Summary

李丰把过去五年的资产定价主线归结为：疫情后史无前例的超级流动性，恰遇2022年中国与欧洲同时被全球资金减配，大钱因而只能拥挤到美国。 He illustrates the scale with “$10T suddenly becoming $25T,” stressing that the truly unprecedented feature was “more money than ever in history, with only one region to invest it in”; the money first went into Treasuries, then crowded into derivatives and U.S. equities, explaining U.S. resilience and the Magnificent Seven rally through the rate-hike cycle.
这一极端配置已经开始反转，港股与欧洲是最清晰的两个价格信号。 In the first 5 months of this year, Hong Kong overtook the U.S. to rank first globally in both IPO count and proceeds, while average daily turnover rose from below HK$100B to above HK$200B; European equities gained about 30% over the same period despite little meaningful improvement in the economy. Li Feng therefore sees capital moving from “the U.S. is the only place to allocate” back toward a more normal distribution across the U.S., China and Europe—but if the original trend took 2 years to build, the reversal could also take 1.5 to 2 years, and only a little more than 6 months have passed.
AI既是技术进步，也是被超级流动性选中的估值支柱，因果顺序不能含糊。 Li Feng’s analogy is that U.S. megacaps originally had “10 floors,” but after capital pushed them to “50 floors,” AI had to serve as the pillars for the additional 40; Meta’s roughly $15B deal for Scale AI shows how the secondary-market ceiling can feed back into and lift primary-market exit prices. China’s most highly valued foundation-model startup is worth about RMB20B, “exactly one exchange rate and one order of magnitude” below its U.S. peers—a gap driven not only by differences in originality, but also by the liquidity position of the two markets.
如果资金逐步回配中国，前半程可能先扩散AI概念，后半程才由经济改善与蓝筹盈利接棒。 The U.S. combined original technology with an explosive liquidity surge; China is more likely to see a longer, slower and more dispersed reallocation: AI applications penetrate from 5% to 10% to 20%, while Tencent, Alibaba and Meituan—companies that carry both an AI narrative and broad economic representation—first absorb the market’s imagination, before more stable capital looks for real growth. “If the market keeps trading concepts, it means the money still isn’t large enough.”
投资人的难题不是看见“天上下钱”，而是决定拿碗还是拿盆，以及何时把盆收回来。 Investors must answer several questions at once: whether to keep funding AI startups valued at hundreds of millions of dollars, whether to partially exit projects already valued at $2B-$3B or even $5B-$6B, whether to accept M&A offers, and whether to add to, hold or cut exposure to the Magnificent Seven over the next year. One divergence worth watching is that, according to Goldman Sachs analysis cited by the show, institutions were reducing exposure to U.S. equities and other risk assets from April to June, while U.S. households and retail investors were adding.
李丰承认以“是否撤侨”判断以伊冲突不会升级是一次明确误判，而其可交易后果主要落在油价与中美通胀数据。 If the conflict keeps oil prices volatile for more than 3 weeks and covers the whole of June, U.S. inflation could reaccelerate, while China’s CPI might return to negative growth; the two countries’ CPIs have moved almost in lockstep for more than 6 months mainly because oil fell to the low $60s, suppressing U.S. energy inflation and offsetting some goods inflation, while pushing down both China’s CPI and the more supply-chain-sensitive PPI.
5月中国金融数据仍是“政府先发力、居民未接棒”，但资金形态正在变化。 Government financing accounted for about 70% of the increase in aggregate social financing, while household loans grew more slowly, indicating that demand for housing purchases and leverage has yet to recover; at the same time, direct financing through government bonds replaced part of bank lending, M1 turned to low-single-digit positive growth, and its gap with M2 narrowed. Household deposits did not surge, while deposits at non-bank institutions continued to grow relatively quickly—more consistent with money moving from deposits into wealth-management products under low rates than with a return to property.

