PPI minus CPI scissors gap widest at 4.1 percentage points! Nearly 500 billion in profits swallowed from mid- and downstream firms, the increase in corporate loans driven entirely by snatching bills, exports to the US off by 10 standard deviations in reconciliation, industrial electricity up 2.4% while value-added output up 5.2%! The clearing-out can only land on bankruptcies and unemployment

By cross-referencing the high-frequency economic data for the third quarter of 2026 released by the National Bureau of Statistics of the Chinese Communist Party (CCP), the central bank, and the General Administration of Customs, this article points out that official macroeconomic data uses aggregate figures to mask micro-level hemorrhaging. Upstream prices have been pushed up while downstream demand has collapsed, resulting in a record-breaking 4.1 percentage point scissors gap. Profits and cash flow for mid-to-downstream manufacturing enterprises have severely deteriorated, and the surge in corporate loans is actually driven by empty bill circulation. Real export data is severely decoupled from U.S. figures and re-export trade, the base for fixed asset investment has been modified across periods, and the true economic clearing can only manifest in business bankruptcies and unemployment.

Video link:

Full text:

Observe things as they are, and reject fragmented cognition. It has been quite lively these past two days, with high-frequency data being released quite densely. The CCP's National Bureau of Statistics, the central bank, and the General Administration of Customs have all released their September data. So, let's summarize the Chinese economy for the third quarter. The most intense piece of data is that the customs reported September exports at 447.18 billion USD, a year-on-year increase of 36.3%, which is the highest single-month value on record. Then there is the Producer Price Index (PPI), which rose 5% year-on-year, also the strongest month since this data turned positive. At the same time, the new social financing in September was 3.55 trillion. With these three press releases placed together, what narrative is the CCP trying to tell everyone? It is clearly a recovery script. But within the same batch of numbers, the new loans for residents in September were zero—not just a decrease, but zero. Moreover, after checking the data, I discovered another problem: the CCP's export data has decoupled from the import countries' data again, and it has reached a point where they cannot be reconciled at all. Through a series of data comparisons, I found something quite interesting: the CCP's macroeconomic data has learned to use aggregate figures to gloss over the results.

The more agreeable the surface appears, the more severe the microscopic blood loss becomes. Let's look at the paper-thin data from the Chinese Communist Party (CCP). First, let's talk about prices, because profit starts moving away from this point. To begin with, the inversion between PPI and CPI has deepened. In September, the ex-factory price of industrial products rose by 5% year-on-year, while the consumer price index (CPI) for residents only rose by 0.1% year-on-year, a gap of a full 4.1 percentage points. Since this scissors gap began to widen in April of this year, it has been growing for six consecutive months. The 4.1 figure in September is the widest month yet. I have mentioned before, what does a larger scissors gap indicate? It means that corporate costs are rising, but commodity prices are completely stagnant. Let's look at how lethal this 4.1 percentage point gap is! Now, let's step into the corporate ledger. In September, the ex-factory price of production materials rose by 6.5% year-on-year, while the ex-factory price of consumer goods was negative 0.1% year-on-year, a difference of 6.6 percentage points. Moving further upstream? Well, the ex-factory price of the mining industry rose by 25.4% year-on-year in September, coal mining rose by 35.3%, and oil and natural gas rose by 24.5%.

What about the prices factories pay for raw materials? The CCP's National Bureau of Statistics has an indicator called the Industrial Producer Purchase Price Index, which measures the cost of raw materials for factories. In September, it rose by 6.9% year-on-year, with fuel and power rising by 13.6%, and non-ferrous metal materials and wires rising by 20%. You might not feel these figures strongly, so let's apply this data to a real production environment. After substituting it, what does this cost inversion look like? Let's take an air conditioner manufacturing plant as an example. First, what do you have to do to start production? You have to buy copper. The cost of copper alone accounts for 20% to 30% of an air conditioner. Copper prices can rise by 20%, but does the air conditioner price dare to rise? In the first half of the year, the national retail sales of home appliances were only 425 billion, a year-on-year decrease of nearly 10%. The decline for air conditioners alone reached 15.9%. In the current environment, if you dare to raise prices, you will inevitably lose market share; if you don't raise prices, you will lose profit. This is what the home appliance companies themselves are saying.

