Follow the paper trail; it always leads somewhere. On August 28, the National Development and Reform Commission convened a national investment work conference with a short, specific agenda: accelerate the issuance and use of local special bonds, move quickly to deploy the new policy-based financing tools, and speed up the major projects in the 15th Five-Year Plan. Read that as the message on the front of the envelope. The receipts inside tell a more interesting story.
The receipts start with the biggest single number in the file. Vice Finance Minister Liao Min confirmed at a State Council Information Office briefing on August 21 that the full ¥800 billion of two-fold heavy funds — the major national strategies and security capacity building programmes — has been allocated, supporting 1,417 key projects in key areas. That is not a commitment or a plan; that is a completed transfer. The money has left the treasury and is in the pipeline.
The second receipt is the special-bond schedule. By the end of July, local governments had issued ¥2.4 trillion of 2026 special bonds, supporting more than 18,000 projects. By August 10, cumulative issuance had reached ¥2.5448 trillion — 57.8 percent of the ¥4.4 trillion annual quota. In the third quarter alone, regions plan to issue ¥1.3726 trillion. Lay those numbers on a timeline and the shape is unmistakable: the second half of the year is where the heaviest volume lands.
The 57.8 percent draw by August 10 deserves to be read alongside the Q3 plan of ¥1.3726 trillion, because the two numbers together describe the pace the authorities intend to keep. Reaching 57.8 percent in roughly the first eight months of the year means the remaining 42 percent of the quota has to be issued in the last four months — a deliberately front-loaded schedule that is exactly what the construction calendar rewards. Issuance concentrated in Q3 lands in the months when projects can actually be built at full speed, before the winter slow-down arrives. The sequencing is not an accident; it is the design of the fiscal year.
Now follow the chain from allocations to actual work. Money allocated is not money spent; money spent is not yet a completed project. The conversion from treasury transfer to steel, concrete and electrical installation is what the market calls the formation of physical work volumes. This is the step where paper trails go dark, and it is precisely the step that the August 28 conference was trying to accelerate. The instruction to move quickly is an acknowledgement that the bottleneck is not funding approval — it is the conversion speed.
The phrase golden quarter for construction is not market colour; it reflects the physical logic of the work. Warm weather extends the working day on sites, concrete cures faster, and labour availability peaks because the agricultural season has finished. A project started in June has a real chance of reaching substantial completion by the end of the year; the same project started in November will sit half-finished through the winter, tying up capital with nothing to show for it. That is why the Q3 window is the hinge of the entire investment programme: the money has to reach the site while the weather still cooperates, or the physical work volume slips into next year.
The leverage arithmetic is where the paper trail gets genuinely interesting. In 2025, ¥500 billion of policy-based financing tools are estimated to have leveraged total investment of around ¥7 trillion — a multiplier of roughly 14 times. The 2026 round is being deployed on top of an existing ¥800 billion base of fully-allocated two-fold heavy funds and a special-bond pipeline running at 57.8 percent. The question is not whether the leverage can work; the 2025 record shows the mechanism. The question is whether the physical work volume can be formed within the Q3 window that the calendar defines as the golden quarter for construction.
It is worth noting the funding architecture, because it shows a deliberate layering. The ¥800 billion of two-fold heavy funds are the strategic layer, aimed at named national projects with defined security and capacity objectives. The special-bond programme is the workhorse layer, financing the broad municipal project base — the ¥2.5448 trillion drawn so far has touched more than 18,000 projects. And the new policy-based financing tools are the leverage layer, designed to pull private and local money behind the public commitments, with the documented 14-times multiplier from the 2025 round as the reference point. Three layers, three functions: targeting, breadth and leverage.
Here is what the conference calendar suggests. The three-month window from July to September is when weather conditions, funding flows and the fiscal year’s rhythm align. The deliberate scheduling of the August 28 conference — with special bonds at 57.8 percent and Q3 planned issuance of ¥1.3726 trillion still ahead — is not a coincidence. It reads like an attempt to keep the issuance curve inside the construction window, so that allocated money becomes physical work before the year’s end. I have seen this pattern before in other investment cycles, and the tell is always the same: the conference dates cluster around the months when the weather and the budget calendar cooperate.
The implementation question is where the paper trail has to be followed with the most care. The conference’s instruction to accelerate issuance is, in effect, an instruction to compress the time between approval and physical work. That is harder than it sounds: it requires project readiness, land and permits settled, financing structures agreed. The number of supported projects — 1,417 for the two-fold heavy funds, more than 18,000 under special bonds — suggests the pipeline is deep. What remains to be proven is the speed at which that pipeline converts. In past cycles, the gap between allocation and construction has been the place where announced stimulus quietly failed to become activity.
