GF SECURITIES Liu Chenming: There is hope for a repeat of deposit migration in the second half of the year, with short-term focus on overseas computing power for gaming and short dramas, while waiting for grand narratives in the medium term

Among the strategy analysts in mainland China, Liu Chenming, the assistant director and chief strategy analyst at GF Securities, stands out. He ranked first in the strategy research category of the New Fortune Best Analyst Awards in both 2023 and 2024.

Liu Chenming's expression is clear, his logic is distinct, and his angles of approach and forecasting abilities are among the best in the industry, making him particularly favored by institutional investors.

On July 8, Liu Chenming rarely shared his latest strategic views in full at an institutional strategy meeting. He believes that there have been two major changes in the international and domestic macro environment over the past six months: the depreciation of the US dollar and interest rate cuts in mainland China, both of which constitute two potential funding reservoirs.

Looking ahead to the second half of the year, there is hope that the money from "deposit migration" in the domestic market will enter the market through actively managed equity funds, potentially forming a trend in the latter half of the year.

From a specific opportunity perspective, before the end of the third quarter and into the fourth quarter, the market may still focus on structural opportunities, primarily stable value and growth assets, with limited downward risk. After that, it will depend on whether the market can once again form a grand narrative consensus on the improvement of long-term corporate operational efficiency.

Key quotes:

1. In the first half of the year, two major "funding reservoirs" formed globally: one is the potential deposit reservoir created by the decline in domestic risk-free interest rates leading to deposit migration; the other is the outflow of funds from dollar assets due to the continued decline of the US dollar.
2. From the perspective of the mainland market, the inflow of funds is not obvious. There are signs of deposit migration in the first half of the year, but the magnitude is not particularly large. The number of new accounts, margin financing balances, and public fund growth have not seen significant increases.
3. There is no need to worry too much about the issuance scale of public funds, as historical patterns show that each round of public fund expansion comes about six months after a certain profit-making effect has been established.
4. This year, the inflow of funds into dollar assets has significantly decreased. Currently, the outflow of funds is mainly directed towards two areas: one is Europe, which has seen a rebound in manufacturing PMI and expectations of fiscal expansion; the other is gold. There is still considerable room for foreign capital inflow into domestic A-shares and Hong Kong stocks.
5. Attracting off-market funds will mainly depend on three aspects in the future: first, whether valuations are sufficiently cheap; second, whether ROE is confirmed to be on the rise; third, whether the grand narrative of rising forward ROE will be believed. All three scenarios can lead to significant upward movement in indices.
6. Looking at the short-term upward trend, the core factor remains PPI, as the majority of the index movements in China are still driven by cyclical assets, which are highly correlated with PPI.
7. Looking ahead to the second half of the year, if the macro trends can be realized, cyclical assets may rise to a new level. This corresponds to opportunities in Hong Kong internet and automotive stocks, as well as leading cyclical core assets such as brokerage stocks, liquor, etc., in A-shares.
8. The structural highlights of performance before August are in overseas computing power, revenue from gaming short dramas, export chains to Europe (motorcycles, sea wind inverters, etc.), and very niche domestic fields.
9. Among all A-share long-term price curves, the only two that have maintained a consistently upward slope are dividends and micro-disks The following text is written in the first person, with some content omitted.

Resonance of Two Major Changes at Home and Abroad

Since the first half of 2025, there have been two major top-down macro changes both domestically and overseas:

One major change is the continuous decline of the domestic risk-free interest rate, with various interest rates in the mainland dropping another level compared to last year. A landmark event was the one-year fixed deposit rate of major banks falling below 1% for the first time. The sustained decline in risk-free interest rates has led the market to continuously associate the possibility of residents moving their deposits, or bringing incremental funds to the equity market and the A-share market.

