Why Korea's export boom isn't reaching consumers — chip jobs are 1.9% of employment, 10.6% in Taiwan
A Bank of Korea research note made public on November 15, 2026 says South Korea's consumer recovery may lag Taiwan's despite the same AI chip boom. It gives three reasons. IT manufacturing employs 390,000 people in Korea, 1.9% of all workers, against 910,000 or 10.6% in Taiwan. The income gains go mostly to high-paid chip workers, who spend a smaller share of extra income. And retail investors' daily stock trading fell from 65.9 trillion won in the second quarter to 37.2 trillion won in the third. Korean exports for November 1 to 10 were $33.5 billion, up 111.9% on the year, yet private consumption fell 0.2% in the third quarter. Taiwan's private consumption grew 5.4% in the first three quarters of 2026, Korea's 2.1%. The note adds that Taiwan itself waited about a year for spending to pick up
The three lines
- The three reasons — IT manufacturing is 1.9% of Korean jobs (10.6% in Taiwan), gains flow to high earners, and retail stock trading cooled after the market correction
- The gap — exports up 111.9% and chips up 331.2% in the first ten days of November; third-quarter GDP up 1.0% but private consumption down 0.2%
- Taiwan waited too — 8.8% growth in 2025 came with only 1.4% consumption growth; the 5.4% arrived in 2026, helped by a cash handout and a stock rally
Key questions
- why are korea exports booming but consumption weak
- **The Bank of Korea note gives three reasons: the chip sector employs a small share of workers, the extra income goes to high earners who spend less of it, and retail stock trading has fallen since the market correction.** | Channel | South Korea | Taiwan | |---|---|---| | IT manufacturing workers (2025) | 390,000, or 1.9% of all workers | 910,000, or 10.6% | | How exports grew | Led by higher memory chip prices | Led by higher volumes at foundries making AI accelerators | | Where the income goes | Concentrated in a small group of high-paid chip workers | Spread across small firms in the server and component supply chain | | Stock market and spending | Retail daily trading down from 65.9 trillion won to 37.2 trillion won | Rising prices and heavier trading supported consumption | | Private consumption, Q1 to Q3 2026 | Up 2.1% | Up 5.4% |
- what is a k-shaped economy
- **It is an economy where one part climbs while another stalls or falls, like the two strokes of the letter K. The Bank of Korea note uses it for the gap between IT and non-IT industries, and says it applies to both Korea and Taiwan.** | | Upper stroke | Lower stroke | |---|---|---| | Industries | Semiconductors, computers and other IT | Other manufacturing and domestic-demand sectors | | Korea, Q3 2026 (quarter on quarter) | Exports +2.3%, facilities investment +4.1% | Private consumption -0.2%; retail, lodging and restaurants -0.4% | | Wages | Gains concentrated in high-paid chip workers | Widening IT versus non-IT pay gap (Bank of Korea) | | What the note recommends | Spread AI adoption to lift productivity broadly | Improve middle-class spending capacity, ease living costs for low-income households |
- bank of korea taiwan comparison report
- **The note is titled, in translation, Features of Taiwan's Semiconductor Boom and Implications for Our Economy. It was written by two economists in the bank's Research Department and reported on November 15, 2026. Its conclusion is that Korean consumption should improve with a lag, but more slowly than Taiwan's did.** | Item | What the note says (as reported by Korean media) | |---|---| | Taiwan GDP growth | 1.1% in 2023, 5.3% in 2024, 8.8% in 2025, double digits expected in 2026 | | Taiwan private consumption | 3.2% in 2024, 1.4% in 2025, 5.4% in Q1 to Q3 2026 | | Timing of the boom | Taiwan from the second half of 2024, Korea from the second half of 2025 | | Why Korea may be slower | Small IT manufacturing workforce, gains concentrated in high earners, weaker stock trading | | Shared constraint | A K-shaped structure limits how far growth spreads to domestic demand | | Warning | If AI demand slows, shocks to the real economy and financial markets could amplify each other |
South Korea's exports have doubled from a year ago, and its consumers are spending less. A Bank of Korea research note made public on November 15, 2026 explains the gap by comparing Korea with Taiwan. The note, written by two economists in the central bank's Research Department, Lee Jun-ho and Jeon Min-geun, is titled (in our translation) Features of Taiwan's Semiconductor Boom and Implications for Our Economy. Its conclusion fits in two sentences. Judging by Taiwan's experience, Korean consumption should improve with a lag. But it may improve more slowly than Taiwan's did.
