Here is what happened. On August 28, the National Development and Reform Commission convened a national investment work conference. The orders were three: speed up special bond issuance and use, deploy the new policy-financing instrument, and accelerate the major projects of the 15th Five-Year Plan. Read against the balance sheet, it is less a meeting and more a departure gate opening.
Fast recap of the last 24 hours in fiscal terms, because the numbers move faster than the rhetoric. Vice Finance Minister Liao Min said the 800 billion yuan “two major” fund is now fully allocated, supporting 1,417 priority projects. By the end of July, 2.4 trillion yuan in new special bonds had been issued this year, backing more than 18,000 projects. And as of August 10, cumulative issuance stood at 2,544.8 billion yuan — 57.8% of the 4.4 trillion yuan annual quota.
The money, in short, has left the building. The question is what happens on the ground.
I tracked this story through the numbers because the headline — “funds allocated” — tells you nothing about whether a single pile is driven. In short, the real story of the third quarter is the distance between an appropriation and a poured foundation. That distance is the whole game.
The window is Q3, and it is not accidental
Why does the third quarter matter? Because construction is seasonal, and this year the calendar lines up with the money. Local governments plan to issue 1,372.6 billion yuan in special bonds in Q3 alone. That is the largest single-quarter tranche of the year, timed to hit the ground before winter weather freezes northern sites.
Combine the three flows and the logic is explicit. First, the 800 billion yuan “two major” fund is fully allocated — it supports national-level major projects and acts as seed capital. Second, special bonds are flowing at pace: 57.8% of the annual quota used by August 10, with a heavy Q3 push behind it. Third, the new policy-financing instrument — the successor to the 2025 vehicle — is being deployed through the same channels that proved it works.
Here is the number that anchors everything: in 2025, the 500 billion yuan policy-financing instrument is estimated to have leveraged roughly 7 trillion yuan in total investment. That is a 14x multiplier. Apply even half that ratio to the new tranche and you are talking about trillions in follow-on spending that has not yet appeared on any tender board.
The signal is in the sequencing, not the headline
The signal worth reading is not the size of the funds. It is the order in which the three instruments are being pushed. Fully allocate the “two major” fund first, then accelerate bond issuance, then deploy the instrument. That is deliberate sequencing: seed capital to start projects, bonds to sustain them, and the leverage instrument to multiply both.
There is no time to linger on the fiscal mechanics, so here is the bottom line. China is executing a synchronized capital push — central funds, local bonds, and policy leverage all firing in the same quarter. The machinery of public investment is at full throttle, and the throttle is set to the third quarter.
Let me think about why this quarter, specifically. The second quarter was dominated by two things: the tail end of the spring issuance season and policy announcements still in draft. The fourth quarter carries year-end execution risk — budgets that run long, weather that turns, and local governments that slow down to close their books. The third quarter is the only block of the year where the money is fully appropriated, the construction season is open, and the delivery targets are still ahead. If the system does not turn paper into physical work now, the arithmetic of the whole year changes.
What to watch: the gap between money and mortar
The honest analyst watches the money, then watches the ground, and never confuses the two. The gap between them is where the risk lives. A 57.8% issuance rate means 42.2% of the annual special bond quota is still unspent — roughly 1.8 trillion yuan waiting in the pipeline. Whether that backlog converts into physical work by year-end is the actual test.
What’s next, concretely: watch three things over the next 90 days. First, local government bond issuance totals for September — the Q3 plan of 1,372.6 billion yuan should show up in the monthly data. Second, construction starts on the 1,417 “two major” projects, which are now fully funded and have no excuse to wait. Third, the tender activity in cement, steel, and equipment — the sectors that physically build. When procurement data turns, the narrative turns with it.
I have been reading China’s infrastructure cycles for long enough to know the pattern: the announcement comes first, the money follows, and the physical work comes last, with a lag measured in quarters, not weeks. The 2025 experience is the template. The 500 billion yuan instrument produced an estimated 7 trillion yuan in total investment because the money was deployed into projects already waiting in the pipeline. The 2026 version is being deployed into a pipeline primed by the fully allocated “two major” fund.
That is why the third quarter matters more than the headline figures. The headlines are already written — 800 billion allocated, 2.4 trillion in bonds, 14x leverage. The story that remains unwritten is whether steel tonnage and earthmoving equipment confirm the paper. Fixed-asset investment is expected to stabilize through the second half; the third quarter is the window that decides whether that stabilization is a trend or a footnote.
