The global economy is entering a downturn phase of the Kondratiev cycle in 2026, a period that historically forces technological change. The world has entered a new era of intense competition in AI technology, with both Eastern and Western civilizations pouring their resources and futures into this pursuit, viewing it as humanity's closest approach to replicating divine intelligence.

Amidst the noise of modern capital and technology, funds are flooding into AI sectors like optical modules, storage, autonomous driving, brain-computer interfaces, and space technology. This has created a K-shaped economic recovery globally, while capital expenditure in traditional economies continues to decline. We see significant losses in commodities linked to traditional infrastructure and real estate, including PVC, rebar, photovoltaic soda ash, glass, iron ore, and coke. These losses have been prolonged and persistent.

The sustained losses in industrial product profits are unsustainable. Is there a potential turnaround trade opportunity? As the fervor to chase AI trends intensifies, can we maintain rationality? Are commodity prices mispriced? Even agricultural products like soybean meal, corn, sugar, and rubber are at historic lows. With the El Ni帽o weather pattern entering a peak phase in the second half of the year, could the agricultural sector resonate with the industrial sector to create an upward catalyst? Chemical futures contracts show significant backwardation, and rising crude oil production costs limit the downside. As the third-quarter peak season approaches, is there potential for restocking demand?

Where the Old Economy Strikes Back

Commodities are fundamentally driven by supply and demand, which are themselves determined by capital expenditure growth. A former head of commodities at Goldman Sachs, Jeff Currie, has argued that the commodity supercycle began in October 2020. Since then, capital has been flowing into AI tech giants like Nvidia, while energy and physical hard assets have been severely undervalued. Capital expenditure in hyperscale data centers is rising, but investment in the traditional economy is being abandoned by capital, quietly planting the seeds for the next capital misallocation. The typical pattern is that each new economic cycle ends with a glut of excess capacity, and the AI cycle will be no exception. From 2000 to the late 2010s, global oil production capacity was built. Between 2010 and 2015, the shale oil industry expanded capital expenditure, keeping crude oil and overall commodity prices low, resulting in low and stable inflation. Central banks then cut interest rates, ushering in a new economic cycle, which from 2020 onward was dominated by technology and energy. AI cannot function properly without ample energy and electricity, and without innovation, there is no progress.

In the 1960s, commodity prices were suppressed, and interest rates fell to 1.5%. This led to the rise of the "Nifty Fifty" stocks like Kodak and IBM. The US stopped investing in the old economy because it was cheap and offered poor returns. All capital chased the new economy, effectively strangling the old one. Then, in the 1970s, the old economy had its revenge. The commodity supercycle of that decade wasn't solely caused by the Arab oil embargo; the real seeds were planted in the early 1960s when investment was halted. Today, we see non-ferrous metals, particularly copper, hitting historic highs. This is also a result of the old economy's revenge. Between 2013 and 2016, capital expenditure by listed Chinese copper companies fell sharply, so when the new AI economy created demand, traditional supply was unable to meet it. The world is once again investing in a new tech cycle, repeating the pattern. This means that as capital expenditure in the traditional old economy declines, commodities linked to it will likely experience their own economic revenge.

Old Economy Profit Margins Have Hit a Limit

Looking at industrial products, we need to identify which parts of the old economy are seeing declining capital expenditure and assess if there are turnaround trade opportunities. If we view the economy as a whole, real estate is its largest and most indispensable component. Even the rapid growth in AI investment cannot fully replace real estate's impact on the overall economy. Real estate investment growth hit a new low in June, falling 18% year-on-year, reflecting a severe imbalance in the economy. While real estate faces challenges like a declining population, the wealth of most Chinese families, often spanning generations, is still tied up in property, which is now experiencing significant shrinkage. The industries behind real estate, from steel to chemicals, pipes to home appliances, are crucial for overall economic prosperity. Regarding infrastructure, we can look at production data for two key commodities. Building highways requires cement and asphalt. Asphalt production data shows a year-on-year decline, partially influenced by the Middle East conflict, but the overall trend still reflects a slowdown in infrastructure growth. Cement is a non-storable good, and its price index has fallen to a five-year low, indicating that infrastructure investment has not increased. This brings us back to Jeff Currie's point about declining capital expenditure in the old economy, with prices and production in a depressed state. However, trading is about the future and expectations. If the macro-traditional economy continues to weaken, could it trigger policy changes? In the real estate and infrastructure sectors, glass is a consumer-facing product. Its price has fallen below 2015 lows. With stricter environmental regulations, the shift from coal-based to natural gas-based production methods has further increased glass production costs. The operating rate for glass used in the photovoltaic industry, as well as float glass for construction and infrastructure, continues to decline. An industry operating rate consistently below 60% is unsustainable. It leads to further declines in factory capital expenditure and triggers a shakeout across the industry chain. The glass industry has reached a very low operating rate. We must ask ourselves: is it wise to persistently short a commodity that is unprofitable to produce and has a declining operating rate? Or, does glass priced below 1000 yuan still have long-term downside? This is a critical question for any investment decision. Everyone will have their own answer, considering risk-reward ratios, profit-loss ratios, and common sense. As the ancient text 吕氏春秋 says, "What is full must inevitably become empty; what reaches extremes must inevitably reverse; what is overflowing must inevitably diminish."

