TSMC earnings: why profit grew as the 2nm ramp reduced gross margin

1. A ramp cost inside a growing profit

In Taiwan Semiconductor Manufacturing Company’s (TSMC) July–September 2026 results, released October 15, gross profit and operating profit rose sequentially while gross margin fell. The company attributes the margin decline primarily to the initial N2, or 2nm, ramp. [2]

Markets can read this combination in two directions. Focusing on the margin highlights the burden of introducing new technology; focusing on the increase in profit highlights revenue growth that more than absorbs the aggregate burden. Choosing one number erases half of the change. An assessment needs to establish where profit was added and where it was lost, rather than attach a single positive or negative label.

The implications also differ for chip designers, customers that outsource manufacturing and equipment vendors. More profit at the manufacturer can support its financial capacity to undertake further work. If ramp costs persist, however, future orders may not convert into profit on the same terms as current business. The useful question connects customer growth with the economics of supplying it.

SG Group’s approach puts the profit added by expansion and the lower margin during a manufacturing transition into one calculation, then follows the effects of prior-quarter items. Revenue, gross profit, operating profit and net income thereby become steps toward assessing repeatability, rather than textbook categories. What a share price already discounts is a subsequent question.

2. Build a profit bridge in the same currency

Start from the table and trace the increments before compressing the result into one growth rate. Retaining the intervening deductions shows improvements and offsets together. Differences between stages help locate the causes of the same profit change rather than grade the company on unrelated scales.

Operating profit does not follow mechanically from revenue. Equally large increases in production and operating costs could leave revenue larger without adding profit. Here the bridge begins with the increase in gross profit, deducts changes in operating expenses and other operating items, then incorporates non-operating results, tax and attribution. Each step locates an improvement or an offset.

The calculation does not mix dollar revenue growth with Taiwan-dollar profit growth. Changing the measurement currency changes the view of the business. A Taiwan-dollar bridge connects Taiwan-dollar costs and profits. An overseas investor’s home-currency return passes through security prices, distributions and exchange rates as well. Keeping business comparison separate from investment return makes the causes easier to trace.

Units matter too. Confusing billions with trillions changes the scale; confusing a margin change with a profit change changes the meaning. The calculations below use the statement’s NT$ million figures and round only the displayed results into billions. This preserves the link between reported amounts and changes better than reconstructing amounts from rounded percentages. The identity of each step matters more than a cosmetic match in the last decimal place.

3. Separate the revenue effect from the margin effect

Gross profit equals revenue times gross margin. Holding the previous margin constant while applying current revenue isolates a revenue-base effect. The gap from that intermediate amount to actual gross profit captures the margin effect. In this ordering, approximately NT$151.6 billion is added by revenue and NT$19.5 billion is subtracted by the margin change, leaving about NT$132.1 billion of additional gross profit.

The NT$19.5 billion is not an estimate of N2 costs. The decomposition measures the company-wide margin change, not the cost of a particular manufacturing generation. Changes in mix, prices, currencies and costs on existing products may all enter that gap. The company’s explanation identifies the principal cause but does not allocate the entire difference to one process. The calculation measures scale at the aggregate level; it does not invent a cost disclosure.

Reversing the order and valuing revenue growth at the current margin gives a smaller revenue contribution and a smaller margin drag. Total gross-profit growth is unchanged; the interaction term moves between the two components. Comparing decompositions therefore requires knowing which margin was held constant. The previous-margin convention used here is a transparent calculation order, not a claim to a uniquely correct economic attribution.

The scale indicates that expansion absorbed the aggregate drag. It does not promise the same revenue increment next quarter. Even a modest margin decline can matter when revenue stops growing; a margin that does not recover immediately can still accompany rising profit if the revenue base expands. Assessing this quarter therefore calls for a combination of scale and margin, rather than margin recovery as the only definition of success.

4. Not all additional gross profit reaches operating profit

The increase in operating profit is approximately NT$22.2 billion smaller than the increase in gross profit. That difference combines the change in operating expenses with other operating items. Some of the gross-profit increment was absorbed after production costs. Treating the entire gross-profit increment as additional bottom-line profit would skip this step.

