, Panic or Opportunity?

On May 29, 2026, Strategy deposited 411.48 BTC (worth approximately $30.3 million) to Coinbase Prime, its first such exchange transfer in nearly two years. The transfer amount is less than 0.05% of Strategy's total holdings (approximately 843,738 BTC as of mid-May 2026).

A transfer to an exchange does not mean the sale is complete; institutional holders often move Bitcoin for custodial adjustments, collateral management, or OTC trade settlements. Strategy's STRC preferred stock—with an annualized dividend yield of about 11.5%—creates reasonable cash flow obligations. This is the primary driver for potentially selling BTC, not a wavering of conviction.

Just two weeks before the Coinbase Prime deposit, Strategy purchased 24,869 BTC for approximately $2.01 billion, confirming its long-term Bitcoin accumulation model continues.

Bitcoin's long-term price trend continues to rely on post-halving supply contraction, institutional ETF demand, and corporate treasury reserve activities—not a single wallet action.

Tiger Research: Korean Retail Crypto Investors are Disappearing, Who Will Support the Market in 2026?

Institutional crypto activities in South Korea have moved beyond the MOU (Memorandum of Understanding) stage into concrete business operations and exchange equity acquisitions.

Institutions are subtly intensifying competition to seize key financial infrastructure, including STO standard-setting, stablecoin payment rails, and the custody market.

Domestic infrastructure builders are becoming the core pillars of institutional business, constructing localized rails that comply with the Bank of Korea's CBDC framework and local regulatory requirements, reducing reliance on foreign technology.

The strategy for overseas Web3 foundations entering Korea has completely shifted from retail community building to collaborating with large corporations and financial institutions, as traditional finance accelerates its takeover of the market.

I've Been a VC in Web3 for Nine Years: Asian Funds Are Experiencing "Hell Mode"

Nine years, three bull-bear cycles. The logic of Crypto VC has fundamentally changed. This cycle offers structural opportunities for research-driven funds. Good projects will proactively seek institutions that can provide genuine non-financial value, rather than those that blindly offer high valuations.

Capital across the entire industry is shrinking. The playbook for US funds is different; many operate on a 10-year cycle. During the downturn, US funds have ample reserves and multiple paths forward. But Asian funds, after being pushed to highs together, find themselves with no way out when the market falls, entering an extremely painful "hell mode." Hence, a large number of Asian Crypto VCs have disappeared.

The biggest problem in the crypto industry is the decoupling of tokens from value.

Truly great projects are only born during the most pessimistic phases of each cycle.

Also recommended: "DAT Failed? The Listed Company Betting on HYPE Has a $1.25 Billion Unrealized Profit" "HYPE Spot ETF Accumulates 1% for 14 Consecutive Days, Is $75 New High Just the Start?.

Web3 & AI

Racing to Beat OpenAI to IPO, Anthropic Aims to Seize AI "Pricing Power"

Anthropic announced Monday that it has confidentially submitted its US IPO application, beating rival OpenAI to the listing process. OpenAI has not yet followed suit with its filing. OpenAI CEO Sam Altman stated he is "not focused on the potential IPO timeline" and that the company "will go public at the right time."

By filing first, Anthropic pushes its competition with OpenAI—spanning models, revenue, and valuation—into the arena of public market pricing. For investors, this competition is no longer just about "whose model is smarter." Now, it's also about who can translate the AI narrative into financial statements the public market is willing to buy into first.

In prediction markets, most participants previously expected OpenAI to file for an IPO before Anthropic. Filing first means capturing the narrative, but also taking on the risk first. The public market will ask: how fast is revenue growing, are computing costs rising faster or slower, how much of total revenue ultimately goes to partners, and are enterprise customers genuinely retained or inflated by short-term AI hype.

The AI PC Battle: Don't Bet on Camps, Bet on Toll Booths

AI PC presents three layers of opportunity:

The first layer is the advanced manufacturing toll booth. No matter who wins, TSMC is more likely to collect the toll.

The second layer is the spillover of computing power and platforms. AMD and NVDA represent the x86 offensive and the extension of GPU software stacks, respectively.

