원문 보기
TSMC has just reported a sales figure that makes you stop scrolling . September revenues increased by 54.6% compared to the previous year. If a company grows that fast, the obvious question is, how could its stock be expensive? Well, imagine buying a business after everyone has already decided that its next few years will be extraordinary.
The business could fulfill exactly what they expected. Your investment could still disappoint because you paid for that success before it arrived. That's the tension in this TSMC report. Growth deserves attention, but it doesn't tell us what price makes sense for the stock. To answer that, we need to know how much of those sales translates into lasting profits and how much future success is already factored into the purchase price.
Welcome to River Pond's Global Capital. Let's pause for a moment on that figure of 54.6%, because understanding what it actually measures changes the way you read the whole story. TSMC reported revenues in September of approximately 511.86 billion new Taiwanese dollars. That's 54.6% above September 2025, but 0.6% below August 2026.
In the first 9 months of this year, revenue increased 41.1% compared to the same period last year. These comparisons are not contradictory. A business can be much larger than it was a year ago while taking a small step backward compared to the previous month. The slight monthly decrease in September does not, in itself, establish a weakening of demand.
And the huge annual increase doesn't mean that growth accelerated by 54.6% in a single month. Nor is it a report that says earnings per share increased by 54.6% or that revenue from artificial intelligence increased by that amount. These are consolidated revenues. It tells us the size of sales that pass through the company before knowing the full picture of quarterly costs and profits.
If we focus on the headline, we miss the part about what the shareholders are actually paid for. Revenue pays suppliers, employees, and the cost of operating the business. What remains is what matters . How reliable that remnant is matters even more when you're buying stocks for years instead of trading a headline for an afternoon.
This is where TSMC's own numbers make the story interesting. Its gross margin for the second quarter was 67.7%. Its previously issued guidance for the third quarter points to a gross margin between 65 and 67%, along with revenue of US$44.6 to US$45.8 billion. That guide describes a business that expects higher quarterly sales, allowing for a lower gross margin than the previous quarter.
It's a forecast, not the result of the third quarter. The full earnings report still needs to tell us what really happened. Gross margin is the portion of revenue that remains after the direct cost of manufacturing the products, before other expenses and taxes. A lower percentage does not automatically mean fewer dollars of profit.
If sales increase enough, total gross profit can still grow, but it means you ca n't simply take the revenue growth rate and apply it to the profit forecast. There is a commercial reason to pay attention here. In its discussion of second-quarter earnings, management said that the rapid ramp-up of its 2-nanometer technology was expected to reduce gross margin by three to four percentage points in the second half of the year.
It was hoped that strong demand and efficiency improvements would offset some of that pressure. Implementing an advanced manufacturing process costs money before it reaches its full profit potential . That could be a wise investment. Customers gain access to better technology, and the manufacturer builds future capacity.
The question for shareholders is how quickly that investment becomes a reliable source of profit. Therefore, a decrease in the margin needs an explanation. Is the cost of preparing for profitable growth temporary? Does the process become more efficient as production increases? Or will a larger portion of each future sales dollar still go toward costs?
These possibilities have very different implications for the value of the same income holder. And even the reported profits deserve a closer look . TSMC's earnings per common share in the second quarter were 27.25 new Taiwan dollars. The administration said that gains related to its shares in Vanguard International Semiconductor contributed 2.24 new Taiwanese dollars to that figure.
Those earnings are actual accounting revenues . Even so, they must be distinguished from the profits obtained through chip manufacturing. If you're estimating what TSMC can earn repeatedly, rolling over an investment profit to each future quarter would make the business appear more profitable than your evidence supports.
That doesn't make the quarter weak. This means that the main earnings figure answers a different question than the one an investor needs to ask. What can this business reasonably earn again without requiring the same investment returns to be repeated? Now, let's put a price on the profits. Because this is where a great company can turn into a disappointing investment.
I'm going to use a fictitious company with simple numbers. These are not actual earnings, stock price, or a forecast for TSMC stock. Let's assume this company earns $10 per share per year and its shares are trading at $300. At that price, investors pay $30 for every dollar of annual profit. That 's a price-to-earnings ratio of 30 .
It's a way of describing the price you pay for the benefit the company produces. One year later, annual earnings rise to $12 per share. That's a 20% growth, a very healthy result. But suppose investors now pay only 24 times the profits. $12 multiplied by 24 gives a stock price of $288. The business earned 20% more.
The share price fell by 4%. We are leaving out dividends, fees, and taxes from this illustration to keep the relationship clear. The change came from the price investors were willing to pay for each dollar of profit. That price could fall if growth expectations cool , uncertainty increases, or if other investments become more attractive.
Earnings growth has to outpace that change before shareholders see a profit. If the same company maintained its multiple of 30, $12 of annual earnings would support a price of $360 in this simplified example. That would be a 20% price increase. Therefore, the lesson is not that strong growth produces poor returns.