The two economies are riding the same wave. Global spending on artificial intelligence has driven up demand for chips, and growth accelerated in Taiwan from the second half of 2024 and in Korea about a year later, from the second half of 2025. Yet in the first three quarters of 2026, private consumption grew 5.4% in Taiwan and 2.1% in Korea. The note's explanation is not about the size of the export boom. It is about how wide the channel is between exports and households. This piece follows that channel, using the note as reported by Korean media and the official statistics around it, to see where it narrows.
1. The gap in numbers — exports doubled, spending slipped
Start with how wide the gap is. The export figures come from the Korea Customs Service and the Bank of Korea's trade indices. The spending figures come from the central bank's advance estimate of third-quarter gross domestic product (GDP).
| Indicator | Figure | Source and date |
|---|---|---|
| Exports, November 1 to 10 | $33.5 billion, up 111.9% year on year | Korea Customs Service preliminary data, November 11 |
| Daily average exports, same period | $4.46 billion, up 97.7% (7.5 working days, half a day more than a year earlier) | Customs data, via Newsway |
| Semiconductor exports, same period | About $16.6 billion, up 331.2%; 49.7% of all exports | Customs data |
| Cumulative exports, January 1 to November 10 | $963.7 billion; $1 trillion for the year expected around November 20 | Customs data, via Newsway |
| Export volume index, October | Up 43.9% year on year (September: 32.2%) | Bank of Korea, November 13 |
| Export value index, October | Up 107.6% | Bank of Korea, November 13 |
| Real GDP, third quarter | Up 1.0% quarter on quarter (the bank's August forecast was 0.3%); up 3.4% year on year | Bank of Korea advance estimate, October 27 |
| Private consumption, third quarter | Down 0.2% quarter on quarter, the first fall since the fourth quarter of 2022 | Bank of Korea advance estimate, October 27 |
Korea publishes customs-cleared exports every ten days, and those short windows are sensitive to how many working days they contain. Stripping that out, the daily average still rose 97.7%, so saying exports doubled is not an exaggeration. One product, semiconductors, accounts for half of everything the country shipped.
Spending went the other way. When the Bank of Korea released the third-quarter estimate on October 27, it said private consumption fell in both goods, such as cars and home appliances, and services, such as finance and restaurants. Lee Dong-won, who heads one of the bank's economic statistics departments, pointed to two temporary factors: lower stock trading volume after a market correction, and strikes at carmakers that delayed vehicle deliveries. Without those, he said, consumption would have risen 0.4%. Even on that basis the figure is 0.4%, in a quarter when facilities investment rose 4.1% and exports 2.3%.
The split shows up by industry too. According to EBN's account of the advance estimate, manufacturing output rose 2.0% in the quarter while services rose 0.2%. Wholesale, retail, lodging and restaurants fell 0.4%, and finance and insurance fell 4.3%. The industries that produced the 1.0% growth figure are not the ones households deal with day to day.
The income figure is wider still. The Seoul Economic Daily reported that real gross domestic income, which adjusts output for changes in the terms of trade, rose 18.8% from a year earlier, the fastest since the fourth quarter of 1977. Chip prices rose and import prices such as crude oil fell. The country as a whole earned far more. The question is whose hands that money reached first.
2. The three reasons — where the channel narrows
The note lists three reasons why Korea's consumer recovery may be slower: a small IT manufacturing workforce, gains concentrated among high earners, and a drop in stock trading after the market correction. Underneath them is a difference in how each country's exports grew.
| Channel | South Korea | Taiwan | What it means for spending |
|---|---|---|---|
| Employment | 390,000 IT manufacturing workers, 1.9% of all workers (2025) | 910,000, or 10.6% | The share of workers who receive the boom directly is about one-fifth of Taiwan's |
| How exports grew | Led by higher memory chip prices | Led by volume, centred on foundries making AI accelerators | Price-driven export growth does not raise output and hiring to the same degree |
| Breadth of the supply chain | Centred on a few large companies | Small and mid-sized firms across servers, components and other AI hardware | In Taiwan, higher output spreads to jobs and income more widely |
| Where income goes | Concentrated in high-paid chip workers; widening IT versus non-IT pay gap | The note says the K-shaped pattern holds in Taiwan too | High earners spend a smaller share of each extra unit of income |
| Stock market | Retail daily trading value down about 44% in the third quarter | Rising prices and active trading supported consumption | Weaker wealth effect and less spending on financial services |
First, employment. In 2025, Korea had 390,000 people working in IT manufacturing, 1.9% of all workers. Taiwan had 910,000, or 10.6%. Working backwards from those ratios, Korea's total is about 20.5 million workers and Taiwan's about 8.6 million. An economy with fewer than half as many workers has more than twice as many people in IT manufacturing. In Taiwan, roughly one worker in ten is in the sector. In Korea it is closer to one in fifty. When chipmakers' profits turn into wages and bonuses, the number of households on the receiving end differs accordingly. Edaily reported the note's observation that Taiwan's small and mid-sized firms are spread across the whole AI hardware supply chain, including servers and components, so that higher output reaches jobs and incomes across a wider range.