Deep dive

1. “是否撤侨”不足以预判战争升级

Li Feng began by admitting error: in the previous episode, he judged that the Israel-Iran conflict “might not escalate,” based on the fact that China, the U.S. and Russia had not evacuated their citizens in advance. He had assumed that major powers always communicate behind a regional conflict and that at least one side would first protect its own nationals. The evidence showed that “going forward, whether citizens have been evacuated cannot be used as a standalone criterion.”
In hindsight, he found that China issued a red tourism warning for Israel on May 18, but it focused more on areas near Gaza and may not have been a warning about an Israel-Iran conflict. China and Russia only began evacuating citizens from Iran this week. The only inference is that the major powers either did not know what was coming or learned too late to take diplomatic action.
Li Feng therefore characterized the operation as “relatively sudden” for other countries: even if Israel had been planning it for a long time, the launch window left outsiders insufficient time to prepare. Li Xiang was struck by Iran’s highly passive response, recalling the military asymmetry exposed at the outset of the India-Pakistan and Russia-Ukraine conflicts. Li Feng said he had not studied the military dimension in depth, but confirmed that Iran’s passivity was “unexpected in several respects.”

2. 港股从“可交易”重新走到了“可融资”

From January 1 through May 31, Hong Kong overtook the U.S. to rank first globally in both IPO count and proceeds. Li Feng said the key was not Chinese ADRs transferring their listings, but the market once again absorbing genuine fundraising: Hengrui Medicine, CATL and the upcoming Haitai offering are all new-share issuances that raise fresh capital.
The contrast with early 2023 is stark. At the time, disputes over SEC audit papers pushed the discussion around Chinese stocks back toward Hong Kong, but the market broadly declared that Hong Kong’s liquidity had “absolutely no chance” of absorbing them. Today it is accommodating returning Chinese ADRs, large A-share companies conducting secondary offerings in Hong Kong and new Hong Kong-listed IPOs from local companies.
The hardest evidence remains turnover. When the market was depressed, Hong Kong’s average daily turnover was below HK$100B; now it can exceed HK$200B. Li Feng said the pessimism of 2 years ago was reasonable at the time, but fundraising capacity and trading depth are no longer those of the same market.
Jane Street’s full-building lease in a prime Hong Kong location, along with Hong Kong’s moves on stablecoins and digital currencies, were also cited by Li Feng as evidence that financial-center resources are returning. In 2023, many people urged him to move to Singapore, but he still believed Hong Kong would become one of the key financial centers for Chinese assets and Asia. 2 years later, that view is beginning to play out.

3. 供给与需求的解释，最终都落到了流动性

Li Xiang offered a three-layer explanation for Hong Kong’s recovery. On the supply side, Chinese companies need a listing venue where they can access international capital and raise dollar-denominated funding. On the demand side, Hong Kong remains a familiar and accessible market for global investors allocating to the world’s second-largest economy. The biggest change, however, has been the revaluation of Hong Kong’s role amid the U.S.-China-Europe geopolitical environment.
Li Feng accepted the supply-and-demand framework but folded both sides into a larger variable: “Hong Kong’s liquidity has changed substantially.” Companies are willing to list and investors are able to absorb the supply only because enough money has returned.
European equities gained about 30% over the same period, making Europe the best-performing regional market in the first 5 months of the year, even though the European economy could at most be described as bottoming out and might still contract year on year. Hong Kong ranking first in fundraising and European equities rising may look unrelated, but Li Feng sees the same force underneath: “Global capital flows have changed.”

4. 疫情后的钱，规模与速度都没有历史类比

Li Feng dates the starting point to Q2 2020, when the U.S. began injecting liquidity on a massive scale in April and the world’s major economies expanded liquidity in parallel. To convey the magnitude, he used a non-quantitative analogy: money that had been $10T suddenly became $25T. Global net debt also increased by tens of trillions of dollars over those 2 years.
This cycle differs from the period after 2008, the aftermath of the European debt crisis and the early mobile-internet era. Previously, liquidity was expanded gradually, allowing each major economy to capture a share. This time, within less than 1 year, the world printed 2x or even several times more money than in any single earlier crisis while pushing interest rates to extremely low levels.
The central question was therefore no longer just industrial fundamentals, but “where this enormous liquidity would go.” When the absolute change in money is large enough, it can overwhelm company-level progress, earnings changes and technology adoption, determining first that assets must be repriced.

