Mu Yiling: China's hard assets will be revalued, and consumption will not be poor; discussing the AI bubble now is meaningless

The Chief Strategist and Executive Deputy Director of SINOLINK SECURITIES Research Institute, Mu Yiling, shared his latest views on the market at an event hosted by the Franklin Templeton Investments, with the theme "China in the World - The Core Context of Equity Investment in 2026." Key Quotes: During the Spring Festival, there was a voice overseas: AI suddenly transformed from a technology filled with hope and the promise of future changes into a "monster" that everyone fears. The profits of large American companies are continuously rising, while the profits of medium-sized companies are fluctuating downwards, and small companies' profits are consistently declining, with employee salary levels also decreasing. Behind this is the increasing adoption of AI by large companies, which is growing faster compared to medium and small enterprises. The reduction in overseas employment will lead to weaker consumption; many shares of small and medium-sized enterprises will also be taken by large companies. In this process, inflation may decrease, giving overseas central banks reason to further cut interest rates, forming a cycle of "investment becoming stronger, consumption becoming weaker." If you pay attention to the U.S. stock market since January this year, you will see an interesting phenomenon: the sectors that have performed well this year are industrials, materials, energy, utilities, and consumer staples. This phenomenon is also related to the disconnection between investment and consumption. The real question is not whether there is a bubble in AI, as AI will continue to develop, but rather when it shifts from thematic investment to industrial investment, and then becomes the most decisive factor in the macroeconomy, which industries will be harmed? Which industries will benefit? This is what truly matters in 2026. After entering a rate-cutting cycle, commodity prices initially rise, investment demand increases, and funds flow back more easily; this return itself creates new demand for commodities, potentially forming a positive cycle. What is the key variable? Emerging markets need electricity for investment, and the U.S. needs electricity for AI. Since last year, a very obvious phenomenon has emerged in China's exports, with three areas maintaining strong export competitiveness. First, those directly related to AI; second, electricity indirectly related to AI; third, broadly defined equipment. Regarding the shift from exports to domestic demand, I believe that by 2026, this cycle will gradually open up. With the appreciation of the RMB, we will see a certain degree of coordination and unity between exports and domestic demand this year, so we should not be pessimistic about consumption. Adopting the first-person perspective, some content has been abridged. Overseas "Investment Becoming Stronger, Consumption Becoming Weaker" During the Spring Festival, you may have noticed a voice overseas: AI suddenly transformed from a technology industry filled with hope and the promise of the future into a "monster" that everyone fears, with various sectors overseas declining, all due to AI. What kind of macro impact will this phenomenon bring? What configuration opportunities will there be for our equity market? We will discuss what will happen in the future from a very interesting perspective - " The Macroeconomic Impact of Industrial Trends." First, you will find that since entering 2025, U.S. consumption has been declining while investment has been rising. Over the past year, many voices predicting a downturn in the U.S. economy have cited numerous data indicating "poor consumption." The optimistic voices about the U.S. economy often focus on capital expenditures driven by AI, with investments booming. A disconnection between investment and consumption is emerging. At the same time, the profits of large U.S. companies continue to rise, but the profits of medium-sized enterprises are fluctuating downward, and small businesses are seeing a continuous decline in profits, with employee wage levels also consistently falling. In other words, while large companies are improving, almost everything else is getting worse. Looking at the number of workers employed by large companies, you will find that this number is also continuously declining. Why are large companies, despite better profits, hiring fewer workers? Data tells us that large companies are increasingly adopting AI, and they are doing so at a faster rate compared to medium and small enterprises. Thus, we seem to see a cycle forming: large overseas companies are making more investments, and this investment is directed towards the field of artificial intelligence; increased investment in AI does not lead to job growth, but rather a reduction in new job creation. The decrease in overseas employment will lead to weaker consumption; many shares of small and medium-sized enterprises will also be taken by large companies, as AI is a process of "cutting the cake." During this process, inflation may decrease because wages are falling. At this point, the Federal Reserve will have reason to further lower interest rates, and lowering rates will allow large companies to further expand credit, creating a cycle of "stronger investment, weaker consumption." AI Reducing Costs for Overseas Society In the fourth quarter of last year, globally representative growth hedge fund manager Cathie Wood (Wood Sister) said something: You all say AI has no applications, how can AI have no applications? The cost of our entire American society is too high, and AI is helping us reduce costs. Think about what reducing costs means. A company's costs are the income of employees