One document is an accident; six are a pattern. And there are more than six documents here. The ¥800 billion fully deployed. The ¥2.4 trillion issued by end-July supporting 18,000 projects. The 57.8 percent draw on a ¥4.4 trillion quota. The ¥1.3726 trillion of planned Q3 issuance. The 14-times leverage documented on the 2025 tools. The 1,417 key projects. Each one is an allocation; together they describe a system pointed at one outcome: fixed-asset investment stopping its slide and stabilising for the year.
Let me correct my own frame before I overstate it. I have been writing the money is moving as if that settles the question, and it does not. Allocations describe intent; physical work volumes describe outcome. The gap between them is where the story will actually be decided. A fully-allocated ¥800 billion is a necessary condition, not a sufficient one. What the market will watch — what it should watch — is the Q3 construction data and the industrial output tied to infrastructure: cement, steel rebar, heavy machinery utilisation. The receipts that matter are the ones stamped by construction sites, not by finance departments.
The accountability angle cuts both ways, and this is where the report’s own discipline should hold. The trail is unusually transparent so far — the numbers are on the public record, dated and attributable: the August 21 briefing, the August 28 conference, the August 10 issuance tally. That transparency is itself the story for anyone who has watched past cycles. What gets announced and when, how quickly funds clear, whether the physical work follows the schedule — these are the audit trails of policy credibility. And audit trails, in my experience, are where the real condition of things is revealed.
For the investor or the contractor reading the file, the practical reading is simple. The pipeline is loaded: ¥2.5448 trillion drawn against a ¥4.4 trillion quota, with a concentrated Q3 schedule ahead. The leverage machinery is documented to have worked at 14 times in 2025. If the physical-work formation holds through Q3, the stabilisation of fixed-asset investment stops being a forecast and becomes a statement. If the conversion stalls — if allocations remain on paper while sites stay quiet — then the ¥800 billion will look less like a stimulus and more like a placeholder.
The confirmation signal, when it comes, will not be a single number but a cluster. Watch the monthly fixed-asset investment print, watch the construction-start series, watch the procurement orders for cement and machinery that sit behind the aggregate. If the Q3 numbers show the physical work volume forming in line with the issuance schedule, then the full-year fixed-asset investment target stops being a hope and becomes a calendar event. If the volume lags the allocations by more than a quarter, the story will need a different chapter. The receipts, as always, will decide — not the announcements.
I keep a file on these programmes, because the difference between announced and delivered is the difference between an intention and an outcome. The current file is well-documented and dated; the transparency is itself a change from earlier cycles. What I cannot yet file is the site-level evidence. That arrives in the autumn data. I will be reading it with the same attention I give the allocation numbers, and the verdict will follow the chain.
One caveat about the autumn data: construction statistics are noisy, and a single month can mislead. The correct habit is to read two or three consecutive prints together, and to read them against the issuance calendar rather than against the calendar year. If the Q3 numbers land within the range that the ¥1.3726 trillion of planned issuance implies, the direction is confirmed; if they lag by a wide margin, the gap between allocation and delivery has grown, and that is itself the finding. Either way, the data will answer the question the August 28 conference raised, and the file only has to record it.
The discipline of this kind of reporting is simple to state and hard to hold: the conclusion follows the chain, not the headline. When the allocation numbers were announced in August, the easy story was that China is stimulating again; the harder story — the one this file is trying to keep straight — is what the money does between the treasury statement and the construction site. Every programme of this scale has a paper life and a physical life, and the two are often quarters apart. The chain from announcement to allocation to procurement to ground-breaking to completion has checkpoints where the story can stall, and each checkpoint has its own documents. The audit is in the sequence.
The paper trail always leads somewhere, and this one leads to a specific place: the third quarter of 2026. The receipts we already have say the money is real, the schedule is front-loaded, and the mechanism has a demonstrated multiplier. The receipts we do not yet have are the ones that matter — the physical volumes, the construction starts, the equipment orders. Follow the chain that far, and the year-end fixed-asset investment number will have already written itself. The only open question is whether it writes itself in the affirmative.
Let me add one more layer of evidence that I keep in mind when reading programmes of this kind: the difference between volume and velocity. The ¥800 billion is a volume fact — it describes how much has been allocated. The velocity fact is how fast that money moves through procurement and into paid construction work, and that fact has not been printed yet. In the programmes I have tracked over the years, the ones that delivered were the ones where velocity followed volume within a quarter or two; the ones that disappointed were the ones where the allocations sat politely in the pipeline while the sites waited. That is the ledger I am watching, and Q3 is the first line in it.