We see that these home appliance companies are miserable enough, right? Well, vehicle manufacturers are even worse. In September, retail sales of passenger cars were only 1.702 million units, a year-on-year decrease of 24%. In other words, cars are simply not selling domestically anymore. The industry's own words are that upstream profits are surging, but price pressure is being transmitted from the upstream to the vehicle manufacturing end. So, this scissors gap in September is being driven by both ends. What is the upstream? It is imported cost shocks, plus the CCP's administrative production restriction order—that is, the No. 1303 document. What do these two things do? They hold up upstream prices. And what is the situation downstream? Downstream, demand has collapsed, suppressing prices. The entire structural state is: upstream prices are rising, there is no volume downstream, and the factories in the middle can only use their profits to cushion the blow. So, how much profit has been cushioned? I went through the profit statements of 41 industries from January to August from the National Bureau of Statistics line by line, and the total number still looks very good. From January to August, the profits of industrial enterprises above designated size were 5.2719 trillion, a year-on-year increase of 15.7%, which is 715.4 billion more than the same period last year.

But who actually earned this 715.4 billion? It was earned by the six upstream industries: coal, oil and gas, non-ferrous metals, chemicals, chemical fibers, and fuel processing. The profits of these six industries were 569.3 billion more than the same period last year. Another industry that made money was computer, communication, and electronics, which earned an additional 443.7 billion. Adding these two together gives 1.013 trillion, which already exceeds the total industrial profit increase of 715.4 billion. Meanwhile, the remaining 30-plus industries earned a combined 297.6 billion less than the same period last year. Simply put, what is this structure? The National Bureau of Statistics' report is for the upstream and chips, but the bill is being paid by the downstream. Let's just mention a few: there is no need to talk about real estate. The profit of the automotive manufacturing industry from January to August was 253.4 billion, a year-on-year decrease of 16%, and the sales profit margin was left at only 3.6%. Steel profits have fucking plummeted, down 62.4%. And industries like non-metallic mineral products such as cement, glass, and ceramics have also seen their profits basically cut in half, down 46.7%.

And the profit in furniture manufacturing also dropped by 51.8 percent. Adding up the manufacturing industries in the middle and lower reaches that experienced this profit decline, from January to August they earned a combined 280.1 billion less than last year. Coupled with electricity and thermal power earning 79.4 billion less, the total comes to 359.5 billion. This is the most typical process by which this scissors gap directly scoops profits away from the mouths of the middle and lower reaches. The profit accounts for September will not be released by the National Bureau of Statistics until October 27, but the prices for September are already out. We can also make a preliminary calculation based on these prices. Let us first look at consumer goods factories; what is the current situation? For every 100 yuan of operating revenue, the material cost is roughly 68 yuan. Let us use the raw materials purchased by consumer goods factories in September for our calculation, and take a weighted average of agricultural sideline products, chemical raw materials, textile raw materials, and fuel inputs according to this proportion. After calculation, the comprehensive cost rose by about 4 percent year-on-year. Then 68 multiplied by 4 percent is the extra 2.7 yuan in cost that must be paid for every 100 yuan of revenue. Meanwhile, the ex-factory price of consumer goods was negative 0.1 percent year-on-year, meaning that the selling price did not increase by a single cent.

With this coming and going, the profit per 100 yuan of revenue is thus squeezed down by another 2.8 yuan. The monthly revenue of the national consumer goods manufacturing industry is roughly 2 trillion, so 2 trillion multiplied by 2.8 percent is 56 billion. Let us look again at the equipment manufacturing industry excluding electronics; the current situation for this trade is that the profit per 100 yuan of revenue has been squeezed down to 2.5 yuan. Moreover, the monthly revenue of this trade is roughly 3.4 trillion, so the drop in profit amounts to 86 billion each month. Adding the two figures together, in the single month of September—under circumstances where factories could not raise prices at all—the profit directly swallowed by this scissors gap was around 140 billion. Adding the solid 359.5 billion less earned from January to August to this projected figure for September, in the past three quarters, the profits eaten up by the scissors gap in the middle and lower reaches are, by my estimation, between 430 billion and 500 billion. And this is still not the most serious part, because after all, profits are only on the books, and the hole in cash flow is even larger. What were the accounts receivable of industrial enterprises above the designated size at the end of August?