Underneath that is how exports grew. According to JoongAng Ilbo and Kukmin Ilbo, the note says Taiwan's export growth came mainly from volume, led by foundries, the contract manufacturers that make advanced chips for AI accelerators. Korea's came mainly from the rising price of memory chips, its main export. More volume means running more production lines and hiring more people. Higher prices go first to company revenue and profit.
| Volume | Price or value | Source | |
|---|---|---|---|
| Korea chip exports, September 2026 | +30.5% | Dollar unit price +182.0% | Based on Bank of Korea trade indices, via JoongAng Ilbo |
| Taiwan chip exports, Q1 2026 | About 80% of the increase came from volume | The remainder from price | The note, via JoongAng Ilbo |
| Korea total exports, October 2026 | Volume index +43.9% | Value index +107.6% | Bank of Korea, via Newspim |
| Korea computers, electronics and optical equipment, October 2026 | Volume +51.2% | Value +214.8% | Bank of Korea, via Newspim |
Multiplying the September figures gives volume of 1.305 times and unit price of 2.82 times, or roughly 3.7 times the export value. Most of the increase came from price. That is arithmetic, not a reported figure. For all exports in October, value doubled while volume rose 43.9%. That volume gain is itself the largest since the index began, but there is still a wide distance between the headline export number and the additional quantity factories actually produced.
Second, where the money goes. ChosunBiz quoted the note as saying the benefits are concentrated in a small group of high-income semiconductor workers and that the pay gap between IT and non-IT has widened. Higher-income households spend a smaller share of any additional income. The same sum turns into less consumption when it goes to a hundred thousand households than when it is divided among a million.
This sits at a different temperature from what the central bank said a few weeks earlier. EBN and Digital Times noted that in an October 22 report the Bank of Korea estimated special bonuses at major companies, chiefly the chipmakers, at 60 to 70 trillion won (roughly $45 to $52 billion at the October average exchange rate of about 1,338 won per dollar), and projected a sharp rise in household spending from early 2027. That report also expected large chip-company dividends and higher corporate tax revenue to flow through to households. The new note says that even when those bonuses are paid, only part will be spent, limiting the lift to overall consumption. The two reports do not contradict each other. Whether money is paid out and how much of it gets spent are separate questions, and this note addresses the second.
Third, the stock market. Retail investors' average daily trading value moved as follows.
| Period | Retail average daily trading value | Change from prior quarter |
|---|---|---|
| Q4 2025 | 34.5 trillion won | — |
| Q1 2026 | 57.2 trillion won | +65.8% |
| Q2 2026 | 65.9 trillion won | +15.2% |
| Q3 2026 | 37.2 trillion won | -43.6% |
Trading value matters for consumption statistics in two ways. One is the wealth effect: when share prices rise, households feel richer and spend more. The other is mechanical: brokerage commissions and similar fees are counted as household spending on financial services. That is why finance and insurance output fell 4.3% in the third quarter, and why the Bank of Korea named stock trading volume as a temporary drag on consumption. In Taiwan, rising prices and active trading pushed consumption up. In Korea, the third quarter went the opposite way. Kukmin Ilbo reported the note's warning that a stock correction could weaken both the wealth effect and financial-services spending, delaying the domestic recovery.