Agricultural Products Below Cost, Weather-Driven Catalysts Intensify

Turning to agricultural products, we see prices for live hogs, soybean meal, and even sugar falling below production costs due to declining consumer demand. Live hog prices have fallen below the cost of breeding, leading to a continuous decline in herd numbers. Even the most cost-efficient domestic producer, Muyuan Foods, faces challenging profit margins, which is forcing the industry to reduce capacity. The pig cycle has been in a prolonged loss phase since 2021, and the number of breeding sows has been declining since this year. We predict a nationwide drop in hog inventory starting in the fourth quarter. A bumper global soybean harvest has also pressured soybean prices, and massive domestic soybean inventories limit further upside for agricultural products. Corn and soybean meal are both at historically low levels. Soybean meal has remained below 3000 yuan per ton, and corn prices have nearly returned to 2000 yuan, which is essentially below production cost. Rubber and sugar are among the agricultural products most severely affected by El Ni帽o. This year, high temperatures have already caused sugar production to decline in India and Europe. Domestic rubber prices are also at historical lows. Based on research in Yunnan and Hainan, the current production cost for rubber is above 15,000 yuan, meaning current production is likely unprofitable. Rubber inventories are also declining, but prices remain depressed due to demand pressure from the automotive industry. If agricultural products remain unprofitable for a long period, it seems unsustainable, suggesting a tipping point may be near. As the El Ni帽o weather pattern develops further, this logic becomes more critical. If a fund could hold agricultural products for more than a year, would it currently be better to short or to go long? The El Ni帽o effect is ongoing, not only affecting normal agricultural production and causing heavy rain and typhoons in Guangdong, but also triggering heatwaves in Europe and India, leading to crop losses. Agricultural production, once lost, is difficult to restore quickly. From a weather-driven perspective, the potential for a breakout in agricultural products seems real. Furthermore, El Ni帽o impacts not just agriculture; it can affect crude oil production through typhoons and coal demand through heatwaves.

Low Inventories and Backwardation Limit Downside for Chemical Products

Coal prices have been rising recently, primarily due to declining inventories. The recent heatwave in East China has caused a surge in demand. Following a coal mine accident earlier this year, national production has declined further. Given the rapidly falling operating rate, the direction for coal prices during the peak demand season seems relatively clear and is at least cautiously optimistic. Returning to crude oil, despite AI improving shale oil extraction efficiency through automation and increasing production, US shale oil extraction costs have risen this year, and output from some wells has declined further. While US crude oil production has been increasing, rapid consumption of floating storage due to the war has accelerated the decline in global crude oil inventories. The US production increase cannot fill this gap. Even with a rising rig count, supply still falls short of demand. The inability to build inventories despite high production suggests strong demand. We often think of the pressure from new energy alternatives, but the fact remains that both gasoline and diesel demand are declining. September marks the start of the traditional peak gasoline and diesel demand season in the US and Europe, which also coincides with the peak hurricane season in the US crude-producing region. We have not seen a significant decline in crude oil prices, even after the Strait of Hormuz reopened. For chemical products, the current 2609 contract is trading at a significant premium to the 2701 contract, which is particularly favorable for long positions to roll over. This is indicated by the unusually high 9-1 spreads for many products. As we analyzed in a previous article, when inventories are low, a demand shock can lead to high price elasticity due to the lack of buffer stock. For example, PP has the highest basis among current chemical products, with the rollover potentially adding 800 yuan to the basis. Other chemical products like methanol, ethylene glycol, PP, and even PTA do not have high inventories. Ethylene glycol is particularly extreme, with inventories falling to around 400,000 tons. Methanol inventories are also at historical lows, partly due to the Middle East conflict. Low inventories naturally lead to high prices, which can suppress demand. Demand is indeed weakening, and building inventories seems difficult. Returning to the initial logic, the K-shaped economy driven by AI and the macroeconomic slowdown caused by real estate have put interest rates in a dilemma. From a demand perspective, lower rates are needed to support the economy. However, if rates are lowered, could these already cheap agricultural and industrial products rally further, causing demand to decline further? Central banks, including the Fed, are in a dilemma. Perhaps, as the trader Fu Haitang once said: "Treat high prices like dirt, and take low prices as precious jade." This means selling decisively when prices are high and buying aggressively when prices are low. Both stock and futures investments are games where only a minority wins. Contrarian thinking might be the first step to success.