Private EducationThe students are vanishing. The spending keeps breaking records.
Private Education
The students are vanishing. The spending keeps breaking records.
Korea is losing its children. The school-age population has fallen by half from its peak, first-grade enrollment just dropped below 300,000 for the first time, and yet private-education spending set its fourth straight record in 2024 at KRW 29.2 trillion — up roughly 60 percent in a decade while the student count fell. This is a study of that paradox — a consumer market that grows as its customer base disappears, because a shrinking cohort competes harder for a fixed pool of elite university and medical seats, and each remaining student spends more — read through Korea's scale leader in after-school tutoring, Megastudy.
A Korean exam-prep lecture hall at full intensity. The students filling rooms like this are fewer every year — the cohort behind them is shrinking on a demographic curve that does not reverse — and yet the money spent on each of them keeps setting records. That contradiction is where this study lives: a private-education market that grows in won even as it loses bodies, read through the company that captured the intensity, Megastudy.
本研究简报以英文发布。
Korea is one of the few consumer markets on earth that grows as its customer base disappears. The school-age population has fallen by roughly half from its 1980 peak; the count of elementary, middle and high-school students slips below five million for the first time in 2026; and first-grade enrollment has just dropped below 300,000 for the first time on record, headed toward an estimated 232,000 by 2030 — a quarter fewer first-graders in three years. And yet private-education spending on those same students hit a fourth consecutive record in 2024, KRW 29.2 trillion, up 7.7 percent in a single year and about 60 percent across the decade. The students are vanishing; the spending keeps breaking records. This is a study of that paradox — what drives it, how durable it is, and what could finally break it — read through Korea's scale leader in after-school tutoring, Megastudy.
The resolution of the paradox is competition. A shrinking cohort does not relax — it fights harder for a fixed pool of elite university and, above all, medical-school seats, and that intensity lifts spend-per-student faster than demographics cut the base. Per-student monthly spending reached an all-time-high KRW 474,000 in 2024, up 9.3 percent in a year the student population fell; for students actually enrolled in tutoring, KRW 592,000, with participation at 80 percent. So demographics set the base, and competition intensity sets the spend-per-head — and the second has been winning. Read the market off the demographic curve and you will get it exactly backwards. Read it off spend-per-student multiplied by competition intensity, and the records make sense.
The market grows as the cohort shrinks
KRW 29.2tn
Total private-education spending on elementary/middle/high students in 2024 — a 4th consecutive record, +7.7% YoY and up ~60% from KRW 18.2tn in 2014, even as the student population shrank. Per-student monthly spend hit an all-time high of KRW 474,000 (+9.3%), with the rise led by elementary (aggregate spend +74.1% over the decade). Demographics shrink the base; spend-per-head more than offsets it. Source: Statistics Korea, 2024 Private Education Expenditures Survey.
Lay the two trajectories against each other and the paradox is stark. Spending ran KRW 23.4 trillion in 2021, then 26.0, then 27.1, then 29.2 trillion in 2024 — four years, four records. Over the same kind of horizon the cohort only fell: the count of elementary-through-high students drops from 5.02 million in 2025 to 4.84 million in 2026, the first time below five million, and the leading edge of the decline — first-grade entry — is the sharpest, below 300,000 in 2026 and projected to lose roughly a quarter of its 2023 level by 2030. The spend line points up; the body-count line points down; the gap between them is the whole market. It is not evenly spread, either: the decade's growth was led by elementary students, whose aggregate spend jumped 74.1 percent at 87.7 percent participation — families now buy 'advance study' for the high-school race years before it begins.
Year
Private-ed spend
Students
Read
2021
KRW 23.4tn
~5.3m
Post-COVID rebound
2023
KRW 27.1tn
~5.2m
Records resume
2024
KRW 29.2tn
~5.1m
4th straight record (+7.7%)
2026
(rising)
4.84m
Below 5m for first time
2030
(rising)
1st-grade ~232k
−25.8% entry vs 2023
The paradox in two trajectories: private-education spend versus the student cohort, past to forecast. Spending is total elementary/middle/high private-education outlay (Statistics Korea, 2021–2024 surveys); the cohort row is the elementary-through-high student count and, in 2030, the projected entry-cohort signal (first-grade enrollment ~232k, −25.8% vs 2023). Spend grows as bodies fall. Sources: Statistics Korea; Korea Ministry of Education school-age projections.