The price paid for the profits matters as much as the profits themselves. Let's apply this again to TSMC. A 54.6% increase in revenue gives us information about the business. It does not tell us what returns investors will expect several years from now, nor what multiple they will be willing to pay when those returns arrive.
The phrase "it's growing fast, so it's cheap" silently assumes that future profit will continue to expand and that the valuation will hold. Both things can happen, but both require evidence beyond a single monthly sales report. You might hear that a stock seems affordable because its forward earnings multiple is lower than its current earnings multiple .
That's worth investigating. Just remember that the future figure uses estimated earnings. The ratio may appear lower because someone expects the denominator to grow very rapidly. If those earnings estimates turn out to be too optimistic, that apparent bargain changes without the share price moving at all; You have to reduce expected earnings and suddenly you're paying more for each dollar of profit than the original estimate suggested.
For TSMC, a useful estimate must connect demand with production capacity, costs, and sustainable margins. An impressive sales forecast is only one part of that job. The more future growth you need to justify your purchase price, the more sensitive the investment becomes to a delay or a less profitable expansion.
Cash gives us another way to verify the story. In the second quarter, TSMC generated NT$783.36 billion from operating activities and spent NT$496 billion on capital expenditures. Its reported free cash flow was 287.36 billion Taiwanese dollars using operating cash flow less capital expenditures. That's a substantial positive cash flow generation after a substantial investment.
It is also a reminder that profits and available cash are not interchangeable. A factory can help generate income for many years, while the equipment bill arrives much sooner. For an investor, the question is not whether the spending is automatically good or bad. It's about whether the capacity being built can generate enough future cash to justify the money committed today.
A strong demand helps that case. The final return on investment has yet to materialize. It must also keep the numbers in the same currency and on the same shareholding basis. The reported growth for September is measured in new Taiwan dollars. The quarterly revenue projection we just discussed is in US dollars.
Changes in exchange rates can affect the comparison even when the underlying business grows. And if you look at TSM in the United States, each US depositary share represents five common shares. You cannot divide its market price in US dollars by the quarterly earnings of a common stock in Taiwan dollars and call the result an annual valuation multiple.
The currency, the stock, the unit, and the earnings period must match. It sounds like accounting, but this is how a stock can appear drastically cheaper or more expensive due to an incorrect calculation. A careful valuation begins with a consistent earnings figure. Then you can discuss how much future growth is worth .
The company's next major event is already on the calendar. TSMC's third-quarter results conference is scheduled for Thursday, October 15 at 2:00 a.m. Eastern Time. Following this investigation's conclusion , that conference and the full quarterly results are still pending. When the report arrives, the monthly sales manager will have fulfilled part of their role.
We will have more information on the quarter's profitability and management's outlook will help test whether the future earnings story deserves more confidence. What would strengthen the case? Evidence that demand is turning into profitable production, that the cost of introducing new technology is behaving as expected, and that the outlook supports sustainable gains beyond an exceptional month.
Strong sales combined with that evidence would tell us much more than just strong sales. What would make me more cautious? A forecast that requires investors to assume that higher costs are temporary without a convincing explanation, or an earnings estimate that requires the recent pace of growth to continue almost uninterrupted.
The concern would be the assumptions needed to support the purchase price, even if demand remains healthy. A company can publish excellent results and still disappoint investors who were expecting even better results. It may also report a slower growth rate and support a higher valuation if earnings appear more sustainable than feared.
Without a verified comparison with expectations, calling a report a hit or a miss doesn't tell us much. That's why the upcoming earnings conference deserves attention beyond just the biggest number in the press release. Listen to what management can explain about the economics of expansion. More customers and more shipments are helpful.
A clearer path from those shipments to lasting profits would be even more helpful for valuation. If you already own TSMC shares, this is also a good time to ask what your position depends on. Would your investment still make sense with slower growth or a lower earnings multiple? What portion of your portfolio relies on the same demand for AI infrastructure, even if the companies have different names?
You don't need to guess the next price move to answer those questions. You need to understand what assumptions make the position work and whether you could tolerate a reasonable disappointment. A solid company can still represent a large exposure to a single set of expectations. TSMC's September report presents an impressive sales result.
This supports taking the company's growth seriously. To decide if the stock is cheap, we still need a credible view of future earnings and a price that leaves enough room for uncertainty. Before you buy based on that 54.6% headline, ask yourself what benefits can actually last and what you're paying for them.
Those questions remain useful, whether the next report spurs action or causes it to falter. Would a stronger revenue guidance or a more compelling margin outlook be more important to your TSMC outlook? Tell me in the comments. If you want financial stories explained through the numbers that matter to shareholders, subscribe to River Pond's Global Capital . Thanks for watching, and see you next time.
Comments 0