and suppliers—essentially, it is "eliminating" the income of others. Similarly, Kevin Warsh, a candidate nominated by the Federal Reserve who caused significant fluctuations in the financial market recently, also believes that the emergence of AI will lead to an increase in U.S. total factor productivity, which may cause inflation to decline. Looking at developed regions, many high-paying professions, such as doctors, lawyers, and accountants, are actually sources of social costs. The costs reduced by AI are essentially their incomes. Thus, a characteristic has formed: stronger investment and weaker consumption. Overseas Interest Rates May Decline Next, people will say that strong investment and weak consumption are unsustainable. Yes, it is unsustainable. But, over what time frame? Do you remember that there were countries that experienced 2 to 3 years of weakening consumption while investment remained strong, with electricity consumption growth consistently outpacing GDP during the same period? Now looking at the U.S., starting from the second quarter of 2025, electricity consumption growth is beginning to outpace GDP expansion, showing an identical situation. If you pay attention to the U.S. stock market since January of this year, you will also see an interesting phenomenon: the sectors that have performed well this year are industrials, materials, energy, utilities, and consumer staples. Does that sound familiar? It resembles China after 2022. Almost all the rising sectors are consistent In a certain sense, developing productivity means increasing investment; and "investment + consumption = savings" implies that some resources for consumption may be squeezed out. Productivity is developing, and the efficiency of the country and society may improve, but corporate profits may not necessarily be good. Thus, the United States is entering such a state: there are some more sensational articles online saying that many industries in the U.S. will be impacted, which is somewhat of a linear extrapolation. However, the macroeconomic impacts for the next 1-2 years are already slowly becoming visible. The prices of investment goods are rising, while consumption is relatively weak, and overseas interest rates may decline. Taking U.S. service sector inflation as an example, it has been relatively high since 2022, but is now declining; after interest rates drop, commodity inflation and investment goods inflation are still on the rise. Debating "Is There a Bubble in AI" is Not Very Meaningful Looking at the three lines related to investment in the U.S. fourth quarter GDP. AI-related investments (including information processing equipment, software, etc.) have been much discussed in recent years, but starting in 2024, the other two lines are declining. However, from the third to the fourth quarter of last year, after excluding AI and real estate, other manufacturing investments in the U.S. also began to rebound. Some say that gas turbines and other sectors are also related to AI, so it is still driven by AI. But I want to ask, just like the Chinese economy, does saying today "excluding new energy" still have significance for the Chinese economy? In fact, it no longer has significance; it has already become a part of the economy. When the development of the new energy industry in 2021-2022 affected various aspects of China's steel, chemical, metal, logistics, etc., it became a very important link in the economy and continues to spread. Historically, many economic cycles, when viewed from a macro-strategic perspective, consist of "quantity" (GDP) and "price" (CPI, PPI), but each round of industrial development is different. When new industries emerge, they become part of the macroeconomy. I believe that before 2026, debating "Is there a bubble in AI" is not very meaningful. Those who invest in technology say that AI does not have a bubble; this is an industrial trend, implying that valuations still need to expand; many macro analysts are more cautious, having risen so much, the bubble may burst. In fact, the real proposition is not whether AI has a bubble or not, AI will continue to develop, but when it shifts from thematic investment to industrial investment, and then becomes the most decisive influencing factor in the macroeconomy, which industries will be harmed? Which industries will benefit? That is what is truly important in 2026. An industry often reaches a certain stage of development, and people will look back to find the industrial chain, directing contradictions to the most scarce areas, rather than pushing forward along the "theme." The current characteristic of artificial intelligence is that this round of AI is about "cost reduction." On your end, there is currently no obvious increase in revenue (of course, it may also be "dividing the cake"). At the enterprise level, it will feel that this kind of cost (advantage) must be utilized with artificial intelligence. So far, there has not been any missed events related to AI capital expenditures, but the capital market pricing does not know how to price it, because it finds that once I use AI, other enterprises will be eliminated. There is no way to conduct "revenue pricing," and panic spreads Finding Tight Links Along Capital Expenditure At this time, a better approach is to find tight links along capital expenditure. We previously discussed that many sectors in the U.S. that are rising are providing supporting services for AI; overall, electric power stocks are also being driven by AI, as it has been discovered that the growth rate of electricity consumption in the U.S. will continue to outpace GDP expansion in the future. There are many aspects to this; traditional industries often design supply based on traditional economics, which has not been good in recent