The demand engine: the medical race and the repeat exam
If per-student spend sets the market's level, the high-school exam-prep segment sets its cycle — and that cycle runs on policy. The decisive event of the current upswing was the 2025 medical-school quota expansion: an announced increase of 2,000 seats, the first major rise in 27 years since 1998, that settled after universities' voluntary cuts to a net 1,509 additional places. Because a medical-school place is the most-prized outcome in Korean education, widening that funnel pulled high-achievers back to re-take the national exam. Repeat test-takers for the 2025 exam numbered 161,784, the highest in 21 years, roughly one-third of all 522,670 applicants. Repeat-prep — online all-pass subscriptions plus premium boarding academies — is the highest-ARPU product in the sector, so a repeat-taker surge flows straight into the high-school market that the scale leader dominates.
The intensity engine. The medical-school race and the repeat exam are what convert a shrinking cohort into a growing market: a smaller class of high-achievers competing for a fixed, widened pool of elite seats, many of them sitting the national exam a second or third time. Repeat-prep is the sector's highest-value product, and the 2025 repeat-taker count hit a 21-year high. This is the demand the high-school core is geared to — and the demand the 2028 reform is scheduled to cool.
The same policy machinery that built the tailwind has scheduled its reversal. The 2028 admission reform replaces the current relative grading of high-school records with a five-tier absolute scale and raises the weight of internal records in admissions, which shifts demand toward in-school, offline support and away from the pure exam-cramming the online-pass model was built on. Crucially for timing, repeat-takers are expected to stay elevated — even to peak — through the 2027 cycle, the last exam under the old rules; some houses project the 2027 cohort as a record. The structural cut lands from the 2028 cycle, when graduates lose evaluation on the advanced and elective subjects that rewarded re-taking and top universities restrict their eligibility in record-based admission. So the shape is a bump, then a cliff — a near-term repeat-exam tailwind into 2027, then a forward repeat-exam cliff from 2028. That single policy outcome is the most important thing to diligence in this sector.
The cycle that policy built — and scheduled to cool
161,784
Repeat test-takers for the 2025 national exam, the highest in 21 years and ~1/3 of all applicants — driven by the 2025 medical-school quota expansion (announced +2,000, net +1,509 seats; first major rise in 27 years). Repeat-prep is the sector's highest-ARPU product. Houses project repeat-takers stay elevated/peak through the 2027 cycle (last under the old rules); the 2028 admission reform is the scheduled cooling. Sources: Korea Institute for Curriculum and Evaluation; Ministry of Health and Welfare; Ministry of Education.
The case in point: growth through the decline
Megastudy is worth studying because it is the cleanest expression of the paradox: a company that nearly doubled revenue through a demographic decline and now sits on a high plateau, defending margin through mix and price. Consolidated revenue ran from KRW 474.7 billion in 2020 to a peak of KRW 942.2 billion in 2024 — a 1.98x rise as the cohort fell — tracking the post-COVID return to in-person prep and the intensifying medical-school and repeat-exam race. The 2025 figure stepped down 6.1 percent to KRW 884.9 billion, but operating profit barely moved, from KRW 123.6 billion to KRW 121.0 billion, and operating margin actually improved to 13.7 percent. That last fact is the tell: this was a margin-stable year, not a demand collapse.
Revenue nearly doubled through a shrinking cohort, then leveled at a steady margin
The 2025 step-down is, importantly, not a single story. It is a roughly three-way mix: the margin-accretive wind-down of the loss-making civil-service exam business — divested end-2024 for KRW 11.4 billion after a roughly KRW 60 billion cumulative four-year operating loss — plus demographic erosion in the elementary-and-middle segment, plus a one-off competitive hit to the offline boarding business. The divestiture's real value is what it did to the bottom line: it removed the only structurally loss-making division, which is why profit held even as revenue fell. The high-school segment remains the engine at roughly 60 percent of consolidated revenue and the most profitable line — online all-pass subscriptions, raised in price in 2025–26, plus the boarding repeat-prep academies. The strategic logic of every recent move is consistent: up the ARPU ladder and out of the loss-makers, a deliberate re-mix for a maturing market rather than a growth sprint.
The margin lever. The online all-pass subscription is the high-margin heart of the high-school core: a star-instructor lecture library that, once produced, scales to more subscribers at near-zero incremental cost, with prices the leader has been able to raise precisely because competition intensity makes families price-insensitive on the highest-stakes product. A mix-shift toward online passes lifts blended margin even as total volume plateaus — which is how the leader held a steady margin through a falling top line.