5. 2022年把中国与欧洲同时变成了“难配置资产”

In China, the Hong Kong Legislative Council and Chief Executive elections, the Shanghai lockdown, Beijing’s near-lockdown, policy debates around the 20th Party Congress, restrictions on travel and reduced operating capacity for foreign media all worsened the external narrative. Li Feng stressed that these factors did not necessarily share a single causal direction, but together made “not willing, not daring or unable to invest in China” a constraint for large pools of capital.
After the Russia-Ukraine war broke out in late February 2022, the U.S. also placed China in discussions linking it to sanctions risk involving Russia. By the 2023 G7, “decoupling” had become “de-risking,” while Europe grew more anxious because of NATO spending, the protracted war and supply-chain risks.
Europe itself could not absorb the capital. It had to continue providing military support while dealing with an energy crisis; Germany and the U.K. even generated headlines about burning wood again. Energy is a foundational cost for industry and households, and the instability of war made Europe subject to underallocation as well.
Emerging markets such as India were too small to absorb incremental liquidity on the scale of “at least the low tens of trillions, or even tens of trillions of dollars.” China, the U.S. and Europe together account for more than half of global GDP. Once China and Europe were difficult to allocate to, the only sufficiently large container left was the U.S.

6. 美国的“exceptional”韧性包含了一次被迫拥挤

Once money entered the U.S., the natural sequence was “bonds first, stocks later.” Expectations of higher rates gave investors a stronger reason to hold Treasuries, but liquidity was too large to remain entirely in bonds; otherwise bonds too would have been bid to extreme valuations, leaving the residual capital to flow into derivatives and U.S. equities.
So even during a rate-hike cycle, U.S. equities were lifted again after brief bouts of volatility. Li Feng called the rise in U.S. capital markets and the resilience of the economy from the second half of 2022 through 2023 and even the first half of 2024 “an extremely unusual economic event,” rather than a normal outcome explained solely by fundamentals.
His core line was: “There has never been so much money in history, with only one region to invest it in.” The two exceptions compounded in sequence, lifting U.S. stocks, bonds and real estate, while the complete inflow of capital into the U.S. became an important backdrop to its economic resilience.

7. 七姐妹需要AI为被顶高的楼层补上柱子

During the conversation, Li Xiang initially placed the moment when GPT attracted broad attention in late 2023, then corrected himself to late 2022. That timing fell in the middle of the rapid liquidity-driven repricing of U.S. assets, making foundation models the ideal “technology-imagination variable” to explain the rally.
Li Feng’s analogy was that a company might originally have only 10 floors, but capital forcibly pushed it to 50, leaving 40 additional floors that needed nominal pillars. Because the huge inflows and outflows were concentrated in mega-cap companies, the Magnificent Seven were lifted first; AI was then used to explain the gains, reinforcing the story and the price in both directions.
Li Xiang noted that the previous Magnificent Seven cycle also reflected the platform economy’s economic influence and high barriers to entry; Alibaba and Tencent experienced similar pricing in China in 2021. Li Feng agreed that mobile internet also came with a liquidity cycle, but maintained that the simultaneous occurrence of “the most money” and “the U.S. as the only destination” made this cycle more exceptional.