TSMC links lower other operating income mainly to prior-quarter recognition of 2024 earthquake-insurance items. This describes a comparison effect, not a new earthquake. The earlier recognition is part of the comparison base; its nonrecurrence changes the apparent growth. [2]

Nor should the entire change in other operating income be relabelled as the insurance amount. A line containing several items is not identical to its principal driver. A comparable adjustment would remove the relevant recorded item on a consistent basis. Removing the entire line without sufficient detail could also erase ordinary business items. Before presenting an adjusted profit, the analyst must be able to identify what the adjustment removes.

Rising operating profit helps locate an improvement that may look less pronounced at the bottom line. But the label “operating” does not make every item repeatable. The insurance comparison shows that period-specific items can affect this stage too. The advantage of the bridge is not unconditional trust in one profit measure; it is the ability to narrow down the source of a change.

5. Prior-quarter share gains slow bottom-line growth

The company attributes the fall in non-operating results primarily to the absence of prior-quarter disposal and mark-to-market gains on Vanguard International Semiconductor, or VIS, shares. Investment-related gains had raised the comparison base. [2]

The calculation adds approximately NT$109.8 billion of operating-profit growth and subtracts NT$61.2 billion of lower non-operating results, leaving about NT$48.6 billion of pretax growth. Changes in tax and non-controlling attribution then take the parent-attributable increase to approximately NT$56.5 billion. Slower bottom-line growth than operating-profit growth therefore need not be explained solely by deterioration in manufacturing.

The NT$61.2 billion is not presented as the exact amount of VIS-related gains. An aggregate bridge and a specific investment breakdown serve different purposes. A rigorous growth rate excluding those gains would also need consistent tax treatment. Subtracting a pretax gain directly from net income mixes bases and can overstate the resulting improvement.

For investors, an explanation of subdued growth is not itself a reason for future growth. A high prior-quarter base explains the comparison but produces no additional sales next quarter. An easier comparison likewise does not improve competitive strength. After period-specific items are understood, the remaining question is how much incremental profit a larger new-generation business can retain.

6. A revenue share is not a process margin

N2 accounts for 8% of wafer revenue, not consolidated revenue or profit. Multiplying consolidated sales by that share and then by the group margin would cross two unsupported boundaries. A wafer is the circular substrate on which chips are manufactured. [2]

A process-level margin needs both process revenue and corresponding costs. Initial products and established production may carry different cost structures, so equal revenue shares need not yield equal profit contributions. The value of wafers sold also differs from the number of wafers started. Inferring yield or utilisation from a revenue mix requires the quantities and prices linking those measures.

A higher N2 share therefore cannot be assumed to raise the group margin. Even a higher selling price can coexist with a lower margin if introductory costs are larger still. Later, more stable production and lower unit costs could let the same shift support margins. The meaning of mix depends on the stage of the transition.

Reading financial statements helps frame the question, but general accounting knowledge cannot supply N2’s missing breakdown. The relevant additional information connects sales expansion and costs for the same process generation. Group figures establish the aggregate outcome: gross profit increased despite transition costs. Ranking individual processes requires a finer unit of analysis than the group average.

7. Where would the margin decline erase gross-profit growth?

A break-even calculation makes the relationship testable. Holding this quarter’s revenue fixed, a gross margin of approximately 57.6% would match the previous quarter’s gross profit. This is neither company guidance nor a production-cost forecast. It measures the arithmetic room above the previous profit amount using actual revenue as the denominator.

The gap between that threshold and the actual margin shows why the observed margin decline did not erase gross-profit growth. But lower revenue would raise the margin needed to preserve the same amount. Carrying 57.6% into a different quarter as a permanent safety line would be misleading. The threshold travels with its revenue and target-profit assumptions.

Looking from the other side, dividing prior-quarter gross profit by the current margin gives a revenue threshold of approximately NT$1,295.4 billion. This holds the current margin constant. In practice, lower sales could change factory cost absorption, so it does not promise that the margin would survive a decline to that level. Both thresholds test the assumptions supporting profit growth rather than forecast deterioration.

Business planners should likewise avoid judging success solely against a revenue target. At equal sales, heavier introductory costs change the profit retained. More stable production could, conversely, preserve profit even with slower revenue growth. A joint view asks how much margin improvement a given revenue path would require, making the quality of growth more concrete than a single pass-or-fail line.

8. SG Group View: follow three clocks through the transition

Three clocks help interpret this transition: sales expansion, the settling of production costs and the turnover of the comparison base. Sales measure work converted into revenue. Costs determine the economics of new production. The comparison base reflects how prior insurance and investment gains affect reported growth. The clocks run together without necessarily moving at the same speed.