The second change is that the US dollar has entered a continuous decline process with almost no rebounds, dropping about 10% in the first half of the year. This has had a significant impact, causing a lot of funds to flow out of US dollar assets, forming a reservoir. There is hope that these overseas funds can return to Greater China and refocus on A-shares and Hong Kong stocks.

Thus, in the first half of the year, two major "fund reservoirs" were formed: on one side, the decline in domestic risk-free interest rates led to the potential movement of deposits, forming a possible deposit reservoir; on the other side, the continuous decline of the US dollar brought about the outflow of funds from US dollar assets, forming another reservoir.

Active Funds Have Not Yet Seen Significant Growth

But what has been the actual fireworks situation of these two reservoirs over the past six months? What hopes are there for the second half of the year?

Let's break it down.

First, in terms of deposit movement, there have been signs of movement in the first half of the year, but the extent is not particularly large.

In terms of the number of new accounts. The number of new accounts at the Shanghai Stock Exchange has remained consistent with past experiences, that is, it is basically in line with the market's profit-making effect and the index's performance. When the index rises sharply and the profit-making effect is particularly good, the number of new accounts increases accordingly. When the index slightly adjusts, such as in November and December 2024, or after the DeepSeek outbreak in March 2025 (coming to a pause), the number of new accounts will also decrease correspondingly.

Therefore, there has not yet been a formation of what is called excessive account openings. In other words, there has not been more residents' money entering the stock market through new accounts than during better market conditions.

Second, looking at the financing balance. The entire financing balance saw a significant wave of growth after "9·24" in 2024, with an increase of about 400 billion. However, from the beginning of 2025 to now, aside from a slight increase in the past week, the financing balance has remained flat without significant growth. In fact, the proportion of financing transactions has continued to decline.

So optimistically speaking, active funds have not yet seen a significant inflow, and there is still a lot of growth potential in the future.

The issuance of public funds is also a similar situation. In the first five months, the issuance of actively managed equity funds was still relatively average. In June, there was a concentrated increase, possibly due to the intensive issuance of more than twenty floating rate funds competing with each other, which indeed led to an increase in volume.

However, this is not a concern because historical patterns show that each round of public fund issuance comes about six months after public funds have formed a certain profit-making effect. For example, during the bear market in 2018, the public funds had a good profit effect at the beginning of 2019, but it wasn't until the end of the third quarter and the fourth quarter of 2019 that the issuance of public funds fully ramped up. This may take about half a year. Therefore, having seen good signs in the first half of 2025, the performance of public funds and actively managed equity products has begun to outperform ETFs, indices, and various broad or narrow-based A-shares, leaving room and possibilities for the subsequent migration of deposits.

Dollar Outflow Funds Flowing to European Stocks and Gold

Let's take a look at overseas funds, which are funds flowing out of dollar assets, or money that should have gone to dollar assets but instead flowed out.

The changes between 2025 and 2024 show a significant difference. In 2024, the money flowing into U.S. stocks under the same criteria was about $550 billion. However, in the first half of 2025, it was less than $160 billion, which is less than one-third of 2024.

Where has the money that should have flowed into dollar assets gone?

First, the most significant change is that it has flowed into Europe. Under the same criteria, Europe experienced a significant net outflow in 2024, while in 2025, it saw a significant net inflow. The underlying reason may stem from the continuous interest rate cuts in Europe, leading to a rebound in manufacturing PMI from a very low position for six consecutive months, coupled with expectations of fiscal expansion.

Secondly, it is easy to understand that it has entered gold.

We see that the Greater China region, including A-shares and Hong Kong stocks, has not attracted a significant inflow of foreign capital in the first half of the year.

Three Hypotheses Corresponding to Different Market Trends

Looking ahead to the second half of the year, we believe there is hope for the migration of domestic deposits. As the profit effect of public funds and actively managed equity products gradually improves and begins to exceed the indices, the money migrating from deposits may gradually form a trend in the market through actively managed equity funds in the second half of the year.