8. 极端配置正在回归常态，但不会原路急转

The Russia-Ukraine war has not ended, but the market has partially normalized it. Investors have formed expectations around European risk, while concerns about uncertainty within the U.S. itself have increased. European equities rising about 30% shows that capital no longer equates an ongoing war with Europe being completely uninvestable.
On China, Li Feng cited “9/24,” allowing wholly foreign-owned companies in key service sectors, stepped-up efforts to attract foreign investment, financial opening, DeepSeek and the symposium with private companies. These efforts are changing expectations, but he does not treat any one event as the complete explanation; instead, he sees them as moments and accumulated changes that accompanied the reversal in capital flows.
The visa-free policy that began in late 2023 was particularly important. Even if China could not immediately influence global media, sufficient renewed movement of people could correct impressions formed in 2022 through firsthand experience. Li Feng also preserved the reverse explanation: perhaps China took more proactive action later precisely because external perceptions had been so poor at the time.
The extreme state took about 2 years to build. If the destination is merely a return to average allocation, rather than the U.S. suddenly becoming “uninvestable,” the unwinding could also take 1.5 to 2 years. Li Feng estimates that only a little more than 6 months have passed; the direction is “a reasonable U.S., a reasonable Europe and a reasonable China,” so the process will be longer, slower and more volatile.

9. 二级市场的高天花板，直接抬高了AI一级市场

The show treated Meta’s roughly $15B transaction for Scale AI as a signal. Silicon Valley funds broadly marked down their portfolios and pursued down rounds from 2022 through early 2023. From mid-2023 onward, foundation-model investment and valuations rose rapidly, and fund-return metrics swung back toward high growth.
Li Feng again used the “10 floors to 50 floors” framework. If a megacap rose from $500B to $1T, the imagined value of related startups could rise from $50B to $150B; a megacap that could previously acquire only a $5B company might now be able to absorb a $20B-$30B deal.
Primary-market valuations therefore do not form independently. Investors are willing to value companies at $5B, $10B or more because secondary-market megacaps have a larger market-cap currency and greater acquisition capacity: “I value it at $5B or $20B and sell it to you for $15B—I still make a lot of money.”
China’s most highly valued foundation-model startup is worth about RMB20B, “exactly one exchange rate and one order of magnitude” below the corresponding U.S. company. Li Feng acknowledged the U.S. advantage in originality, but also stressed that Chinese tech giants and AI startups remain “at the bottom of the cliff,” without the U.S. secondary-market ceiling to map against.

10. 流动性总能为上涨找到一个非流动性的理由

Li Feng said the first step is to isolate the largest variable and then distinguish cause from effect. If the money supply is stable, technology, earnings and company conditions can be rationally disaggregated. If liquidity undergoes an absolute change, prices are more likely to move first and the market to search for reasons afterward. “Once money decides to make an asset rise, it can always find a reason.”
Li Xiang added mobile internet as an example: O2O and online-offline integration were good reasons to absorb capital at the time. Li Feng then cited 2020 consumer investing—Haitian soy sauce, vinegar, mineral water and baijiu could all be packaged as “the Maotai of soy sauce” or “the Maotai of water,” because super-liquidity needed to find enormous valuation space in consumption.
This does not mean technology and companies have no value; it means the magnitude of the rally may exceed the underlying value. Li Feng expects that if the bubble bursts, someone will rewrite the history and debate how much of the AI boom came from technology and how much from an unusual financial and liquidity structure.

11. 一级投资的成就感来自陪跑，二级投资来自逆势判断

Li Xiang asked whether primary-market investors lose their sense of achievement if valuations are driven mainly by the tide. Li Feng’s answer was that early-stage investing is the smallest wave within a larger tide and has relatively limited macro influence. Its value lies in watching something “go from 0 to 1, from 1 to 3 and from 3 to 7.”
He used the listing of the “film and television technology” company mentioned on the show as an example. At the IPO, the founder’s parents came specifically to thank the earliest supporters, while the founder repeatedly expressed his gratitude. For Li Feng, accompanying a person and a company through the entire journey from nothing to something is the unique return of primary investing.
In the secondary market, the achievement comes from “thinking differently when nobody else thinks that way,” then turning 20 into 50 or 100 by staying the course. If a primary investor wants to continue participating, they generally have to sell all the shares acquired in the primary market and buy them back in the secondary market, and they can no longer serve as a director, supervisor or senior executive.