Let everyone guess: 29.48 trillion, a year-on-year increase of 9 percent. So can everyone see what the so-called newly added 715 billion in profit actually is? It is all fucking accounts receivable. Just the accounts receivable item alone surged by 2.5 trillion year-on-year compared to last year. Another item is the inventory of finished goods, which directly hit 7.36 trillion, also a year-on-year surge of 11 percent. Accounts receivable pressed an extra 2.5 trillion compared to the same period last year, and inventory pressed another 730 billion, which totals 3.16 trillion. This 3.16 trillion is money parked on the books that cannot be collected or sold, and this 3.16 trillion accounts for 60 percent of the total industrial profit of 5.27 trillion from January to August. So what is the situation for this bunch of factories in the middle and lower reaches today? They are being struck simultaneously by both profits and operating capital. This also explains why no one even bothered to borrow working capital loans in September. Having finished with prices, let us look at money. Newly added RMB loans in September were 1.08 trillion. What kind of time node is September? It is the end of the quarter. And what kind of time is the end of the quarter?

That is when banks rush to inflate their scale; hey, this is the rule, so the data must be compared with September of previous years. I have checked the newly added loans for every September since 2016, and there was not a single year lower than this year's figure. The last time it was lower than now, one has to go back to 2015. Let us look at the breakdown by sector; newly added household loans in September were zero, short-term loans decreased by another 90 billion, and medium- and long-term loans increased by 874 billion, with the additions and subtractions balancing out. In September of last year, household loans increased by 389 billion. Let us make a final tally for the first three quarters: cumulative net household loans decreased by 1.03 trillion, whereas in the same period last year there was a net increase of 1.1 trillion. In all previous years, household loans for the first three quarters were positive, but this year, for the first time, an entire quarter is entirely negative, so stop listening to bullshit about the real estate market warming up. Once everyone looks at this capital flow, everything becomes clear. On the enterprise side, it is even more aggressive; enterprise loans in September were 1.29 trillion, an increase of 700 billion compared to the same period last year, which hey, is a bright spot in Tonghuashun. But let us look at the structure: enterprises' short-term loans increased by 250 billion less than last year, medium- and long-term loans increased by 200 billion less, and bill financing was 50 billion less.

Last September it was minus 402.6 billion, and this year it's just that the degree of minus is 352.6 billion less than the same period last year. What does this mean? It means that the 70 billion more added by enterprises all came from the 352.6 billion reduction in commercial bills. If we take the bills away, the loans to enterprises increased by 270 billion less than last year. The bill sector is also very interesting. Let me explain to you, what is this bill stuff for? It's that enterprises issue commercial drafts of several months to a year to suppliers, and suppliers can take these bills to the bank to discount them, exchange them for cash, and the bank will treat these bills as loans and record them into its own credit scale. Why do banks still use bills now? Because at the end of every quarter, they have to assess this loan scale, and real loans cannot be lent out, so what to do? They collect bills in the market to pad this number, and you can see this by looking at the numbers. From July to August, the six-month bill discounting interest rate had already reached the floor price of 0.5. At the end of April, the one-month bill discounting rate once fell to 0, which is the most direct evidence of banks grabbing bills on a large scale.

In the first three quarters, enterprise bill financing accumulated to 1.24 trillion, while the same period last year was only 475.2 billion. The proportion of bills in enterprise loans rose from 3.5 percent to 9.9 percent, while enterprise medium and long-term loans in the first three quarters were only 6.53 trillion, directly 1.76 trillion less than 8.29 trillion in the same period last year. The bank's scale indicators are filled with bills, so what is the situation? The true investment willingness of enterprises is entirely in its, this is two sides of the same coin. Oh, by the way, such a low bill discount rate also leads to a more fun thing. What can it explain? It explains why China's deposits skyrocketed! Because I read the second-quarter monetary policy report of the People's Bank of China (the central bank of the CCP), how was it written? It was written that the weighted average interest rate of bill financing in June was 0.94 percent, while the weighted average interest rate of newly occurred time deposits in the same period was 1.28 percent, and that of three-year time deposits was 1.63 percent. If an enterprise uses a cost of 0.5 to 0.9 percent to issue and discount bills to get cash, and then turns around and deposits it as a one-year to three-year time deposit, there is an interest spread of 30 to 110 basis points in the middle, and it can completely make money lying down.