What the mechanism actually guarantees — and what it does not
Let me be precise about the limits of the design, because precision is the difference between a signal and a slogan. What the mechanism guarantees is that money will reach projects: the “two major” fund is allocated, the bonds are issued, the instrument is deployed. What it does not guarantee is the rate at which money becomes physical output. That rate depends on local-government execution capacity, on land and environmental approvals, and on the contractors who actually pour the concrete. The capital side can be perfect and the ground side can still lag — which is why the third quarter is measured in construction starts, not in appropriation tables.
There is a second limit worth stating. The 4.4 trillion yuan special-bond quota is the headline, but 57.8% issued by August 10 leaves 42.2% — roughly 1.8 trillion yuan — still to be placed. A large share of that remaining balance will be spent in Q3 by design, but some of it will slip into Q4, where execution risk rises. The pipeline does not empty on a schedule; it empties at the speed of approval chains and procurement cycles. The design is sound; the timing is the variable.
And the leverage number deserves a second look as well. The 14x ratio from 2025 is a realized outcome, not a forward guarantee. It tells you the instrument is capable of mobilizing follow-on investment at that magnitude when the projects are ready. It does not tell you the new tranche will hit the same multiple, because the multiplier is a function of pipeline readiness, and the pipeline this year is competing with a wider set of priorities. The honest framing: the mechanism is proven, the multiplier is conditional, and the condition is physical readiness on the ground.
What the market should take from all of this is the signal embedded in the sequencing: the central government has chosen to spend its seed capital first and to time the bond and instrument flows to the same quarter. That is the signal. It says the authorities believe the third quarter is the window that converts appropriation into work, and they have loaded every instrument they control into that window. The rest is execution.
Why the multiplier deserves skepticism — and attention
Let me be honest about the 14x figure, because it is the most quoted number and the easiest to misuse. A 14x leverage ratio describes what the 2025 instrument did on a gross basis: 500 billion yuan of instrument, roughly 7 trillion yuan of total investment associated with it. It does not mean every new yuan deployed this year will reproduce the same ratio. The multiplier depends on project readiness, on how much of the financing the market itself absorbs, and on the health of the local-government borrowers who execute. The right way to read it is not as a promise but as a proof of mechanism: the channel works, capital can be multiplied, and the government is reusing the channel on purpose.
That is precisely why the sequencing matters. The “two major” fund was fully allocated first, which tells you the central government is front-loading the seed capital that makes the multiplier possible. The 1,417 priority projects are the ready-made pipeline. Special bonds are the sustainer — 2.4 trillion yuan already issued, backing over 18,000 projects. And the policy-financing instrument is the accelerator. Three instruments, three roles, one calendar. That is the design.
Reading it like a flight board
Think of the third quarter the way a dispatcher reads a departure board: a full row of flights showing “boarding,” not one flight that left early. The capital push is not a single event; it is a cluster of movements coordinated to arrive in the same window. The 800 billion yuan fund has boarded. The 2.4 trillion in special bonds is taxiing. The policy-financing instrument is scheduled next. What a dispatcher would watch is whether they leave in sequence without one flight blocking another — whether the seed capital, the bonds, and the leverage all convert to physical work without a jam in the middle. That is the Q3 question, and it is a logistics question as much as a fiscal one.
The reason the flight-board framing matters is that the failure mode here is not a lack of money. It is a lack of sequencing discipline: money that arrives faster than the projects can absorb it becomes parked capital, not work. The design, read correctly, is built to avoid that — seed first, sustain second, leverage third. If the sequencing holds, the physical work follows with the usual lag. If it jams, the numbers still look fine on paper while the construction sites wait. Which of those two outcomes actually occurs is what the next 90 days will decide, and it is why the ground data matters more than the appropriation data from here on.
The bottom line, fast
Signal first, context after: the capital side of China’s investment push is done. Every instrument is spent or in motion. The third quarter is now a physical-work story, and the data that will tell it is already on the calendar — September bond issuance, construction starts, equipment tenders.
What’s next matters more than the headline. And what’s next is 90 days of watching whether the 14x leverage bet turns into poured foundations before winter. No time to linger on the allocation numbers; they have already done their job. The only number left that matters is the one you cannot read in a press release — the one printed on a construction site.
In short: the money is out the door, the window is open, and the third quarter will tell you whether the machinery of public investment actually builds things. The signal is the sequencing; the test is the mortar. What’s next is the only question worth asking.