years, and suddenly there is new demand; there are also commodities, which we will discuss further later. Some derivative effects are also emerging. Since the beginning of this year, the PMI of major developed markets has risen. The U.S. is not to mention, and the industrial production activities in Germany and Japan have also significantly improved. Since December, the equipment investment or shipment volume of major economies has clearly increased. People might say that these have risen after interest rate cuts; won't there be inflation? It's still the same principle; this time, many investments (especially in the U.S.) leading to wage growth and service sector inflation will not hinder interest rate cuts. Capital Flowing Back to Emerging Markets Another point worth noting is emerging markets. Although everyone says that the supply chain will be rebuilt, the reconstruction of the supply chain in emerging markets during 2022-2023 seems to have "paused." What does this mean? The proportion of FDI (Foreign Direct Investment) to GDP has actually decreased during this round of interest rate hike cycle. The trend in these areas is good, but the cycle is weak. This round of emerging markets has two differences. First, many developing places have mines. For example, Indonesia, which has recently become very popular; even Zimbabwe and Congo have seen demands for "resource democracy." In the past, the demand was "mines should be nationalized, and I want to sell them at high prices"; this time, it is different. Previously, there might have been a rejection of foreign investment and a desire to hold onto the mines, which was referred to as the "resource curse"; but this time, it seems that everyone is saying you can invest, but I won't sell the raw ore; I want to sell the mid and downstream products; I need industry and construction. Therefore, we see that as Indonesia advances its nickel industry chain, "raw ore is not exported," but the production of nickel products and even batteries is expanding. In this process, the growth of electricity consumption continues to outpace GDP expansion, leading to a significant pull on demand from China. So at this stage, we see that after entering the interest rate cut cycle, commodity prices have risen, investment demand has increased, and capital is flowing back more easily; this flow itself creates new commodity demand, potentially forming a positive cycle. What is the key variable? Emerging markets need electricity for investment, and the U.S. needs electricity for AI. A phenomenon is beginning to emerge globally: with this round of interest rate cuts, capital is flowing back to emerging markets, and we can see that the growth of electricity consumption is starting to outpace GDP expansion globally. Everyone loves growth stocks. What are growth stocks? Broadly speaking, if an industry clearly grows faster than GDP, it is a growth stock; if it grows slower than GDP, it may be a value stock; negative growth may indicate a sunset industry. What grows faster than GDP? The most certain and resonant globally is "electricity." Electricity is what grows faster than GDP. Who can move China's "electricity" overseas? What opportunities are there for China? I remind everyone that China does not have a shortage of electricity; it is overseas that faces a shortage. Electricity prices in China may still be declining. So who can move China's "electricity" overseas, or use products made in China to solve overseas power generation issues, unstable grid problems, and electricity distribution issues? There are huge opportunities in that. Of course, this will also increase China's electricity consumption, but it's a different level. Why did we buy power generation companies in the past two years? Because at that time, Chinese companies were making investments, and our electricity was relatively tight, so power generation companies were better than electricity consumption companies. This round may favor electricity consumption companies more. There are many opportunities involved here, related to commodities close to electricity, such as copper and aluminum. Aluminum has another characteristic: the production side requires electricity, and the demand side is also related to electricity and AI, which is a very good direction. There is also China's new energy industry. Broadly speaking, China's manufacturing industry is also on this path, as electricity is one of the most important resources for China's manufacturing industry. In the past few days, I have also seen a statement that overseas computing power is expanding significantly, leading to insufficient electricity. Chinese companies can also sell abroad, calculating here, using China's electricity, and then "selling out" through fiber optic cables. No matter which industry you are in, whether it's growth stocks, tech stocks, consumer goods, manufacturing, or commodities, you can develop around this line, which may be a key focus in the near future. Since last year, a phenomenon has become very obvious in China's exports, with three areas maintaining relatively strong export competitiveness. First, directly related to AI; second, indirectly related to electricity; third, broadly defined equipment. This year's exports and domestic demand will form coordination and unity; we should not be pessimistic about consumption. At this point, everyone will realize that we share a similar view with the mainstream market perspective, which is to be bullish on China's exports. However, there is a very twisted phenomenon in the mainstream market perspective: when everyone believes exports will be good, they often say domestic demand will be poor. Why is there this contradiction? Because people say that housing prices have been falling, and the decline in housing prices