When sales move first, profit may grow before the margin improves: growth is real, but additional work remains costly to undertake. If costs settle first, the margin might improve despite weaker revenue growth. Both could be described as good results, yet their next requirements differ. The first needs steadier economics; the second needs broader sales.

If only the comparison clock advances, bottom-line growth can change even with similar business activity. Once a large earlier gain leaves the comparison, a subsequent growth rate may look better. That alone would not prove a successful manufacturing transition. Comparing changes at the operating and net-income stages helps distinguish additional business from a changing base.

9. The global transmission runs beyond revenue headlines

More profit at a manufacturer gives customers information about its financial ability to provide future production. It does not turn the profit increment automatically into spending on the next factory. Investment plans, the allocation of funds and production readiness intervene. Financial expansion in this quarter and availability for a particular customer at a particular date remain connected through business decisions.

During an unsettled ramp, the quality of customer commitments may matter more. Large volumes alone may not support unchanged costs if specifications or delivery schedules shift materially. Stable plans could reduce duplicated preparation and abrupt changes. This is a possible transmission between ramp economics and customer planning, not a statement about the terms of TSMC’s current contracts.

Further downstream, effects could reach product costs and launch plans. A lower margin at the manufacturer does not establish a lower procurement price for its customers. Internal cost absorption and changed selling prices would transmit differently to customer margins. The group margin cannot justify adding a fixed number of points to another company’s profitability; intervening prices and quantities matter.

The guide to supply chains and cash flow explains the sequence of goods and payments between businesses. The additional ramp question is who bears the cost, and when. Procurement focuses on delivery, finance on timing and product teams on specifications. They can read the same revenue news while needing different follow-up evidence. Tracing the transaction is more useful than immediately projecting one company’s results onto the entire industry.

10. Connect equipment activity with foundry economics

The article on ASML’s equipment supply examines upgrades and production workloads. This TSMC analysis examines how a manufacturing business containing such equipment translates into earnings. Available tool time, good output, revenue and retained profit are connected through conditions. Placing equipment-vendor and foundry figures side by side does not guarantee that their earnings will move together.

Even when equipment spending grows, some spending may prepare for future business while other assets support current revenue. Strong expectations at ordering need actual operations to become profitable production. The margin-and-amount decomposition here does not feed expectations back into a vendor’s orders. It measures the profit already retained across the manufacturer’s business.

For suppliers, prolonged ramp costs could create demand for problem-solving equipment and services, or encourage more cautious investment. A burden can become another company’s opportunity if spending solves it; it can become a headwind if the manufacturer delays expenditure to protect returns. The outcome depends on the problem and investment decision, not simply on the direction of a gross-margin change.

The guide to AI demand and semiconductors asks which products and processes receive demand. This profit growth is not a common recovery indicator for every chip company. Firms serving similar end markets still differ in product generations, costs, contracts and assets. A hypothesis about another company needs shared activities and costs, rather than merely a shared demand label.

11. Read next-quarter guidance as two conditions

Fourth-quarter company guidance is US$52.3–54.3 billion in revenue and a 65–67% gross margin, with an exchange-rate assumption. These are forecasts. Selecting only the upper endpoints would erase the ranges rather than establish a profit outcome. [3]

Turning the historical bridge into a forecast would require consistent currencies. Dollar revenue cannot be added directly to prior Taiwan-dollar gross profit, and conversion at an assumed rate does not establish the realised average. Rather than manufacture a point profit estimate, the guidance can be read as two tests: whether revenue reaches its range and whether the corresponding cost structure supports the margin range.

Revenue reaching its range with margin near the lower end could indicate additional work alongside persistent burdens. At equal revenue, a margin nearer the upper end retains more profit. Currencies and mix can also move the rate, however. Assessing whether N2 ramp economics have settled needs both the reported outcome and a consistent explanation of the change.

The guide to results and guidance explains why prices respond to the gap against expectations as well as the level of a forecast. This guidance alone does not determine the next share-price move. If strong sales are already discounted, the persistence of costs could become the larger dispute. Keeping operating conditions separate from expectations helps avoid equating strong business performance with a strong investment return.

12. Three combinations for the next results

The first combination is broader revenue with easing ramp costs. Scale and margin would then support gross profit together, potentially improving the retention of profit. That assessment requires actual outcomes: mix alone cannot establish better costs. The stated drivers and the change in gross profit should move consistently.

