祝融说。

When analyzing any major "storm" in the Chinese economy, there is one word you can never avoid — infrastructure construction, commonly referred to as "infrastructure."

In my four decades of observation, infrastructure investment plays a role in the Chinese economy much like the "reserve force" in ancient warfare. During calm and favorable times, it may not be the main attacking force at the front. But whenever the economy encounters headwinds, when external demand weakens and internal momentum falters, this government-deployed "reserve force" is immediately rushed to the battlefield. Its mission is clear and critical: hold the line, stabilize the position, and buy precious time and space for the reorganization and counterattack of subsequent forces.

Therefore, I define infrastructure investment as the "ballast stone of policy." When the great ship of the economy encounters rough seas, decision-makers turn first and foremost to adding this heavy "ballast stone" of infrastructure to steady the vessel and prevent it from capsizing.

From the issuance of long-term construction bonds to cope with the 1997 Asian Financial Crisis, to the "Four Trillion" stimulus to address the 2008 global financial crisis, to the massive issuance of local government special bonds in recent years to counter the impact of the pandemic and economic downturn... virtually every time the Chinese economy faces a major test, we see infrastructure investment on the scene.

Analyzing infrastructure investment is essentially interpreting government behavior and macroeconomic policy intent. Unlike manufacturing investment, which is purely driven by markets and profits, and unlike real estate investment, which is closely tied to financial cycles and household expectations, its pulse resonates in lockstep with the policy rhythm of Zhongnanhai, the budget reports of the Ministry of Finance, and the project approvals of the National Development and Reform Commission (NDRC).

In this chapter, we will don the hat of a "policy researcher" and delve deeply into the composition, operational mechanisms, and analytical framework of this "ballast stone." We will learn how to define its scope, how to observe it comprehensively from the three dimensions of projects, funding, and policy, and most importantly, how to use fiscal data and special bonds — the "real money" clues — to anticipate its future direction.

Understanding infrastructure allows you, during economic headwinds, to see the signal of a "policy bottom" earlier than others and to more deeply grasp the resilience and challenges of the Chinese economy.

5.1 The Scope and Meaning of Infrastructure Investment

To analyze infrastructure, we must first understand exactly what "infrastructure" includes. This question is far more complex than it sounds. There is no official statistical indicator directly called "infrastructure investment." What we currently use are "pieced-together" scopes agreed upon by research institutions and the market.

The "Broad" and "Narrow" Infrastructure: Distinguishing Two Scopes

In the market, you typically encounter two scopes for infrastructure investment: the old scope (narrow) and the new scope (broad).

• Old Scope (Narrow): This is the most traditional scope. It directly selects three core categories from the fixed asset investment industry classifications published by the National Bureau of Statistics:

  1. Transportation, warehousing, and postal services
  2. Water conservancy, environment, and public facilities management
  3. Production and supply of electricity, heat, gas, and water

  The advantage of this scope is that it is clear, simple, and has continuous historical data. It essentially covers what we traditionally know as "railways, highways, and infrastructure" and public utilities. Before 2018, almost all research used this scope.

• New Scope (Broad): In 2018, the NBS adjusted its industry classifications, separating out some infrastructure-related activities that had been classified under other industries. To more comprehensively reflect the full picture of infrastructure, the market mainstream began shifting to the "new scope." It builds on the old scope by removing "production and supply of electricity, heat, gas, and water" but adding infrastructure-related components from other industries. A common broad-scope calculation formula is: Infrastructure investment (broad) = Total fixed asset investment completed - Real estate development investment - Manufacturing investment - Agriculture, forestry, animal husbandry, and fishery investment - Mining investment - ... (subtracting a series of non-infrastructure industries)

  The advantage of this scope is that it is more comprehensive and better reflects the full picture of "big infrastructure." The disadvantage is that the calculation is relatively complex, and different institutions may have slightly different deduction items, leading to minor differences in results.

• The Emergence of "New Infrastructure": The Evolution of Content In recent years, as China's economy has entered a phase of high-quality development, the content of infrastructure has also been evolving. The 2018 Central Economic Work Conference first proposed the concept of "new-type infrastructure" (new infrastructure). It mainly includes three aspects:

  1. Information infrastructure: 5G base stations, big data centers, industrial internet, artificial intelligence, etc.
  2. Integrated infrastructure: Using information technology to upgrade traditional infrastructure, such as smart transportation, smart energy, etc.
  3. Innovation infrastructure: Major scientific and technological infrastructure supporting scientific research and technological development, such as national laboratories.