3. Taiwan waited too — a one-year lag, and what ended it
Reading the note as a verdict that Korea cannot do what Taiwan did misses half of it. Its starting point is that Taiwan's boom did not reach consumers straight away either.
| Period | Taiwan GDP growth | Taiwan private consumption growth | What happened |
|---|---|---|---|
| 2023 | 1.1% | — | Before the boom |
| 2024 | 5.3% | 3.2% | Growth accelerates from the second half, led by chip exports |
| 2025 | 8.8% | 1.4% | Growth speeds up while consumption slows; recovery begins in the fourth quarter |
| Q1 to Q3 2026 | 14.22% | 5.4% | Consumption recovers in earnest; Korea manages 2.1% over the same period |
| Q3 2026 | 14.36% | 5.37% | Fastest third-quarter growth in 48 years; second straight quarter of consumption above 5% |
The year 2025 is the key line. As growth jumped from 5.3% to 8.8%, consumption growth fell from 3.2% to 1.4%. That is the picture Korea presents now. It is the basis for the note's view that Korean consumption will also improve with a lag, and it lines up with the fact that Korea's boom began about a year after Taiwan's. Edaily and Kukmin Ilbo cite the note's figure that Taiwan's economy grew a cumulative 25.6% between the third quarter of 2024 and the third quarter of 2026.
What turned Taiwan's consumption around? The reported contents of the note point to three things: a strong stock market, fiscal support, and more overseas travel. According to Edaily, Taiwan's government paid cash to every resident between November 2025 and April 2026, worth about 420,000 won per person. Taiwan's statistics agency, the Directorate-General of Budget, Accounting and Statistics, said on October 30, 2026 that third-quarter consumption growth of 5.37% reflected a recovery in the car market and the wealth effect from record share prices.
The point to notice is that Taiwan's recovery did not come from wages alone. Even in an economy where 10.6% of workers are in IT manufacturing, what turned spending was share prices and a cash transfer. The note itself says that because much of Taiwan's improvement came from stock trading and overseas travel, the pace could change with share prices and the exchange rate. Taiwan's 5.4% does not stand on especially firm ground either.
Setting the two side by side shows what Korea lacks.
| What supported spending in Taiwan | Korea's position in 2026 |
|---|---|
| A broad IT manufacturing workforce (10.6%) | 1.9%; a narrow group of beneficiaries |
| A rising stock market and heavier trading | Retail trading value down about 44% in the third quarter |
| A universal cash payment (November 2025 to April 2026) | The note calls for improving middle-class spending capacity and easing living costs for low-income households |
| About a year from the start of the boom | The boom began in the second half of 2025; the third quarter of 2026 marks one year |
The note says the K-shaped structure constrains, in both countries, how far the warmth of growth spreads into domestic demand. The sentence quoted by ChosunBiz and JoongAng Ilbo is that improving the spending capacity of the middle class and easing the living-cost burden of low-income households is urgently needed to support the lower stroke of the K. It is worth recording that the research arm of an institution responsible for monetary policy wrote distribution and living costs into its prescription.
A warning came attached. As reported by Edaily, the authors said Korea and Taiwan are closely linked through the AI chip supply chain and both are running sharply larger current account surpluses, so if AI demand slows, shocks to the real economy and to financial markets could move together and amplify each other. Read alongside the rest of the note, the implication is that if exports turn before consumption has caught up, there is little else to hold growth up. The authors also argued that for the current supply-chain windfall to be more than another chip cycle, the two economies need broad productivity gains from wider AI adoption and need to ease structural constraints such as polarisation.
4. The November 26 rate decision — what is open, what is unconfirmed
The timing of the note is notable. The Bank of Korea's Monetary Policy Board, the seven-member committee that sets the base rate, holds its last meeting of 2026 on November 26. The base rate stands at 3.00%. The bank raised it in July and August, then held unanimously on October 22.
One thing should be clear first. No outlet reports that the note says anything about interest rates. It is an analysis by research staff, not a signal from the board. The views below come from outside the note.
| Who | What they said | Basis |
|---|---|---|
| Governor Shin Hyun Song (October 22) | A restrictive stance needs to continue; the bank will keep responding until it is confident inflation is converging stably on target | Press conference after the board meeting |
| Woori Finance Research Institute (reported November 9) | Expects a rise from 3.00% to 3.25% on November 26 | Chip-led growth, core inflation in the high 2% range, housing prices in the Seoul area |
| Lim Jae-kyun, KB Securities (Newsis, November 15) | Expects a further hike alongside higher growth and inflation forecasts, but a hold cannot be ruled out because falling share prices are shrinking household financial assets | The stock market correction |
| Bank of Korea (October 27) | Growth of 3.6% for the year is possible even with zero growth in the fourth quarter | Cumulative growth of 3.6% through the third quarter |
The case for raising rates is stacked up in numbers. Consumer prices rose 2.8% in October and core inflation was also 2.8%, both above the 2% target. Annual growth looks likely to be the highest since 4.7% in 2021. In the United States, expectations that the Federal Reserve will raise rates in December have firmed, with the size of the move now under discussion, and the gap between US and Korean policy rates is 1.00 percentage point at the upper bound.