The second combines rising revenue with persistent costs. Profit amounts and margins could again diverge. Calling that failure would ignore profit added by expansion; calling it success solely because the amount rose would overlook the burden of sustaining growth. Comparing gross profit retained from additional revenue across periods gives the assessment more substance.

The third combines slower revenue growth with persistent costs. A weaker scale effect could allow margin pressure previously absorbed by growth to reach the profit amount more forcefully. The issue becomes whether costs adjust, and the remedy differs depending on changes in prices, mix or production volumes. Waiting for stronger demand alone may leave a cost problem unresolved.

These are reading conditions, not probability-weighted forecasts. Revenue and gross margin could both improve while non-operating items weaken the bottom line. Begin with manufacturing economics, proceed through operating costs, then examine investment-related results and tax. Locating the driver makes it possible to explain what needs reassessment even if net income disappoints.

13. Test the argument that a lower margin means deterioration

Concern about a lower margin has a rationale: retaining less gross profit per unit of revenue can reduce room for weaker demand. Strong revenue expansion this quarter does not disprove that concern. It identifies how much scale helped absorb the burden. A margin-focused critic can make the argument concrete by examining costs if revenue growth becomes harder to sustain.

It would still be too strong to ignore actual profit growth and conclude that costs damaged the business. Profit earned while carrying a burden can fall below a hypothetical no-burden outcome yet exceed actual prior-quarter profit. Those comparisons answer different questions. Switching between them without explanation shifts the standard of the argument.

The bullish claim that a larger profit proves a smooth ramp also goes too far. Established business can generate enough profit to lift the group amount even if the new activity has weaker economics. Group results alone do not establish process-level improvement. The stronger bullish evidence would be further incremental profit accompanied by an explanation connecting the change to settling ramp costs.

14. Different claims need different additional evidence

Measuring business growth without prior-period items needs insurance and share-related results adjusted at consistent profit stages. Combining operating and after-tax adjustments may produce a neat-looking number that mixes unlike amounts. Keeping both reported and adjusted figures makes the chosen growth concept traceable. An analyst’s adjustment should not be presented as a company-reported measure.

A claim about passing costs to customers requires contract units, repricing dates and product terms, not just manufacturing costs. A falling supplier margin does not prove successful customer price negotiations; higher internal costs offer another explanation. A high margin likewise does not prove that higher prices were imposed. The aggregate change points toward a question about terms without answering it.

Quarterly accounts may also be too coarse to establish production stability. Conditions can differ between the beginning and end of a period while the accounts aggregate both. Monthly revenue would not become monthly profit without corresponding cost information. Combining frequencies requires care not to substitute a short-term activity change for period-wide profitability.

15. Assess the business and the price paid separately

An investment assessment first identifies the source of profit growth, then considers the value assigned to future earnings. Understanding operating growth does not establish a cheap share price. If prices already assume substantial growth, the same results may be insufficient; if cost deterioration had been heavily feared, the profit retained despite the burden could support valuation.

Expectations and order flow, discussed in the guide to share-price formation, matter here. Prices depend on views about future earnings and on the buying and selling of participants holding those views. The profit bridge measures company results, not positions or trading intentions. Understanding the figures does not create a short-term trading timetable.

Comparing earnings yields with interest rates likewise depends on the assumed repeatability of earnings. Capitalising a temporary gain as permanent can make a price look more generous than it is; treating transition costs as permanent can understate possible improvement. These results help test which profit is being valued, and for how long, before attempting a single valuation answer.

16. Record the conditions before the next growth rate

A useful note for the next results records expectations at the gross-profit stage, not just a net-income target. State the expected revenue scale and the margin needed to preserve profit. When results arrive, the analyst can identify whether revenue or profit retention differed. This leaves a reusable explanation rather than merely a score for whether a forecast was right.

Then record operating expenses and other operating items, followed by non-operating results and tax. State whether prior-period items were assumed to recur. With material comparison effects, separate the business forecast from the base-effect forecast. Even an accurate net-income total can conceal a different path, with different implications for the following quarter.

The question is how a growing manufacturer absorbs the cost of moving to a new generation. Expansion supported the result while prior-period items shaped the comparison. The next test is whether additional work keeps generating additional profit as the transition advances, and where that profit is generated. Those questions provide a consistent standard for both welcoming strong amounts and scrutinising weaker margins.