12. AI泡沫里的四个决定，都比识别趋势更难

For early-stage investors in the U.S., the first question is whether to keep investing when an AI startup is already valued at hundreds of millions of dollars. The second is whether to sell at least part of the position and realize gains when an early-stage project rises to $2B-$3B, $3B-$5B or even $5B-$6B and continues raising money.
The third question is whether to push for or approve an acquisition. Li Feng compared it with the 2021 consumer bubble: it was difficult to persuade people to invest in consumer startups valued at RMB200M-RMB300M, but it was equally difficult to persuade them to sell shares in famous consumer companies that had already risen several-fold. Founders faced the same difficulty when presented with acquisition offers of $1B or $1.5B.
The fourth falls directly on every U.S. equity holder: over a 1-year horizon rather than 20 years, should the Magnificent Seven be increased, held firmly, trimmed or exited entirely? Goldman Sachs analysis cited by the show said institutions reduced risk assets substantially from April through June while U.S. households and retail investors added, creating a small contest between the two sides.
For newly listed AI projects whose lockups have expired, whether to sell, hold or sell only a little also depends on the listing venue. The liquidity cycles in U.S. stocks, Hong Kong stocks and A-shares are “one beat apart,” so the same valuation and exit rules cannot be applied mechanically. The hosts repeatedly stressed that this was abstract speculation and “not technical investment advice.”

13. 中国接到的不是暴雨，而更像一场持续回流

Li Xiang framed the question this way: if “money is falling from the sky,” should investors bring a bowl or a basin? Li Feng first imposed a strict condition: only if the cycle is in its first half or has just passed its inflection point, and the liquidity effect will continue expanding, is it worth looking for assets capable of absorbing it.
The U.S. combined original foundation models, an extremely rapid liquidity surge and a super-large capital market, so companies closest to the base model “exploded quickly.” If China is merely returning from extreme underallocation to normal allocation, rather than the U.S. suddenly becoming uninvestable, it will not replicate the same speed or concentration.
China is more likely to pair liquidity with application penetration: a product moves from 5% to 10% to 20% penetration in consumer behavior, while company revenue and risk appetite rise gradually. This process would be longer, slower and more dispersed than the U.S. foundation-model rally, and would require economic expectations to improve in parallel.
The only event that could accelerate the pace sharply would be the U.S. suddenly becoming “uninvestable”; Li Feng used internal conflict within the U.S. as a purely hypothetical example. In that case, capital would shift in relatively dispersed fashion toward China, Europe and other emerging markets, and a new explosive concept would be needed in the short term to absorb it.

14. 回流前半程炒映射，后半程买真实增长

Li Feng’s conditional speculation about China is that the first phase would absorb the continuation of the U.S. wealth effect, with the AI concept spreading to domestic names. Because the U.S. already provides reference PE multiples, valuations and technical standards, concept-driven pricing is easier than creating a new narrative from scratch.
Tencent, Alibaba and Meituan are particularly suitable intersections: they can all be included in the AI narrative while representing important parts of the economy—entertainment, retail and services, respectively.
As longer-duration, more stable and lower-risk-tolerance capital enters, the market must see the economy emerge from its trough and see earnings and business growth improve at mid- and large-cap blue chips. Only when capital begins allocating to companies that represent changes in economic categories does it qualify as a larger-scale, rational reallocation. “If the market keeps trading concepts,” it means the money that has truly returned is still not large enough.

15. 创新药与美国VC都提供了泡沫后半程的识别样本

After Hong Kong’s Chapter 18A launched in 2018, China’s innovative-drug sector accelerated clinical development and commercialization. Just as the bubble may have been close to bursting, the 2020 pandemic pushed biotech valuations higher; money printing added another layer, creating “two exceptions stacked on top of each other.” When liquidity reversed in the second half of 2021, both bubbles burst together, producing exceptionally severe losses.
The subsequent recovery first traded the possibility that a drug might be sold at a high price: companies with drugs that could be acquired by big pharma at a premium received higher valuations. But Li Feng’s test was whether capital would eventually return to the entire innovative-drug chain, including early-stage R&D, CROs and clinical trials. If the market keeps trading only the possibility that a particular drug will be sold at a high price, it has merely passed the bottom and has not returned to rational allocation.
For U.S. AI VC, he offered no definitive conclusion, but proposed a checklist for “assuming the bubble is in its second half”: the number of new investments falls materially while early-stage valuations remain elevated; exits, mergers and partial acquisitions increase while late-stage valuations continue rising; large financings concentrate in a small number of leading projects, with investors still competing over who dares to bid higher.
The comparison is SoftBank’s Vision Fund at the end of the mobile-internet cycle in 2014-2015. Some companies eventually survived to benefit from the super-liquidity cycle in 2020, but SoftBank’s Vision Fund went through an extremely painful period from 2017 to 2019. Large financings at the tail end do not automatically prove that the cycle is still in its early stages.