However, whether overseas money will return may depend on several changes. This reservoir is likely to see a significant increase in water volume in the second half of the year. As the probability of the Federal Reserve cutting interest rates in the second half of the year increases, there should also be a considerable amount of money in the overseas reservoir. Whether this money can come in may consider three scenarios.

First, compared to other markets, is it cheap enough?

Second, is there an expectation for ROE to rise?

This "upward" can be divided into two situations:

One is the reality of going up. For example, if the ROE level can rise quickly by the third or fourth quarter of 2025, it will definitely stimulate a lot of incremental funds to flow back, whether it is domestic deposit migration or foreign capital. Historically, there have been many typical cases, such as the cycles of 2006-2007, 2009-2010, 2016-2017, and the post-pandemic recovery cycle of 2020-2021. All four rounds were accompanied by the upward movement of Chinese indices, pro-cyclical assets, and the entry of foreign capital, as well as the increase of domestic off-market funds Secondly, although the current ROE performance is average, there exists a long-term unverifiable story, or a grand narrative logic, that makes people believe that the ROE will eventually rise. That is, assuming that the ROE level in the third and fourth quarters is not good, but people expect that next year, or in the longer term, the ROE will eventually go up, which can also attract domestic investment. There have been many historical cases of this.

Therefore, to attract external funds, whether it is a foreign capital reservoir or a domestic deposit relocation reservoir, it can be summarized into three aspects: sufficiently cheap; realistic upward movement of ROE; and a grand narrative of upward movement of ROE that is unverifiable in the long term.

As long as one of these conditions is met, it can bring about a significant level of index upward movement and inflow of external funds.

A-shares Valuation is at a Historically Cheap Range

Let's take a look at the valuation position of A-shares. Since the beginning of this year, the PB level has been maintained around 2.2 times PB. Compared to historical time series, this is relatively cheap, already positioned in approximately the last 30% percentile of history. This is attractive to domestic funds, as most domestic funds cannot flow out.

If we want to attract overseas funds, the ROE also needs to be compared on a country basis.

We summarized the corresponding relationship between ROE and PB at the historical equilibrium of major developed countries, such as the US, Japan, South Korea, the UK, France, and Germany, which can generally categorize global assets into three valuation tiers.

The first tier has an ROE level roughly between 6%-8%, with a PB level only at 1-2 times. Many industries in East Asian countries fall into this tier. This aligns with our intuitive feeling—East Asia has a large population, mainly engaged in manufacturing competition, leading to companies not making much profit.

The second tier has an ROE level between 10%-14%, with a PB level at 1.5-2.5 times. This has stepped up a level, with industries in Europe represented by the UK, France, and Germany in a similar situation. Europe has a relatively small overall population, lower levels of competition, and a focus on quality work. Therefore, Europe can achieve higher levels in certain top segments. Additionally, the service industry in Europe is relatively developed, and the inherent PB and ROE levels of the service industry are higher.

The third tier has an ROE level between 16%-20%, corresponding to a PB level of 2.5-4 times. Major industries in the United States are basically in this tier, including discretionary consumption, staple consumption, industrial, pharmaceutical, technology, etc. Companies in this tier possess technological barriers, brand premiums, globalization capabilities of multinational companies, and advantages in stock repurchase and cancellation. All these factors enable them to maintain a higher and more stable ROE level. This is somewhat similar to the result of the global industrial chain division of labor smile curve.

Financial and Dividend Advantages in Global Comparison

From a valuation perspective, the main focus for global investors is to assess the steady state of pricing. The latest ROE level in China is around 6.5%. Assuming that capacity stabilizes, PPI stabilizes, and ROE can also stabilize, a PB of 2.2 times is roughly reasonable from a domestic perspective, as domestic broad interest rates are declining and domestic funds cannot flow out However, from an overseas perspective, a 6.5% ROE corresponds to a 2.2 times Pb, which is actually not cheap compared to stable levels.