leads to poor consumption; to solve poor consumption, housing prices need to rise; housing prices are difficult to rise, requiring government stimulus; without government stimulus, housing prices will have to force down; how to force down? Exports must be poor. It seems like a tight logical cycle, but it ignores one point: the impact of housing prices on consumption may be non-linear. Looking at a set of data, the decline in Japan's housing prices from their relative peak and the quarterly changes in consumption after removing trends (relationship). When the top 20% of prices fell, the negative correlation was very obvious: the more housing prices fell, the worse consumption became; but when the decline was between 20% and 40%, further declines in housing prices had less obvious effects on consumption. I joke that money is gone, but people are still here. At this stage, consumption is affected very little. Looking at the relationship between Chinese consumption and housing prices over the past two years. Since 2021, housing prices have fallen, and consumption has worsened; but since the fourth quarter of last year (2025), housing prices have not stabilized, and consumption seems to have stopped declining. This is the logic of "bottoming out." Is there a positive logic? Let me mention another thing: over the past two years, China's so-called "exports," people say exports have nothing to do with domestic demand. In fact, it's very simple: the money earned by export companies has not flowed back into the domestic market. We are a surplus country; goods go out, money comes in, but the "money coming in" part is missing If we subtract the international balance of payments surplus from the actual foreign exchange settlement, you will find that the "gap" in 2023 and 2024 is quite large. Clearly, companies are making money, but the money has not been settled under the foreign exchange settlement standards in banks. This phenomenon will start to correct itself in 2025. Note that this does not mean that the money that has already flowed overseas will return; there is no need to make such a strong assumption. We only need to believe that, as exports continue to improve or remain strong, there will be more "new money" coming back compared to 2023 and 2024. Because 2023 and 2024 are not the norm. Under optimistic expectations, the foreign exchange settlement volume, which accounts for 4%-5% of GDP each year, will return. Do you remember 2006 and 2007? During that bull market, this volume was 12% of GDP. At that time, China's economic scale was not that large, and the impact on domestic inflation was more pronounced. Back then, the focus was not primarily on stimulating real estate; real estate was more like a physical asset, and the increase in housing prices did not keep pace with the new construction area, resembling "better living," like consumer goods. What we saw was a shift from exports to domestic demand. I believe that in 2025 and 2026, this cycle will gradually open up, as the renminbi enters an appreciation channel, this state is beginning to emerge. This year, we will see a certain degree of coordination and unity between exports and domestic demand, so we should not be pessimistic about consumption. Since the fourth quarter of last year, we have already seen some recovery in high-end consumer goods. There are also details in consumption; in the third quarter of 2024, many micro-consumer companies experienced a cliff-like decline in performance, and distributors began to struggle and sell off inventory. The deflation or price decline we see at this stage is not entirely due to a collapse in demand, but rather because the industrial chain is undergoing a fierce destocking process. We expect that from the fourth quarter of last year (2025) to the first quarter of this year (2026), this will come to an end. It cannot be said that inventory will be replenished immediately, but once the inventory destocking is completed, inventory will shift from "supply" to "demand"; even if there is no replenishment, it will still show significant improvement. Therefore, the performance of consumer companies in the first quarter of 2026 is very critical. The macro turning point has long appeared, and the industry is waiting for this phenomenon. From the perspective of annual allocation, if there are investors present who have been waiting for 2-3 years on consumer stocks, it might be worth waiting a little longer, as we have come this far. We believe that domestic and foreign demand will coordinate and unify this year. Looking at Chinese consumption from a different perspective may be more reasonable. When people say "export settlement returning to boost consumption," it seems a bit convoluted. More simply, is it not the case that only places with wealth are qualified to consume? Returning to classical economics, what is wealth? Adam Smith's "The Wealth of Nations" tells us: productivity is wealth, it is a country's ability to create goods and provide services, not money, not management. In classical economic theory, who has the strongest productivity globally? I believe there are two. First, artificial intelligence is continuously developing; second, China on the physical side. Recently, Goldman Sachs released a report discussing "irreplaceable high-barrier assets," all of which are in Chinese manufacturing, and no one in the world can replace China. We possess the strongest productivity, which means we have the strongest wealth globally; it just needs to be realized through trade and money repatriation. The "illusion" brought by currency in the past was too strong, and we will gradually walk out of it. This is an important supplement to our unification of domestic and external demand. Industrial metals are the focus among commodities