This is the process of capital idling, and not a cent has entered the workshop. The projection of this process on the data is M1. M1 is cash plus demand deposits, right? Didn't the CCP broaden this caliber last year? At that time, it was because the scissor-edge difference between M1 and M2 was getting wider and wider, and the CCP, fucking hell, shifted its mind upward: if I can't change the financing structure, can't I change this statistical caliber? And then it was changed, how was it changed? Personal demand deposits were also counted. With such a broad caliber, the growth rate of M1 dropped from 5.9 percent in February to 3.2 percent in September, dropping 0.9 percentage points in a month. And enterprise deposits in September increased by 1.31 trillion, but M1 was still falling instead. What does this indicate? It indicates that the money in the hearts of enterprises, fucking hell, all went into time deposits and didn't enter demand deposits at all. The money just walked from the banking system through the bill end to the deposit end, just not entering the real economy. So where did the skyrocketing deposits come from? Everyone understands now, right?

Then who is supporting the 3.55 trillion in aggregate financing to the real economy in September? It is still government bonds. The net financing of government bonds in September was 1.53 trillion, accounting to 43.1 percent of the newly added aggregate financing that month. In July, this proportion was 93.6 percent, in August it was 60.8 percent, and in the first three quarters, the net financing of government bonds had already reached 10.3 trillion. It sounds like the CCP government stepped up, but the aggregate financing in September only increased by 19.2 billion year-on-year, and it also formed a seesaw effect here, which is quite fun. The CCP government issued more than 300 billion more bonds here, and the private sector financing side was 300 billion less, and the addition and subtraction just evened out. And what is even more worthy of attention is the quota: the budget for the scale of newly added government debt for the whole year this year is 11.89 trillion, but 10.3 trillion has already been used up in the first three quarters. That is to say, the whole fourth quarter only has around 1.6 trillion left, so this year it is estimated that next year's quota will still have to be used. This is an arithmetic problem and doesn't require reasoning at all.

Does the CCP central bank have any tricks up its sleeve now? Not a damn one! To this day, the policy interest rate has remained unchanged for seventeen consecutive months. Now, they don't even dare to let the market expect an interest rate cut. Why? Because rate cuts are useless for the downstream. What is the weighted average interest rate for corporate loans? It is 3.04%. With that same 3.04%, for the energy sector where ex-factory prices are up 24.5% year-on-year, the real interest rate is negative 20% or more. Borrowing money at such times is practically free. But for an automaker whose ex-factory prices have fallen by 2% year-on-year, the real interest rate is 5%. The overall profit margin for industrial enterprises above designated size is only 5.66%; it looks higher than the loan interest rate, but that average is inflated by the upstream and chips. Look at automobiles, which are only at 3.62%, furniture at 2.21%, and agricultural by-products with a profit margin of only 0.76%; steel is even worse—zero. These industries also have slow capital turnover, turning over only about once a year, and their return on assets is only 2 to 4 points. This is either on par with or lower than the 3.04% loan rate. When the return on investment from borrowing money is already lower than the cost of borrowing, cutting interest rates in this situation will cause massive volatility in asset prices. The CCP is most afraid of price fluctuations right now.

Having finished with money, let's look at exports. Oh, this is the most eye-catching data for September. The CCP claims that September exports were $447.18 billion, up 36.3% year-on-year. Exports to the U.S. were up even more, by 39.6%. Integrated circuit exports were $45 billion, up 136%, accounting for 10% of that month's total exports. The inflation of the CCP's export figures is due to this state of increasing prices without increasing volume. I have already told you all about this. Besides, I reconciled the accounts, and oh, it gets even more interesting. I went to check the U.S. Customs records for imports from China, and this time, to shut the mouths of those claiming different statistical standards, I used a variance assessment method. Let's look at the historical data first; I used the four years from 2021 to 2024 as a baseline. The mean difference between the two sides' standards is about 14.8%, with a standard deviation of around 4.6 percentage points. However, starting last year, this discrepancy reached 36.1% for the whole year, deviating from the mean by 21 percentage points—this is 4.5 standard deviations. By the first eight months of this year, the gap in import/export standards between China and the U.S. reached 63.6%, deviating by 49 percentage points, which is 10.5 standard deviations. Under a normal distribution, the probability corresponding to 10 standard deviations is on the order of 10 to the power of negative 23. Come on, who can tell me if this can still be explained by statistical standards? Even if we double the standard deviation to 9 percentage points, this year's event is still outside of 5 standard deviations.