  The introduction of "new infrastructure" marks the transformation and upgrade of infrastructure investment from "hard" connectivity of "railways, highways, and infrastructure" to "soft" connectivity of data and information. It focuses more on shoring up weaknesses and promoting innovation, serving the digital economy and national strategies. In our analysis, we need to recognize that "new infrastructure" currently accounts for a relatively small share of total infrastructure investment, and traditional infrastructure remains the "main force" for stabilizing growth. However, "new infrastructure" is growing very quickly, represents the future direction, and analyzing it helps us grasp the trajectory of economic structural transformation.

As an analyst, which scope should I use? My advice is:

• For long-term trends: Use the old scope (narrow), as it has a longer, more stable historical data series, making it easier for cross-cycle comparisons.
• For current analysis and forecasting: Use the new scope (broad) whenever possible, as it better reflects the comprehensive picture of current policy efforts.
• For storytelling and the future: In analytical reports, mention "new infrastructure" separately as a structural highlight and future growth point.

In my team, we calculate and track both scopes' growth rates and compare them. If they move in the same direction, the conclusion is robust. If they diverge, for example, if the new scope grows significantly faster than the old scope, it may indicate that some non-traditional infrastructure areas (such as warehousing and logistics, ecological and environmental protection) are gaining momentum.

Characteristics of Infrastructure Investment: Why Infrastructure?

Having understood the scope of infrastructure, we also need to understand its intrinsic properties. Why is it that whenever counter-cyclical adjustment is needed, decision-makers always think of infrastructure first?

• Strong government leadership, quick to take effect: Compared to manufacturing investment, which relies on the spontaneous decisions of market entities, the power to initiate infrastructure investment is largely in the hands of the government. Through the NDRC accelerating project approvals, and the Ministry of Finance and the central bank providing funding support, the government can quickly launch a number of major projects in a short period, forming real physical workload with rapid execution.
• Long industrial chain, broad driving effects: A high-speed railway project can drive upstream industries like steel, cement, and machinery, and create a large number of jobs downstream. Its multiplier effect is very significant, effectively amplifying the government's "initial push" into total social demand.
• Dual attributes of short-term demand and long-term supply: Building a road is an investment activity that pulls demand in the short term. Once completed, it reduces logistics costs, improves regional connectivity, and becomes a quality asset that enhances long-term economic efficiency (future supply). This attribute gives infrastructure investment the theoretical advantage of "killing two birds with one stone."
• Public goods attribute, areas of market failure: Many infrastructure projects have the characteristics of huge investment amounts, long construction periods, and low rates of return. Private capital is often unwilling or unable to enter. This requires the government to play the role of supplementing the "market failure."

Of course, this "prescription" of infrastructure is not without side effects. Excessive, inefficient infrastructure investment can lead to a rapid expansion of local government debt, a continuous decline in the return on investment, and a "crowding-out effect" on private investment. These are all aspects we need to view dialectically in our analysis.

5.2 Three Dimensions for Observing Infrastructure Investment: Projects, Funding, and Policy

To accurately predict the trajectory of infrastructure investment, it is far from enough to simply stare at the final published data on investment completed. That is just "driving by looking in the rearview mirror." A professional analyst must construct a forward-looking, multi-dimensional observation system. I summarize it into three dimensions: projects, funding, and policy. Together, these three determine the "ceiling" and "acceleration" of infrastructure investment.

Dimension One: Projects — Where Are the "Troops"? (Construction Potential)

Without projects, all investment is a water without a source. Observing the project pipeline is the foundation for judging future infrastructure potential.

• Leading indicators: Project approval and reserves

  • NDRC project approval: The National Development and Reform Commission is the "master gate" for major infrastructure projects. Monitoring the project approval data published monthly on its website (including the number of projects and total investment amount) is the first step in observing future project supply. If approvals accelerate and the total approved investment increases significantly, it indicates that the "project pool" is being filled.
  • Medium and long-term plans (e.g., 14th Five-Year Plan): Infrastructure projects do not emerge from thin air. Most originate from national and local medium and long-term development plans. Carefully studying the layout of transportation, water conservancy, energy, and other areas in these plans helps us grasp the general direction and key areas of infrastructure investment for the next few years.
  • Construction permit area: For the construction industry, the area of construction permits issued is a direct reflection of future construction volume.

