#我的七夕交易分享 Macroeconomic Landscape and Asset Outlook Amid the Transition Between the Old and New Economies: Remain Bearish on Crude Oil, Gold, and Silver
When many people discuss Japan’s “lost three decades,” they habitually interpret it as the collective decline of the entire country. But that is not what happened. The losses were structurally distributed: ordinary people who took on heavy debt and bought property at high prices at the peak of the real estate bubble genuinely endured three decades of shrinking assets and stagnant wages; meanwhile, large capital that sold assets at high levels during the bubble used the opportunity to accumulate enormous wealth. For them, it was three decades of spectacular gains.
After the yen bubble burst, the yen appreciated sharply, further magnifying the overseas purchasing power of this capital. Holding enormous wealth in the local currency, and with the domestic environment of persistently low interest rates, capital could borrow cheaply at home and make large-scale overseas acquisitions of high-return assets, driving the scale of Japan’s overseas assets to expand continuously for decades.
By contrast, within Japan, household wealth shrank while debt burdens mounted, consumption continued to contract, and returns in the domestic real economy declined. Even though the prices of production factors became cheaper, capital lacked the incentive to expand in traditional domestic industries.
A similar logic is now unfolding domestically. Large numbers of residents bought homes at high prices during the upswing in the property cycle and took on long-term debt. Domestic demand will need to digest these liabilities for a long time in the future. Meanwhile, large capital that cashed out and exited at the cycle’s peak has already accumulated substantial wealth, with some assets allocated overseas. This can also explain why, even when the exchange rate strengthens, funds have not flowed back into the domestic capital markets.
Capital that remains domestically is likewise unwilling to invest in traditional industries. End-user demand in traditional industries comes from ordinary residents. Against the backdrop of household debt not yet being fully worked off, consumption continues to shrink, making it difficult for traditional businesses to generate substantial returns. As a result, large amounts of capital are flowing into the AI technology sector, while global capital is likewise betting on the productivity transformation brought about by AI.
But the AI growth story faces an unavoidable underlying constraint: regardless of how much technology improves production efficiency, capital investment must ultimately be completed through end-user consumption by ordinary people; otherwise, enormous investments will be unable to generate cash flow returns.
At this stage, AI is concentrated mainly in B2B enterprise services, computing power, and large-model software. It has yet to produce a large-scale durable consumer product for millions of households, comparable to automobiles, smartphones, or housing. Without such a mass-market consumer vehicle, AI cannot fully transmit the benefits of industrial growth to ordinary residents. Therefore, large capital will continue positioning itself in the AI sector until AI end-user products capable of igniting mass consumption are truly deployed on a large scale.
At the same time, household debt reduction is a prolonged slow-moving variable. Starting from 2021, the entire debt-repair cycle will most likely take close to ten years. Only after households gradually complete debt repayment, combined with a new generation no longer bearing the pressure of high housing prices, will household consumption capacity be released once again.
In short, two major conditions need to be met simultaneously: first, AI must produce large-scale consumer products for civilian use; second, domestic household debt must be largely worked off. Only when both conditions are met can the dividends of AI technology be transmitted throughout society and the overall economy return to expansion. Before then, traditional industries lack a basis for bottom-fishing. But the path to technological deployment in the AI sector itself is full of uncertainty, making it difficult to accurately bet on which specific subsector or company will ultimately prevail.
Now let us turn to the commodities market. Crude oil has remained elevated amid geopolitical conflicts, objectively playing the role of a “targeted rate hike.” High oil prices raise production and logistics costs for traditional industries, compressing traditional-sector profits and forcing existing capital to flow out of the old economy and migrate toward the new AI economy; at the same time, high oil prices drive up inflation expectations and US Treasury yields, also completing an internal bubble cleanse within the AI sector by eliminating purely narrative-driven speculation and leaving leading companies with real orders and realizable profits.
And this mechanism has now reached a stage-specific inflection point.