The Peterson Institute for International Economics also just released a reconciliation on October 2nd. Take a look: from January to July this year, Chinese customs reported exports to the U.S. of $258.1 billion, while U.S. records of imports from China and Hong Kong were only $185.3 billion. The gap reached $99.8 billion—basically $100 billion—which is equivalent to 63% of U.S. imports, and this is five times the historical normal data. U.S. records show imports from China from January to August were only $183.5 billion, a year-on-year decline of 16.3%, and August alone only saw a 6.9% increase. So when I saw Chinese customs claiming exports to the U.S. rose by 34.9%, I fucking crashed. Two mirror images of the same number: one curve says it dropped 16%, the other says it rose 40%. Can these two curves possibly be true at the same time? If anyone says this is still a matter of statistical standards, those bastards in statistics should be peeking through your windows every day. So where did the goods go? Oh, look at Vietnam, and it's exposed immediately. Let's look at the data from the General Statistics Office of Vietnam for the first three quarters: in the first three quarters of this year, Vietnam's imports from China were $187.3 billion, with a deficit against China of $121.4 billion, up 43%. Meanwhile, Vietnam's exports to the U.S. were $140 billion, with a surplus against the U.S. of $122.6 billion. Look at this structure: Vietnam's deficit with China and its surplus with the U.S. are growing almost at a 1:1 ratio. Is it possible for normal processing trade to be this symmetrical? This is the path where goods enter Vietnam from China, get a new country-of-origin label, and then go to the U.S.

Another set of data from the Peterson Institute is even more direct. From January to July, the U.S. recorded $149.6 billion in imports from Vietnam, while Vietnam reported only $105 billion in exports to the U.S. What is the $44.6 billion difference? These are goods that entered the U.S. with Vietnamese labels but never appeared in Vietnam's export records. This accounts for more than half of the excess U.S.-China trade gap, and this doesn't even include false declarations or tax rebate fraud. Imports have the same problem. What is being said? They claim that imports rose by 27.5% in the first three quarters, but retail growth in the same period was less than 1%. With domestic demand so weak, why are imports growing by double digits? It's still because of integrated circuits. In the first three quarters, the import value of integrated circuits rose by 56.6%, and imports of storage components rose by 1.5 times. At the same time, exports of integrated circuits rose by 99.1%. In other words, from the moment a chip is imported for packaging to the moment it is exported, both the import and export are recorded once, and both sides have doubled simultaneously, so this has absolutely nothing to do with domestic consumption.

Finally, let's look at the books of the National Bureau of Statistics of the CCP. This is the most solid piece of evidence in the September data. This data point is fixed asset investment. The completion amount announced by the CCP's National Bureau of Statistics for January to August was 29.3 trillion yuan, a year-on-year decrease of 7.2%. If I take this 29.3 trillion and divide it by (1 minus 7.2%), I can reverse-calculate that the base used by the Bureau to calculate the year-on-year growth this year was 31.58 trillion yuan. But what was the number announced by the Bureau itself on September 15 last year? It was 32.61 trillion yuan for the completion amount from January to August. For the same period of January to August 2024, they reported 32.61 trillion last year, but when used as a base this year, it only remains at 31.58 trillion. They secretly cut 1.02 trillion in the middle, a full 3.15% cut. If you compare the original announced figures from last year and 2024 directly, the real decline for January to August this year is not negative 7.2%, but negative 10.1%.