Let's talk about commodities. Many people are concerned about our views on commodities. By the end of 2024, the Federal Reserve will start to cut interest rates, and commodities will begin to rise. After taking the logarithm, the more expensive the value per ton of goods at the starting point at the end of 2024, the more it has generally risen in the past two years. This is very interesting. If you are an entrepreneur, you would tend to substitute the expensive: if you can use steel, you won't use aluminum; if you can use aluminum, you won't use copper; if you can use copper, you won't use silver. Under what circumstances would you instead buy the expensive? It's simple, suppose you have 30 million and say, "I want to buy metals to store," you would buy the metal with a higher value per ton. This corresponds to what everyone has been talking about for the past two years: "the collapse of the dollar system." We published a series of reports in 2022 discussing the dollar system and physical assets. I believe this is a process of more than a decade, but it has been excessively accelerated in 2024-2025. Those who study fundamentals will tell you that if this logic holds, gold and precious metals should be prioritized over industrial metals. Because industrial metals affect the real economy; if copper and aluminum rise too much, it will raise industrial costs; gold rising to the sky doesn't matter. This was correct in the past, but it may not be in the future. You can understand it as follows: the higher up, the more global M2 there is; the lower down means "the price of gold × reserves" is larger, and the total value is greater. Around the year 2000, we issued a lot of currency, but gold did not rise, as everyone trusted the dollar; however, since 2018, M2 has grown at an annualized rate of 8%, while gold has grown at an annualized rate of 20%. This line is infinitely close to the position around 1985 (which can also be understood as the phase after the collapse of the Bretton Woods system around 1975). Can gold return to the level of 1980? There may be room, but it is also facing resistance. When gold reaches 5,300 USD, it means that the scale of U.S. Treasury bonds held by central banks can no longer keep up with gold. Gold will not be consumed by industry, which is its downside, as it accumulates a large volume. If it rises further, a large amount of fiat currency will need to flow into precious metals, which will create a strong contraction effect at the level of the real economy; the medium is represented by interest rates on U.S. Treasury bonds, and rising interest rates lead to credit tightening, which will bring about restrictions and resistance from the government and various sectors. The earliest phenomena include Kevin Warsh wanting to reshape dollar credit upon taking office, limiting the ability to expand the balance sheet. Additionally, it was previously believed that the U.S. would suppress crude oil, but what the U.S. has done recently is unrelated to the decline in oil prices, but rather points to rising oil prices. The reason is simple: only when oil prices rise will Middle Eastern countries have more money to return to the U.S. to buy U.S. Treasury bonds. Who is under less pressure? Industrial metals. Industrial metals account for about 0.76% of the total market value of U.S. stocks, and inventory levels account for less than 20% of annual consumption, which historically could reach 100%. Why compare with the total market value of U.S. stocks? It's simple; when housing prices were very high, rebar, as the most important physical asset for houses, had a very large price elasticity. This round of technological development has been "choked" by industrial metals, which also have large elasticity. However, a significant portion of industrial metals still comes from traditional demand, leading to new and old demands constantly competing for territory until effective supply is stimulated. Old demand will definitely incur losses gradually, and this process may last for a long time So now the industrial attributes are more core; in the same year, the value of goods worth hundreds of billions of dollars has less impact on the real economy. At the same time, we see that industrial production activities have begun to recover. China's wealth will be revalued, and consumption will not be poor In conclusion, our logic roughly focuses on the following aspects. First, commodities. Those that are "choke points," whether industrial products or crude oil that supports the dollar system, as well as the oil transportation industry, can be focused on. Second, China's advantageous industry exports. Power grid equipment, energy storage, lithium batteries, engineering machinery, all starting from AI, and broadly speaking, China's hard asset industries (such as chemicals), gradually transitioning to consumption. Technological development continues, but it does not mean that the development of technology must be expressed through technology stocks. Technological development has formed an important impact on the economy, affecting monetary policy, traditional fields, and more supply-demand contradictions. Everyone talks about what hard assets are, right? These are hard assets. Similarly, China is the country with the most hard assets, and China's wealth will be revalued, and consumption will not be poor. Risk warning and disclaimer The market has risks, and investment needs caution. This article does not constitute personal investment advice and does not take into account the specific investment goals, financial situation, or needs of individual users. Users should consider whether any opinions, views, or conclusions in this article are suitable for their specific circumstances. Investing based on this is at your own risk