What the note adds to that picture is a question. If 3.6% growth and a 0.2% fall in private consumption describe the same economy, which one should the rate be set for? A rate increase reaches indebted households and domestic-demand businesses before it reaches large chipmakers. A prescription meant to cool the upper stroke of the K can press harder on the lower one. That is not a sentence from the note. It is the question that appears when the note's diagnosis is laid over the rate calendar. It is at least notable that the reason the KB Securities analyst gives for a possible hold, shrinking household financial assets, points to the same place as the note's third reason.
| Open question | Where it stands |
|---|---|
| The note's full text | Not opened for this article; seven outlets' reports were cross-checked |
| Whether the note mentions interest rates | Not in any report |
| Korea versus Taiwan on household debt or dividends | Not found in the reported contents |
| Fourth-quarter private consumption | The bank attributed the third-quarter fall to temporary factors; the fourth-quarter advance estimate normally comes in January of the following year |
| Spending effect of 60 to 70 trillion won in bonuses | The bank's October 22 estimate; the new note says only part may be spent |
| The November 26 base rate | A hike is the more common forecast, with a hold also discussed; revised economic forecasts come the same day |
| $1 trillion in annual exports | Expected around November 20 on customs data |
| Taiwan's 5.37% third-quarter consumption growth | An advance estimate, subject to revision |
In short, the Bank of Korea's research staff treat the failure of exports to reach domestic demand as a matter of structure, not of the cycle. Chipmaking employs few people, money earned through higher prices goes first to profit, and that profit reaches households that spend less of it. Taiwan carried the same structural constraint and still turned consumption around within a year, with a broader workforce, a stock rally and a cash payment. Korea's one-year mark falls in the second half of 2026. Each time the export statistics set a new record, the figure worth checking alongside it is how many paychecks that record turns into.
Our earlier explainers cover the background: one on the trade balance explains why a surge in exports does not raise the surplus, or how people feel, by the same amount; the pieces on base effects and on working days show how much of a figure like 111.9% comes from the year-earlier number and the calendar; and the explainer on the memory supercycle traces how chip prices move Korean exports. Our October 26 report on the Bank of Korea's hold at 3.00% covers how the board left rates unchanged in October.
Sources
- ChosunBiz via Nate News — Bank of Korea: same chip boom, but Korea's consumption recovery may come later than Taiwan's (Korean, 2026-11-15)
- Edaily via Nate News — Taiwan is smiling thanks to chips, so why not Korea? (Korean, 2026-11-15)
- JoongAng Ilbo via Nate News — Taiwan's consumption rose 5% in the chip boom, Korea's only 2% (Korean, 2026-11-16)
- EBN via Nate News — The chip trickle-down reached Taiwan; why is it slow in Korea? (Korean, 2026-11-15)
- Kukmin Ilbo — Korea's chip trickle-down may arrive more slowly than Taiwan's (Korean, 2026-11-16)
- Digital Times — Wallets stay shut despite the chip boom; BOK says recovery may be slow (Korean, 2026-11-15)
- Seoul Economic TV — BOK: domestic consumption may recover slowly despite chip boom (Korean, 2026-11-15)
- Newsway — Exports and imports for November 1 to 10, Korea Customs Service preliminary data (Korean, 2026-11-11)
- Sisa Magazine — November 1 to 10 exports reach $33.5 billion (Korean, 2026-11-11)
- Newspim — October export volume index up 43.9%, Bank of Korea trade indices (Korean, 2026-11-13)
- Seoul Economic Daily — Surprise 1% growth in Q3; consumption falls for the first time in nearly four years (Korean, 2026-10-27)
- EBN — Chips lift Q3 growth to 1.0%; private consumption retreats (Korean, 2026-10-27)
- Taipei Times — Q3 GDP growth hits 14.36%, highest in 48 years (2026-10-31)
- Newsis via Nate News — With a Fed 'big step' in play, will the BOK raise rates at its next meeting? (Korean, 2026-11-15)
- Global Economic News — Woori Finance Research Institute expects BOK to raise the base rate to 3.25% in November (Korean, 2026-11-09)