On October 11, a weekly report from Liang Hong, founder of the billion-yuan private fund Hainan Xiva, went viral across the capital markets.
This outspoken private fund tycoon did something many of his peers would not dare to do — publicly criticize insurance capital.
He said that certain major insurance asset management giants bought a pile of tech stocks at high prices in July, then significantly cut positions and stopped losses in late September. "Such large institutions chasing rallies and dumping on dips is truly baffling."
To translate: insurance capital, you are supposed to be the ballast stone — why are you more panicked than retail investors?
From glory to pressure, Liang Hong publicly takes aim at insurance capital
Before the criticism, Liang Hong first laid out his own performance cards.
Liang Hong revealed that most of his funds were expected to see net value declines of more than 2% this week, with drawdowns exceeding 20% from their June peak.
A cumulative 20% drawdown is no small figure for a billion-yuan private fund.
Take Xiva Xiaoniu No. 19 as an example — its year-to-date return once exceeded 58% in June, but now stands at only 27.43%, with returns cut in half.
From 50% to a drawdown of over 20%, only a little more than three months passed.
Just in the first half of the year, Liang Hong was the return champion among billion-yuan private fund founders.
Data from Private Fund Ranking Network shows that Liang Hong led the billion-yuan private fund founder rankings with a first-half return of 50.03%, followed by Evolution's Wang Yiping at 37.70%, while Dan Bin, Zhu Xiaokang and others all posted returns above 20%.
Measured by the average return of the company's products, Hainan Xiva's first-half average also reached 36.10%, ranking among the top billion-yuan subjective private funds.
Over a longer horizon, Liang Hong's three-year return stands at 107.51%, with long-term returns also at the doubling level.
This is the core reason the weekly report resonated so widely: the most prominent billion-yuan private fund of the first half is now also mired in the agony of drawdowns.
He honestly disclosed his position adjustments over recent months: due to regulatory restrictions, US individual stocks can only be sold, not newly bought.
Xiva gradually liquidated its US storage holdings, retained a 20% wafer manufacturing position, and shifted toward Hong Kong internet stocks, wafer manufacturing, and A-share tech stocks, betting on valuation repair and growth potential.
The result was the opposite of what he intended.
The Hong Kong and A-share tech stocks he bought continued to be hit by liquidity shocks and sentiment-driven selling in September.
Meanwhile, the one heavily weighted US stock he retained hit a record high during the same period.
This cross-market reallocation actually amplified net value volatility.
Liang Hong's core judgment is clear: the decline is not about fundamentals — it is about capital flows.
The AI industry development logic has not changed, and the US GPU leader even hit a new high, but the same type of company in China saw its stock price keep falling.
He stated bluntly, "The domestic capital flow situation is not good, which indeed surprised me."
The implication: his stock-picking logic is fine — it is the market capital that is not giving the rally.
This judgment is only half valid.
The tech stock decline does face capital flow pressure — the Fed's rate cut pace keeps fluctuating, foreign capital is flowing out, and the stock game lacks buyers.
But the other half is that he himself admitted many A-share tech stocks initially "had bubbles."
When the bubble is squeezed out, anyone inside gets cut.
As for his claim that "insurance capital chased the rally in July and cut losses in September," no insurance institution has responded publicly so far, and the veracity remains a question mark.
However, a research report from Zhongtai Securities provides some supporting evidence.
How much did insurance capital actually sell during the tech stock pullback?
On October 11, Zhongtai Securities analysts Xu Chi and Zhang Wenyu published a strategy report titled "STAR Market Breaks Below Annual Line — How Much Selling Pressure Has Cleared?", dissecting the sources of internal selling pressure behind A-shares' relative weakness versus overseas markets.
The report directly concluded: the main sellers in this round came from insurance and fixed-income-plus products.
By observing the capital structure since September, the two analysts essentially ruled out leverage, quant funds, retail investors, foreign capital, and long-term funds as the main sellers.
Public fund selling pressure has already been largely released: the trading concentration of the top 5% of stocks by turnover has fallen from a high of nearly 54% in June-July to around 43%, the historical median.
What is still selling is insurance capital and fixed-income-plus products.
In the second quarter, they concentrated on adding tech exposure, and their correlation with the tech sector has fallen from its high by about 70% to 80%, but remains about 20% to 30% above the 2025 normal level, corresponding to potentially 20% to 30% of newly added tech positions still to be digested.
In other words, insurance capital clearing is in its later stage but not yet finished.
This is interesting.
The role of insurance capital in the capital markets was supposed to be "smart money," "long-term capital," and "ballast stone."
But based on Zhongtai's data, chasing into tech in the second quarter and retreating in the third quarter looks no different from retail investors.
Liang Hong says it is "baffling," but from the insurance capital perspective, it may not be baffling at all.
Insurance funds have liability durations exceeding 10 years, and long-cycle assessment reform was published in July 2025, but old assessment inertia remains — quarterly rankings and annual absolute returns pile pressure layer by layer.
When tech stocks rise, not allocating means falling behind peers; when they fall, not cutting means ugly reports.
Under assessment pressure, long-termism is a luxury.
More critically, insurance capital faces inherently less pressure when selling tech stocks.
Out of every 100 yuan of insurance capital, roughly 50 yuan is in bonds, just over 10 yuan is in stocks, and tech stocks are only a small portion of that equity allocation.
Profits are icing on the cake; losses are only skin-deep.
But public and private funds are different — equities are their main net worth, and a 20% drawdown is real pain.
Insurance capital's "chasing rallies and dumping on dips" is a passive action forced by assessments; public and private funds' "toughing it out on the left side" is a fight with no retreat.
Different pressures naturally produce different postures.
Liang Hong's criticism is not fundamentally about accusing insurance capital of poor investment ability, but about exposing the asymmetry in this capital structure.
On the public fund side, in April this year the Asset Management Association of China issued the "Fund Management Company Performance Assessment Management Guidelines," requiring that among fund investment return indicators, medium- and long-term indicators of 3 years or more carry a minimum weight of no less than 80%, directly tying assessments to fund manager compensation.
After the July tech stock pullback, many tech-themed public funds that had doubled in the first half reopened subscriptions, signaling confidence in the sector's long-term value.
CITIC Securities says "don't rush," but what is the premise?
On the other end of the market, brokerage leader CITIC Securities also could not sit still.
A few days ago, CITIC Securities published a piece titled "Don't Rush," with the core message being just one line: A-shares are not expensive, but cheapness is not a reason to rise — it is only a safety cushion.
Objectively speaking, this judgment is not wrong.
The ChiNext Index has a price-to-earnings ratio of 35.83 times, at the 20.90% historical percentile — indeed at a low level.
The STAR 50 valuation stands at 125 times, at the 78.85% percentile — hanging at a high level.
Data source: Wind, compiled by CITIC Securities, as of October 7, 2026
But the two words "don't rush" carry a subtext: a valuation safety cushion only provides psychological comfort to fully invested investors who are trapped; for those sitting on cash, it represents a potential opportunity.
Liang Hong stated that if his products continue to decline, Xiva will launch a new round of self-purchases; insurance capital is using position cuts to shrink tech sector allocations; and brokerages are speaking out when the market feels the worst.
Three roles, three postures.
For ordinary investors watching this capital game, the biggest takeaway is seeing one thing clearly: in this market, no one is born a long-termist — all long-term beliefs must withstand market tests under the constraints of assessments, risk control, and liabilities.
Comments