16. 油价把中美CPI绑在同一方向，却产生相反叙事

Li Feng recommended reading the long explanations that follow official data. Since Q4 last year, China has proactively explained components and causes when releasing macro data. The publisher will not emphasize the worst part, and neither will the U.S., but if the goal were truly to conceal it, “it would be better not to publish the data or to frame it from a completely different angle.” The more effective approach is to test the explanations and observe relationships within the data and between China and the U.S.
China’s CPI can be roughly understood as energy plus food, while the U.S. is housing plus gasoline. After buying a home, Americans impute the equivalent rental service into CPI; transportation, electricity and heating are also affected by energy. China’s manufacturing chain is longer, so oil prices transmit repeatedly through ethylene, propylene, packaging bags, packaging boxes and other intermediate links, making PPI especially sensitive to energy.
In China’s negative CPI growth in May, lower energy prices contributed roughly half. Core CPI and the measure excluding energy and food were relatively acceptable. PPI was about -3% year on year, with little month-on-month improvement; lower energy and commodity prices contributed more than half of the monthly PPI decline, while lower prices for intermediate and industrial goods contributed about 20%. Ordinary consumers can reverse-check the data using the timing and range of domestic gasoline price adjustments instead of tracking crude oil directly.
Over the past 6 months or more, oil prices falling to the low $60s suppressed U.S. inflation but pushed China’s CPI and PPI toward deflation, causing the two countries’ indicators to move almost in lockstep. If oil-price volatility caused by the Israel-Iran conflict lasts more than 3 weeks and covers June, the U.S. could face renewed inflation, while China’s CPI “might return to negative growth.” This remains a scenario conditional on those specific assumptions.

17. 5月金融数据仍靠政府托底，居民资金先去了理财

About 70% of the increase in aggregate social financing in May came from government-related financing, while household loans grew more slowly, showing that household borrowing demand has yet to recover. Li Feng therefore believes that fiscal support and government backstopping may need to continue for some time; this is a weakness mentioned in the official explanation but not emphasized.
Special-purpose bonds, ultra-long special sovereign bonds and bonds issued to replace local-government debt are all direct transactions between issuers and subscribers, so they count as direct financing. Slower loan growth does not necessarily mean financing has contracted: part of corporate and government borrowing demand has been replaced by bond issuance, consistent with China’s structural shift from bank lending toward bond and equity financing.
M1 turned to low-single-digit positive growth in May, narrowing its gap with M2. Because the M1 definition now includes the reserve funds of third-party payment companies such as Alipay’s Yu’e Bao and Tencent-related products, year-on-year comparability is somewhat weaker, but month-on-month data also improved. Official commentary attributes this to a modest recovery in households’ willingness to spend, with retail-sales data offering some corroboration.
In April, slower or declining household deposits and rapid growth in non-bank deposits showed money moving from term deposits into brokerages, insurers, funds and wealth-management products. Household deposits recovered in May but did not surge, while non-bank deposits continued to grow relatively quickly, extending part of the trend. Since household loans did not accelerate, the money looks less like a clear return to property and more like asset reallocation under low interest rates.

Bobby Goes Live: Vakee on Its Creation and Everyday Investing Coatue Pt2. Open AI’s Kevin Weil Dives into All Things ChatGPT | BG2 w/ Bill Gurley & Brad Gerstner