• Coincident indicators: Project commencement and construction intensity

  • Total planned investment in new projects: We discussed this indicator in detail in the previous chapter. In the infrastructure field, it is equally crucial. Observing the growth rate of total planned investment in new projects in infrastructure-related industries (such as transportation, water conservancy and environment) is the core leading indicator for judging the turning point of infrastructure investment.
  • Construction PMI: This is the most direct monthly indicator reflecting the health of the construction industry. The construction business activity index reflects the current month's construction intensity, while the new orders index indicates future workload. If these two indices continue to rise, it indicates that infrastructure construction is heating up.
  • High-frequency physical quantity data:
    • Cement output/dispatch rate: Cement cannot be stored for long periods. Its output and dispatch situation very truthfully reflects the current level of busyness at construction sites. Tracking the dispatch rate of cement enterprises in key regions (weekly data) is a "grassroots" indicator for grasping construction progress.
    • Operating rate of petroleum asphalt equipment: Asphalt is an essential material for road construction. The seasonal and year-on-year changes in its operating rate serve as a good indicator of road construction progress.

Dimension Two: Funding — Where Does the "Provisions" Come From? (Construction Capacity)

With projects but no money, projects remain only on paper. Funding is the "blood" that determines whether projects can be smoothly started and continuously advanced. Analyzing the funding sources of infrastructure is the most core, most quantifiable, and most predictive part of the entire infrastructure analysis framework.

The funding sources for infrastructure can be summarized into "four accounts":

1. Budgetary funds: The portion of general public budgets used for infrastructure. This portion is relatively stable but limited in size, usually only playing the role of "seed money."
2. Domestic loans: Primarily loans issued by banks and other financial institutions to infrastructure projects, including supporting loans from policy banks (CDB, ADBC).
3. Self-raised funds and others: This is the largest and most complex portion of infrastructure funding. It includes:
   • Local government funds (the main player): Mainly the portion of local government special bonds used as project capital. This has been the most core incremental source of infrastructure investment in recent years.
   • Financing of Local Government Financing Vehicles (LGFVs, or chengtou companies): Funds raised by chengtou companies through issuing bonds, bank loans, non-standard financing, etc.
   • Social capital: Funds invested by enterprises under models such as PPP (Public-Private Partnership).
4. Utilization of foreign capital: A very small portion, negligible.

Among these "four accounts," we should focus on fiscal funds (especially special bonds) and supporting bank credit.

How to track funding?

• Ministry of Finance data (monthly):
  • The expenditure progress of items such as "urban and rural community affairs" and "transportation" in the general public budget expenditure.
  • Government-managed fund budget expenditure, especially the issuance scale, issuance pace, and investment direction structure of local government special bonds.
• Central Bank and Banking Regulatory Commission data (monthly):
  • The net financing amount of "government bonds" and "corporate bonds" in aggregate social financing.
  • The increment of medium and long-term loans from financial institutions, especially loan data directed to the infrastructure industry.
• Market high-frequency data (weekly/daily):
  • Issuance and bid yields of special bonds and chengtou bonds: Through financial terminals like Wind, we can track weekly bond issuance in real time. Accelerating issuance pace and falling yields are both signals of easing funding conditions.

Dimension Three: Policy — When Does the "East Wind" Arrive? (Construction Willingness)

Projects are the "troops," and funding is the "provisions." Policy is the "commander's order" that determines when and where to attack — it is that east wind.

• Top-level meetings set the tone:
  • Central Economic Work Conference (every December): Sets the overall tone for the next year's macro policy. Phrases in the communique such as "proactive fiscal policy should be stepped up and made more effective" and "appropriately advance infrastructure investment" are the most authoritative signals for judging policy direction.
  • Politburo meeting (every quarter): Analyzes the economic situation and deploys work for the next phase. If the meeting emphasizes the increasing pressure to "stabilize growth" and clearly mentions the need to leverage the key role of investment, then subsequent infrastructure stimulus measures are imminent.
• State Council Executive Meetings: This is the "execution layer" for policy implementation. State Council executive meetings frequently discuss topics such as "stabilizing investment," "issuance and use of special bonds," and "major project construction." Reading the meeting press releases allows you to understand specific policy details and the pace of implementation.
• Ministry documents and press conferences:
  • NDRC: Policy guidance on project approval and investment areas.
  • Ministry of Finance: Regulations on fiscal deficits, special bond quotas, and efficiency of fund use.
  • Central Bank / Banking Regulatory Commission: Statements on monetary policy, credit policy, and financing support for infrastructure projects.