The third layer is architectural diffusion and turnaround narratives. Both ARM and INTC have flexibility, but position discipline must be stricter.

From "Old-Guard Stocks" to "New Aristocrats": How AI is Revaluing Old Infrastructure

As AI moves from model parameters to real data centers, the market naturally re-evaluates companies with delivery capabilities and infrastructure prowess. This is why Dell, HP, Nokia, and others are being seen again.

The AI era doesn't require all data to sit in the most expensive high-speed storage. Vast amounts of cold data, training data, log data, video data, and archived data still need cost-effective, high-capacity hard drives.

To determine if an old company is genuinely being revalued, look for at least three criteria: confirmed orders and revenue; guidance upgrades; and whether profit quality can keep pace.

Prediction Markets

Polymarket Starts Testing, Kalshi Gets Approval; Perpetual Contracts Arrive for the Prediction Market Duo

Last week, both Polymarket and Kalshi saw key developments. Polymarket's perpetual contract Beta version has been opened for testing to some users, with access gradually expanding over the next 4 weeks. Kalshi received CFTC approval to list the Bitcoin perpetual contract BTCPERP. One is conducting small-scale product testing first, the other securing regulatory approval first. Different paths, but the same signal: prediction market platforms are no longer content with just event trading; they are moving into higher-frequency, more standardized derivatives markets.

For prediction markets, launching Perps isn't because prediction markets aren't profitable, but to add a more mature contract business alongside event trading. However, when competing with established giants, the brand and traffic of prediction markets won't automatically translate into competitiveness in contract trading. It remains a challenging path.

Policy & Stablecoins

US Government Allows Crypto Perpetual Contracts for the First Time: What Does It Mean for the Market?

On May 29, the US Commodity Futures Trading Commission (CFTC) issued supervisory guidance for 7x24 trading, emphasizing that crypto-related derivatives are better suited for around-the-clock trading and clearing due to their digital infrastructure and global continuous trading characteristics.

This means the US, previously considered a "forbidden zone for crypto perpetual contracts," has opened up for the first time. The CFTC has officially opened this market—which previously had virtually zero share—to US citizens and domestic crypto platforms and CEM exchanges.

Direct beneficiaries of the new policy: Kalshi, Coinbase, CME. The US market is poised for an explosion in derivatives trading.

CeFi & DeFi

Regulation Relaxes, but the Path Narrow for Hyperliquid?

The CFTC has liberalized contracts. However, Kyle, Hyperliquid's biggest critic and former Multicoin partner, poured cold water on the Hyperliquid community: "What you have now is a guarantee that no regulated US company will ever distribute Hyperliquid liquidity."

To legally operate a perpetual contract trading platform in the US, three types of businesses and licenses are required: DCM, for the trading platform itself; DCO, for the clearinghouse, i.e., the centralized clearing counterparty; and FCM, for the intermediary broker. All three are indispensable. However, the entire regulatory framework for operating trading platforms was designed from the outset to exclude entities without DCO qualification, including Hyperliquid, from the broker access list, as such Perp DEXs fundamentally do not rely on a "clearinghouse."

Another lingering concern is the age-old question: "Who takes the blame?" The regulatory instinct is to find an accountable entity: if something goes wrong, who do I summon, who do I penalize. In the traditional framework, the regulated entities are tangible intermediaries like FCMs, DCOs, and DCMs. But under the banner of "decentralization," "who takes the blame" remains a legal vacuum.

Hyperliquid faces three paths: remain offshore; fully onshore; continue pursuing decentralization until passing the "8-prong decentralization test" of the Clarity Act.

Commanding 94% Market Share, Who Is the Real Winner Behind the Stock Token Competition?

Whether CeFi or DeFi, virtually every stock token trading service you see has Alpaca behind it. According to data disclosed by Alpaca on December 4 last year, the company holds a 94% market share in the tokenized US stock and ETF asset market.

Unlike traditional brokerages that directly serve end investors, Alpaca's core clients are developers, fintech companies, and other brokerage institutions. Al