The ten-year US Treasury yield has already risen to around 4.6%, while the 30-year US Treasury yield has exceeded 5.2%. If oil prices remain elevated, long-term yields will rise further. Traditional industries already face weak demand and have no more funds available to flow out; excessively high interest rates would instead backfire on the AI ecosystem, raising financing costs for numerous AI application companies and hindering commercialization, which does not serve the overall interests of capital.
Confirmation can also be seen at the market level: various AI subsectors have experienced significant volatility, with themes such as large models and optical modules undergoing pullbacks and their bubbles being sufficiently compressed; meanwhile, some large-model companies have begun reporting solid profitability data, while semiconductor and model-inference costs continue to decline, favoring further expansion of AI applications.
Therefore, the current economy has an inherent need for interest rates to decline moderately, and crude oil already has the momentum to retreat. Geopolitical developments will of course cause short-term pulse-like disruptions, but over the medium to long term, as conflicts remain at a low-intensity stalemate, the market will gradually become desensitized to the risk premium, and the center of oil prices will return to fundamentals.
Now let us look at gold and silver. The current elevated levels of precious metals are mainly driven by the safe-haven premium arising from the downturn in the traditional economy, along with support from geopolitical conflicts. In the future, as the old industries continue to be cleared out and the new AI economy continues to grow, risk appetite will rise, and funds will continue shifting from safe-haven assets toward high-return technology assets; at the same time, improving economic conditions will push up real interest rates, raising the opportunity cost of holding non-yielding precious metals. Gold and silver will therefore face sustained downward pressure. With weak traditional industrial demand added to the equation, silver will have greater price elasticity. Of course, global central-bank gold purchases will provide some downside support, making a one-off collapse unlikely; a prolonged, volatile decline is more probable.
Overall, several major trends for the future are already relatively clear: AI is the long-term main theme of the new economy, but it is difficult to pinpoint specific subsector opportunities; traditional industries have no bottom-fishing value; and crude oil, gold, and silver have a medium- to long-term downward logic.$XAGUSD
When many people discuss Japan’s “lost three decades,” they habitually interpret it as the collective decline of the entire country. But that is not what happened. The losses were structurally distributed: ordinary people who took on heavy debt and bought property at high prices at the peak of the real estate bubble genuinely endured three decades of shrinking assets and stagnant wages; meanwhile, large capital that sold assets at high levels during the bubble used the opportunity to accumulate enormous wealth. For them, it was three decades of spectacular gains.
After the yen bubble burst, the yen appreciated sharply, further magnifying the overseas purchasing power of this capital. Holding enormous wealth in the local currency, and with the domestic environment of persistently low interest rates, capital could borrow cheaply at home and make large-scale overseas acquisitions of high-return assets, driving the scale of Japan’s overseas assets to expand continuously for decades.
By contrast, within Japan, household wealth shrank while debt burdens mounted, consumption continued to contract, and returns in the domestic real economy declined. Even though the prices of production factors became cheaper, capital lacked the incentive to expand in traditional domestic industries.
A similar logic is now unfolding domestically. Large numbers of residents bought homes at high prices during the upswing in the property cycle and took on long-term debt. Domestic demand will need to digest these liabilities for a long time in the future. Meanwhile, large capital that cashed out and exited at the cycle’s peak has already accumulated substantial wealth, with some assets allocated overseas. This can also explain why, even when the exchange rate strengthens, funds have not flowed back into the domestic capital markets.
Capital that remains domestically is likewise unwilling to invest in traditional industries. End-user demand in traditional industries comes from ordinary residents. Against the backdrop of household debt not yet being fully worked off, consumption continues to shrink, making it difficult for traditional businesses to generate substantial returns. As a result, large amounts of capital are flowing into the AI technology sector, while global capital is likewise betting on the productivity transformation brought about by AI.