The analyst's role: Our job is like that of an intelligence analyst, piecing together information fragments from these three dimensions into a complete strategic map.

• Policy sets the tone ("why"): From high-level meetings, we sense the urgency of "stabilizing growth" is rising.
• Funding is in place ("what to fight with"): From Ministry of Finance and central bank data, we see special bond quotas being issued ahead of schedule and banks increasing credit supply to infrastructure projects.
• Projects are ready ("where to fight"): From NDRC and local government developments, we see a batch of new major projects being approved and commenced at an accelerated pace.

When the signals from all three dimensions point in the same direction, we can predict the trend of infrastructure investment for the next few quarters with great confidence.

5.3 Data Linkages: Fiscal Expenditure, Special Bonds, and Infrastructure Growth

Theory must be verified by data. In infrastructure analysis, the hardest-core skill that best demonstrates an analyst's ability is to quantitatively link the "money" data with the "investment" data and establish a forecasting model. In this section, we will focus on the most critical transmission chain: fiscal/quasi-fiscal funds → infrastructure investment completed.

The Transmission of Fiscal Funds: From "Issuance/Expenditure" to "Investment Completion"

From the moment fiscal funds — especially special bonds — are recorded as "issued" or "expended," to when they actually become "physical workload" (tying rebar, pouring concrete) on the construction site, and subsequently recorded as "fixed asset investment completed," there is a transmission time lag. Understanding and quantifying this time lag is key to prediction.

• Breakdown of the transmission path:

  1. Funds arrive: The Ministry of Finance announces successful issuance of special bonds / fiscal funds are allocated to project entities. This is the starting point.
  2. Project commencement/construction: The project entity, having received the funds, begins bidding, procurement, and hiring workers, converting funds into construction activities.
  3. Statistical reporting: At the end of the month, the construction unit reports the "investment completed this month" to the statistical department based on the month's project progress. This is the end point.

• How long is the time lag? According to our team's years of measurement experience, this time lag is approximately 1 to 3 months. That is, the large-scale issuance of special bonds this month will primarily be reflected in the infrastructure investment completion data of the next 1 to 3 months. This time lag is not fixed.

  • During periods of urgent "stabilizing growth" when policy emphasizes "forming physical workload as soon as possible," the government will use various means to push projects to accelerate, and the time lag may shorten.
  • When project reserves are insufficient, preliminary work is not solid, or there are obstacles such as environmental protection or land acquisition, a phenomenon of "waiting for projects" may occur, and the time lag will lengthen.

• How to build a forecasting model? A relatively simple and effective model is to regress the current period's infrastructure investment growth rate against the growth rate of fiscal/quasi-fiscal funds from the past few months. Infrastructure investment YoY growth(t) = a * Fiscal expenditure-related item YoY growth(t-1) + b * Fiscal expenditure-related item YoY growth(t-2) + c

  The "fiscal expenditure-related item" here is a carefully constructed "infrastructure funding proxy variable." A good construction method is: Infrastructure funding proxy variable = Infrastructure-related portion of general public budget expenditure + Infrastructure-related portion of government-managed fund expenditure + Infrastructure-related portion of policy bank loans + ...

  By regressing historical data, we can obtain coefficients a and b, which reflect the weight and time lag structure of the fund transmission. This way, once we have the latest fiscal and financial data, we can plug them into the model to make quantitative predictions about future infrastructure growth.

The Core Driver: The Rise of Local Government Special Bonds

In recent infrastructure stories, if there is one "protagonist," it is undoubtedly local government special bonds. From their debut in 2015 to now totaling trillions of yuan in annual issuance, they have become the most core marginal force driving infrastructure investment. Not understanding special bonds means not understanding today's Chinese infrastructure.

