Manufacturing growth: how to tell a recovery from a stronger market

Start with production, not the sales headline

UNIDO's second-quarter manufacturing report records global production growth of 1.1% from the preceding quarter. Medium-high- and high-technology industries expanded by 1.7%, while lower-technology manufacturing was broadly flat. Computers and electronics and other transport equipment were among the stronger categories.

The report also records a 4.1% rise in manufacturing exports. That figure is not directly comparable with production growth: the production measure tracks volume, while trade is recorded in current US dollars. Prices and exchange rates can affect the trade value without an equivalent change in physical output.

This distinction matters when evaluating customers. A rise in revenue could reflect more units, higher prices, a different product mix, or currency movements. Before concluding that a buyer needs more production capacity, establish which explanation fits its business.

An illustrative example makes the problem clear. Suppose a factory sells 100 identical units at 10 currency units each. Revenue is 1,000. If it sells the same quantity at 11, revenue reaches 1,100, a 10% increase, while unit output is unchanged. This is arithmetic, not a description of a surveyed factory. Real businesses require a more careful comparison because their product mix rarely stays fixed.

Growth rates need a starting point

A positive growth rate says activity increased relative to a chosen base. It does not establish that production has recovered its earlier level, that utilization is high, or that margins have improved.

Consider another hypothetical factory with an output index of 100. A 20% decline takes it to 80. A subsequent 10% increase takes it to 88, leaving it below the original level. Describing only the rebound would omit the remaining gap. Describing only the gap would omit the improvement. Both are relevant to a capacity decision.

This is why a short review should show the latest level, the recent direction, and a meaningful earlier comparison. Select the earlier period because it helps answer the business question, not because it produces the most dramatic percentage. State whether the comparison is monthly, quarterly, annual, or annualized.

The Federal Reserve's September release illustrates a modest increase alongside spare capacity. It reported US manufacturing output up 0.2% in August and manufacturing capacity utilization of 76.8%. Those national estimates do not identify the bottleneck in a particular plant, but they discourage treating every positive production reading as evidence of a capacity shortage.

Look beneath the regional average

Geographic labels can conceal different industrial structures. A supplier to a narrow subsector needs evidence about that subsector and its customers, not merely a positive regional headline.

Eurostat's September 16 release estimated that euro-area industrial production rose 0.3% in July from June. Capital goods increased 1.3%, while energy production fell. The aggregate therefore combined different directions of movement. It also covered industry excluding construction, a broader scope than manufacturing alone.

These figures are useful when read on their own terms. They should not be treated as a direct test of another organization's global quarterly manufacturing estimate. Differences in period, geography, and industrial coverage can produce different results without either series being wrong.

A manufacturer can translate this into a customer map. Group customers by what they produce and where their end demand originates. A component maker serving industrial equipment, passenger vehicles, and food-processing lines may face three different demand patterns even when all three customers operate in the same country.

The next step is to compare those patterns with actual enquiries, order releases, cancellations, and delivery requests. Public statistics should help formulate the questions. They cannot replace the company's evidence about its own order book.

A survey measures a different dimension of activity

The Institute for Supply Management's September US Manufacturing PMI stood at 49.1, below the level associated with expansion in the sector. Yet its production subindex was 51.0. A weak composite reading and a stronger production component can occur together because the composite includes several dimensions of business activity.

PMI readings are diffusion indexes built from survey responses. A reading of 49.1 is not a 0.9% fall in manufacturing output. Nor does a one-point increase mean factories made 1% more products. The index tells readers about the distribution and direction of reported changes under its methodology.

Use survey information to probe the composition of conditions. Are businesses producing against old orders? Are new orders improving? Are inventories, staffing, or supplier deliveries changing? These questions can reveal a less comfortable situation than a production headline alone suggests.

For internal reporting, maintain the same separation. Record production performance alongside incoming demand and the inventory it creates. A busy plant deserves a different interpretation when it is completing customer orders than when it is adding stock without an identified buyer.

Faster shipments may reflect timing

The WTO's October update identified early importing ahead of tariff changes and demand for AI-related goods as important features of trade in the first half of 2025. Its forecast at that point put merchandise-trade volume growth at 2.4% for 2025 and 0.5% for 2026. These were projections, not completed annual results.

The timing issue deserves attention. A customer that brings purchases forward can strengthen one period's shipments while reducing the need to order later. That possibility is a reason to inspect the delivery schedule and inventory position; it is not proof that every recent order is temporary.

Ask customers what changed. Is the order associated with a new product, additional capacity, replacement stock, or an earlier purchasing schedule? Where evidence is unavailable, preserve the uncertainty in the forecast instead of assigning a confident explanation.

Make the capacity decision at the right level

Before approving additional capacity, prepare a short record linking the market evidence to the proposed investment. Identify the relevant customers, the operations that constrain delivery, the expected product mix, and the assumptions about recurring orders.

Separate demand growth from an operational improvement opportunity. A machine could be justified because it removes a quality problem or replaces unreliable equipment even in a slow market. Conversely, a strong industry forecast may offer little justification for equipment that does not address the factory's actual constraints.

Test the proposal against a less favorable order profile. What happens if customer releases arrive later, if the mix changes, or if the initial increase does not persist? Use the business's own numbers, retain the calculations, and identify which assumption most affects the decision.

Early October's manufacturing evidence supports targeted investigation rather than a single recovery label. The most useful growth figure is the one that can be connected to a specific product, a consistent measurement basis, and demand the factory has a credible chance of serving.