Why does the CCP's data look so exhausting now? Because the CCP's data now is really like a fucking 'Lao Liu' (a backstabber/troll), hitting you out of nowhere. So if you look at a piece of data, what else do you have to do? You have to simultaneously open three separate analytical paths for cross-comparison. Through the comparison between 7.2% and 10.1%, what does everyone see? It's just to prevent the two numbers from showing a decline in perception. Pushing back another year, the completion amount announced for January to August 2024 was 32.93 trillion. Comparing the original values of the two years, it dropped by 1% directly, but if you multiply the growth rates announced by the CCP for the two years, it is only negative 6.7%. The 4-percentage-point difference is the result of cutting the base twice. Do you understand why I say the CCP's data is a 'Lao Liu'? You can fucking change the same data across periods for two consecutive years, and then the growth increments, adjustment reasons, and details are not published at all. The mathematical significance of this operation is only one thing: to spread the collapse of the total amount into the growth rates of two years, so that the decline announced each year is 3 to 4 percentage points smaller than the real decline. Isn't that just it? The CCP's 5.3% unemployment rate follows the same logic. Hey, as long as you work for one hour during the survey week and get paid, you are considered employed. Then, starting in 2024, the unemployment rate for those aged 16 to 24 excluded students. Furthermore, once the migrant population with agricultural household registration returns home, they are removed from the urban sample. Look at these big moves; if you don't pay attention to the statistical process, who could distinguish these bullshit data from the CCP? I previously estimated that the median real urban unemployment rate is around 11%, and the September data gave me no reason to raise the credibility of the CCP's metrics.

Let us calculate using another independent path, looking at the account of secondary industry electricity consumption against the value-added of industrial enterprises above designated size. In 2024, secondary industry electricity consumption increased by 5.1%, corresponding to a 5.8% increase in industrial value-added—meaning one point of industrial growth consumed 0.88 points of electricity. From January to August this year, electricity consumption increased by 4.4%, and value-added was 5.3%, yielding 0.83 points of electricity per point of growth; cumulatively, these figures still seem to match up. Where does the problem lie? It lies in the single-month ledger for August. Secondary industry electricity consumption grew by only 2.4%, yet industrial value-added surged by 5.2%. This means one point of growth consumed only 0.46 points of electricity, effectively cutting the energy consumption intensity in half within a single month. Industrial energy consumption intensity is determined by the product structure of equipment; it is impossible for it to drop by half in a single month. Therefore, this 5.2% figure for August is either an artistic creation by the National Bureau of Statistics, or the September electricity data will have to make up for this gap. The National Energy Administration will release the September electricity consumption data at the end of the month, and we will know as soon as we cross-check them.

Now, let us put these ledgers together and discuss them. After cross-checking these data, we can see the situation very clearly. The upstream sector is collecting taxes through prices: coal mines, crude oil, and non-ferrous metals made an extra 569.3 billion, which is the resource tax bill of the Chinese Communist Party (CCP). And this tax bill is paid for using the profits and cash flow of the downstream sector. The 360 billion less in profits earned by the middle and downstream sectors, along with an additional 2.5 trillion in overdue accounts receivable, constitutes this payment record. On the residents' side, they are also paying their bills through early mortgage repayments. In the first three months, household loans decreased by a net of 1 trillion. On the side of the CCP government, it is relying on treasury bonds to single-handedly fill the social financing gap: the 43% share and the meager remaining quota of 1.6 trillion speak for themselves. None of these matters eased in September compared to August. Why is it said that this situation lacks a self-healing mechanism? Because what does normal deflationary clearing rely on? It relies on prices. When prices drop to the bottom, demand naturally returns. But right now, upstream prices are kept from falling by imported cost increases and production restrictions under Document No. 1303, while downstream prices cannot rise because they are suppressed by collapsing demand. Prices are choked at both ends, so clearing cannot happen through prices; it can only happen through business closures and rising unemployment. Furniture profits have been cut in half, steel profits have been essentially wiped out, and whole-vehicle profit margins continue to bleed—this is what the clearing process looks like. How can this situation be repaired? Either the downstream sector regains pricing power, or the upstream sector yields its profits. However, the September data show that both of these things are moving in the opposite direction. Therefore, in this cycle, where is the most dangerous position? It is downstream manufacturing and the creditors of downstream manufacturing.

After examining these accounts, I can draw a very clear conclusion: for every single penny of good news reported on the CCP's balance sheets, the profits of the people in the middle layer decrease by a penny. Once the National Bureau of Statistics releases whatever numbers, everyone can evaluate them for themselves. Alright, let's compare notes up to here for today.