Of course, this is based on historical stability. Let's take a look at the latest situation, specifically the global PB and ROE levels as of last weekend. From the latest levels, Chinese assets still appear relatively cheap. At this position, the sectors that can be held long-term are still mainly financial and dividend-paying, at a cheaper position.

Other industry categories will rely on the judgment of ROE rising in the future, including technology sectors that need to see an order explosion.

This can be understood as two types of investment approaches. From a valuation perspective, if viewed from a foreign capital perspective, horizontal comparisons show that valuations are not cheap enough or sufficiently undervalued, thus requiring a judgment on ROE's upward movement.

The core focus in the short term is on PPI

Specifically analyzing the upward movement of ROE can be divided into two categories: short-term upward and long-term stories, macro narratives moving upward.

First, looking at the short-term upward movement, the core focus is still on PPI. Because the majority of the changes in ROE driven by China's indices are still cyclical assets, which are highly correlated with PPI. If we want to see ROE rapidly rebound from 6.5% in the second half of 2025, it may show a comprehensive upward trend.

The current situation of PPI is that the mid-point in the second quarter has slightly decreased compared to the first quarter. Looking ahead, PPI in 2025 may still be mainly supportive; the current core policy direction is more about cross-cycle adjustment rather than a true counter-cyclical approach. Policies need to leave room for the future without excessively over-leveraging.

Historically, we have also implemented many counter-cyclical measures, and if the intensity is sufficient, it can comprehensively raise PPI. Correspondingly, the increase in the fiscal share of GDP should quantitatively be at least 5 percentage points, which is a pattern observed in the previous 20 years across four instances.

For example, in 2007, it rose from just over 1 to just over 7, a 6-point increase; in 2009, from just over 3 to just over 8, a 5-point increase; in 2015 and 2016, the monetaryization of shantytown renovations saw an increase from just over 4 to just over 14, a 10-point increase over two years; and in 2020, in response to the pandemic, there was a 5-point increase in one year. All four instances are typical counter-cyclical measures, which correspond to a comprehensive rebound in PPI and ROE.

This time, we define it as cross-cycle. So far this year, the increase in ROE is about two points. If it reaches 5 points, it would basically align with the previous counter-cyclical policy state. However, this year still belongs to cross-cycle, primarily focusing on support.

This is our judgment on recent ROE and fundamentals: ROE can stabilize around 6.5%, but it is indeed quite challenging to see a rapid increase in the second half of the year, particularly in the third and fourth quarters.

Long-term macro narratives are worth paying attention to

We also need to look again to see if there are any new macro narratives (or short-term unfalsifiable logic) worth considering In the past 10 years, there have been two market trends driven by new macro narratives: one is the "9·24" in 2024, where everyone believes that housing prices will change and that the importance of stocks will change significantly.

In hindsight, over the past three quarters, ROE and PPI have not rebounded, but the index has risen by about 800 points. This is also similar to the influx of resident funds brought about by macro narratives, including foreign capital inflows.

The last time was during the state-owned enterprise (SOE) reform in 2014 and 2015.

At the end of 2014, the first batch of pilot lists for SOE reform was released, including state-owned capital investment companies like China Investment Corporation and COFCO, as well as state-owned capital operation companies like Chengtong and Guoxin, which were all pilots in the second half of 2014.

At that time, there was a belief that SOE reform could improve the operational efficiency of SOEs and central enterprises, thereby enhancing the logic and story of ROE.

Going further back, it also includes the first proposal of the Belt and Road Initiative, referred to as China's version of the Marshall Plan. Going abroad alleviated overcapacity, improved the ROE of traditional industries, and ultimately led to mass innovation and entrepreneurship, transforming the economic structure and improving ROE. All of these created expectations among the public for the potential upward movement of future long-term ROE, bringing in off-market funds and driving the index upward.