But the AI growth story faces an unavoidable underlying constraint: regardless of how much technology improves production efficiency, capital investment must ultimately be completed through end-user consumption by ordinary people; otherwise, enormous investments will be unable to generate cash flow returns.
At this stage, AI is concentrated mainly in B2B enterprise services, computing power, and large-model software. It has yet to produce a large-scale durable consumer product for millions of households, comparable to automobiles, smartphones, or housing. Without such a mass-market consumer vehicle, AI cannot fully transmit the benefits of industrial growth to ordinary residents. Therefore, large capital will continue positioning itself in the AI sector until AI end-user products capable of igniting mass consumption are truly deployed on a large scale.
At the same time, household debt reduction is a prolonged slow-moving variable. Starting from 2021, the entire debt-repair cycle will most likely take close to ten years. Only after households gradually complete debt repayment, combined with a new generation no longer bearing the pressure of high housing prices, will household consumption capacity be released once again.
In short, two major conditions need to be met simultaneously: first, AI must produce large-scale consumer products for civilian use; second, domestic household debt must be largely worked off. Only when both conditions are met can the dividends of AI technology be transmitted throughout society and the overall economy return to expansion. Before then, traditional industries lack a basis for bottom-fishing. But the path to technological deployment in the AI sector itself is full of uncertainty, making it difficult to accurately bet on which specific subsector or company will ultimately prevail.
Now let us turn to the commodities market. Crude oil has remained elevated amid geopolitical conflicts, objectively playing the role of a “targeted rate hike.” High oil prices raise production and logistics costs for traditional industries, compressing traditional-sector profits and forcing existing capital to flow out of the old economy and migrate toward the new AI economy; at the same time, high oil prices drive up inflation expectations and US Treasury yields, also completing an internal bubble cleanse within the AI sector by eliminating purely narrative-driven speculation and leaving leading companies with real orders and realizable profits.
And this mechanism has now reached a stage-specific inflection point.
The ten-year US Treasury yield has already risen to around 4.6%, while the 30-year US Treasury yield has exceeded 5.2%. If oil prices remain elevated, long-term yields will rise further. Traditional industries already face weak demand and have no more funds available to flow out; excessively high interest rates would instead backfire on the AI ecosystem, raising financing costs for numerous AI application companies and hindering commercialization, which does not serve the overall interests of capital.
Confirmation can also be seen at the market level: various AI subsectors have experienced significant volatility, with themes such as large models and optical modules undergoing pullbacks and their bubbles being sufficiently compressed; meanwhile, some large-model companies have begun reporting solid profitability data, while semiconductor and model-inference costs continue to decline, favoring further expansion of AI applications.
Therefore, the current economy has an inherent need for interest rates to decline moderately, and crude oil already has the momentum to retreat. Geopolitical developments will of course cause short-term pulse-like disruptions, but over the medium to long term, as conflicts remain at a low-intensity stalemate, the market will gradually become desensitized to the risk premium, and the center of oil prices will return to fundamentals.
Now let us look at gold and silver. The current elevated levels of precious metals are mainly driven by the safe-haven premium arising from the downturn in the traditional economy, along with support from geopolitical conflicts. In the future, as the old industries continue to be cleared out and the new AI economy continues to grow, risk appetite will rise, and funds will continue shifting from safe-haven assets toward high-return technology assets; at the same time, improving economic conditions will push up real interest rates, raising the opportunity cost of holding non-yielding precious metals. Gold and silver will therefore face sustained downward pressure. With weak traditional industrial demand added to the equation, silver will have greater price elasticity. Of course, global central-bank gold purchases will provide some downside support, making a one-off collapse unlikely; a prolonged, volatile decline is more probable.
Overall, several major trends for the future are already relatively clear: AI is the long-term main theme of the new economy, but it is difficult to pinpoint specific subsector opportunities; traditional industries have no bottom-fishing value; and crude oil, gold, and silver have a medium- to long-term downward logic.$XAGUSD