• The "Special Mission" of Special Bonds

  • Earmarked for specific purposes: The funds from special bonds must be used for public welfare projects that generate certain returns. They cannot be used for recurrent expenditures (like paying salaries). The repayment of principal and interest must be covered by the project's own returns.
  • The policy innovation of "acting as capital": In 2019, policy began allowing some special bond funds to be used as capital for major projects (project capital is typically required to be about 20% of total investment). This is an extremely important "leverage" innovation! This means that 1 yuan of special bond capital can leverage about 4 yuan of bank loans or other social financing, creating a total investment of 5 yuan. This greatly amplifies the multiplier effect of fiscal funds.
  • Bypassing the "deficit" constraint: Special bonds are included in the government-managed fund budget, not the general public budget deficit. This makes them an excellent tool for implementing a proactive fiscal policy without expanding the "visible" deficit.

• How to analyze infrastructure through special bonds?

  1. Look at the total quota: At the annual "Two Sessions" in March, the Premier's "Government Work Report" announces the new special bond quota for the year. This number directly determines the upper limit of the year's infrastructure "ammunition depot."
  2. Look at the issuance pace: The Ministry of Finance regulates the issuance pace of special bonds according to the needs of "stabilizing growth." If the requirement is "basically all issued in the first half of the year," it means policy is front-loaded, and infrastructure growth in the first half of the year will be strongly supported.
  3. Look at the investment structure: The Ministry of Finance publishes the investment areas of special bond funds. We can see what proportion of funds is going to transportation, municipal works, industrial parks, new infrastructure, etc., thereby judging the structural focus of the policy.
  4. Look at the proportion "used as capital": This is a key "leverage" indicator. If policy allows more areas of special bonds to be used as capital, or increases the proportion used as capital, the same amount of special bond issuance can leverage a larger total investment.

Case Study: The "V-Shaped Reversal" of Infrastructure in 2022

At the beginning of 2022, the Chinese economy faced the triple pressure of demand contraction, supply shock, and weakening expectations. Let us see how the decision-makers used the "ballast stone" of infrastructure:

1. Policy tone (end of 2021 - early 2022): The Central Economic Work Conference clearly stated "appropriately advance infrastructure investment."
2. Funds front-loaded (Q1): The new special bond quota for 2022 was 3.65 trillion yuan, of which 1.46 trillion had been issued ahead of schedule by the end of 2021. This ensured that projects had funds to start in the new year. In Q1, special bond issuance was exceptionally rapid.
3. Transmission effect materializes (from Q2 onwards): Although the economy was severely impacted by the pandemic in Q2, infrastructure investment, supported by massive funding, showed strong resilience. From April onwards, the cumulative year-on-year growth rate of infrastructure investment began to stabilize and recover, taking the lead among the three types of investment in "turning positive" and continuing to rise.
4. Incremental tools added (mid-year): To further offset the impact of the pandemic, hundreds of billions of yuan in policy-based development finance instruments (commonly known as "funds") were established through policy banks mid-year. These injected equity capital into major projects, effectively serving as another round of "quasi-fiscal" ammunition.
5. Result: For the full year, infrastructure investment (NBS definition, excluding electricity) grew by 9.4% year-on-year (market-estimated broad measure including electricity: about 11.5%), far exceeding manufacturing (9.1%) and real estate (-10%), playing a decisive "supporting" role in "stabilizing growth" for the whole year.

This case perfectly illustrates the complete analytical chain of "policy → funding → projects → investment."

Chapter Summary

In this chapter, we went deep inside the "policy toolkit" of infrastructure investment. We clarified its "broad" and "narrow" scopes and understood why it always serves as the "ballast stone" for counter-cyclical adjustment.

We constructed a forward-looking framework for observation from three dimensions: projects, funding, and policy. We know that to understand infrastructure, you must not only see how busy the construction sites are but also read the NDRC's approval documents, study the Ministry of Finance's budget reports, and track the destination of each special bond.

Most crucially, we mastered the "hardest-core" quantitative logic in infrastructure analysis — the transmission chain of "fiscal/special bonds → infrastructure investment." We understood that the growth rate curve of infrastructure investment is largely determined by the pace of fiscal fund allocation.

At this point, we have dissected the first of the investment "three carriages." It represents the will of the government. Next, we will turn to a more controversial and more emotionally charged player — real estate. It was once the "engine" of rapid economic growth, but now it has also become the source of many risks. It represents the financial cycle. Understanding it will allow us to touch the deepest pulse of China's economic cycle fluctuations.