Looking ahead to the second half of the year, the probability of a macro narrative with long-term logic being believed by everyone is greater. Just like the "9·24" in 2024. The market believes that future ROE can improve, which can lead to a central upward shift in cyclical assets.

If the macro trend can be realized in the second half of the year, whether relying on central meetings or the 14th Five-Year Plan, it will provide a good long-term outlook for everyone, and cyclical assets may rise again.

Corresponding opportunities include Hong Kong stocks in the internet and automotive sectors, A-shares in brokerage firms, and even liquor, as well as leading cyclical core assets.

Therefore, in the second half of the year, before the end of the third quarter and the fourth quarter, the market may still focus on structural opportunities, with a primary emphasis on stable value and growth assets, while the downside risk is relatively limited. However, if we rely on the current fundamentals and policies, the first type of assets will need to wait for the larger macro narrative story to achieve rapid upward movement.

Where are the structural highlights?

In this process, from a bottom-up perspective, there are several details to pay attention to. Especially for the third type of growth assets, as we are approaching the mid-year report period, the latest main board disclosure rules require the release of half-year performance forecasts within 15 working days. During this period, including the peak of formal mid-year report disclosures in August, the relationship between stock price performance and current fundamentals is more closely correlated.

Therefore, since mid-June, we have been continuously communicating with researchers about the situation in specific industries. From an overall beta perspective, the overall performance during this period has been relatively average, with the PPI central growth rate around -3% in the second quarter, so we see more structural highlights.

Here are some bottom-up structural highlights: **overseas computing power, revenue from gaming short dramas, export chains to Europe (motorcycles, marine wind inverters, etc.), and very niche domestic fields **

This batch of companies can be roughly divided into three categories.

The first category consists of companies with very good current fundamentals (regardless of price, orders, or revenue), mainly concentrated in overseas computing power, gaming, copper, and aluminum, continuing to follow the trend.

The second category has average current fundamentals (mid-term report), but very good long-term expectations that cannot be falsified, including military trade, innovative drugs, and solid-state batteries.

The third category consists of major sectors with significant expectation differences. For example, domestic computing power. Looking ahead to the second half of the year, there are several things to pay attention to regarding domestic computing power: including the fixed price of Cambrian's private placement, the Hong Kong IPO of Tianzuo Zhixin, and the industrial chain orders of 910c. These are all points that may have new changes in the future.

Keep an eye on dividends and micro-accounts

Now let's take a look at dividend-type or stable-type assets.

Assets with long-term holding value must be cheap enough. Among all A-share long-term price curves, the only two that have maintained a consistently upward slope are dividends and micro-accounts.

This does not mean that the quality of micro-accounts and dividend companies is particularly good; the core point is that these two indices are the only two in all A-share indices that adjust positions based on buying low and selling high. This means that stocks that have risen significantly will be removed from the index, while those that have fallen significantly will be added to the index.

Other indices, such as the Shanghai 50, CSI 300, ChiNext, and STAR Market, all chase rising stocks and sell falling ones. Stocks must rise significantly to meet the criteria for inclusion in the index, and those that fall significantly are removed from the index, constantly chasing gains and cutting losses. Since most companies in A-shares are not cheap, the most effective strategy in A-shares remains to buy low and sell high, allowing for a more stable curve.

However, the volatility difference between dividends and micro-accounts is significant. Dividends have relatively low volatility, making them suitable for holding, and they correspond well to the concept of moving deposits.

On the other hand, micro-accounts have very high volatility and are only suitable for increasing positions after a round of crisis, as there are usually two instances each year where they drop by 20%. For example, in January, when annual report forecasts are released, they may drop by 20% during a black swan event in A-shares.

Risk Warning and Disclaimer

The market has risks, and investment requires caution. This article does not constitute personal investment advice and does not take into account the specific investment goals, financial conditions, or needs of individual users. Users should consider whether any opinions, views, or conclusions in this article align with their specific circumstances. Investment based on this is at their own risk