Leader versus disruptor
On absolute scale, Megastudy dominates — roughly 3.5 times the size of the next listed high-school incumbent, Digital Daesung, and spanning high-school, adult/professional and elementary-through-middle where peers are narrower. But the competitive energy sits elsewhere. The fast-rising challenger is Sidae Injae, the premium repeat-prep operator run by Hi-Consi: privately held, focused on the highest-value boarding segment, it grew 15.3 percent to KRW 381.8 billion in fiscal 2024 while trade press described it as the upstart stalling the traditional players' growth. Among the listed peers, Digital Daesung posted a fiscal-2025 record — KRW 253.8 billion at a roughly 12.4 percent margin — lifted partly by consolidating a medical-track boarding subsidiary, while Visang Education, often mis-cast as a laggard, is in fact a profitable textbook-and-publishing-led recovery, not a loss-maker. The leader's scale is secure; it is the growth edge in premium repeat-prep that is contested.
Player
Revenue
Margin
Role
Megastudy
KRW 884.9bn
13.7%
Diversified scale leader
Sidae Injae / Hi-Consi
KRW 381.8bn
~9.3%
Premium repeat-prep disruptor (+15.3%, FY24)
Digital Daesung
KRW 253.8bn
~12.4%
#2 listed high-school incumbent (FY25 record)
Visang Education
KRW 268.8bn
~8.0%
Textbook/publishing-led recovery
The competitive hierarchy by latest-year revenue: a diversified scale leader against a focused premium disruptor. Megastudy and Digital Daesung are FY2025; Visang is FY2025; Sidae Injae / Hi-Consi is private and FY2024 (latest audited) — do not read its +15.3% growth as same-period versus the incumbents' FY2025 results. Margins are operating margin. Sources: DART (Megastudy, Digital Daesung); ETNews/Eduplus (Sidae Injae); DigitalToday/FnGuide (Visang).
Two qualifications matter for how the disruption reads. First, Sidae Injae's growth is concentrated in premium boarding — the highest-ARPU product — where it opened a roughly 1,500-capacity academy in early 2025, billed as a single-academy maximum, challenging the incumbent boarding trio. Second, the link from that expansion to the leader's softer offline year is analyst interpretation, not a sourced fact: Megastudy itself attributes its offline weakness to a shrinking repeat-student market and higher offline operating costs, and the offline-division operating profit fell far more than its revenue, consistent with a one-off competitive and cost hit rather than a structural loss of position. The clean read is that a focused challenger can take share in one high-value segment without threatening the leader's overall scale — the risk is a growth-edge erosion, not displacement.
Profit held while revenue fell — the margin tell
13.1% to 13.7%
Megastudy's operating margin actually rose in FY2025 even as revenue fell 6.1% (KRW 942.2bn to 884.9bn) — because operating profit held at KRW 121.0bn (from 123.6bn) and the loss-making civil-service unit was divested. The strongest single piece of evidence that the 2025 step-down was a deliberate re-mix and demographic/competitive adjustment, not a demand collapse. The civil-service exit is not yet fully clean in the books, so the full margin benefit shows from FY2026. Source: DART FY2025 Annual Report (rcept_no 20260318000849).
Where the market goes to 2030
Project the market forward and it turns on two variables pulling against each other. The demographic floor sets how fast the shrinking cohort caps the headcount even rising spend can monetize — already visible below the high-school stage, in elementary-and-middle student numbers and school closures, and in a 2025 survey that showed elementary-and-middle private-ed spend itself dipping as participation fell. The repeat-exam cliff sets how much the 2028 reform cools the high-school cash engine, after a likely demand bump into 2027. Against both runs the per-student-spend tailwind — credential competition in a winner-take-most labor market — that has not reversed with the birth rate, plus an edutech and AI-tutoring wildcard that is at once a cost lever and a new competitive front. The market environment for a scaled, diversified leader sorts into three regimes, set by whether intensity and the new in-school/offline demand keep outrunning the cohort decline through the reform.
The floor under everything. This is the demographic the whole market is built on, and it is shrinking on a curve that does not reverse — half the school-age children of a generation ago, first-grade entry below 300,000 for the first time, schools with single-digit incoming classes. For two decades, intensity per remaining student has outrun this decline; the forward question is whether it keeps doing so once the cohort floor and the 2028 reform arrive together. Demographics are the headwind every player runs into; intensity is the tailwind that, so far, has been stronger.
Regime
What would have to be true in the market
Trajectory to 2030
Intensity wins
Competition intensity keeps lifting per-student spend faster than the cohort shrinks; the 2028 reform expands the offline/naesin market more than it cuts repeat-prep; adult and edutech lines scale.
Mid-single-digit growth; margin widens past the mid-teens
Managed plateau
Per-student spend roughly offsets the demographic drag; a 2027 repeat-exam air-pocket as the reform bites, then mild recovery on naesin/offline and ASP.
Roughly flat; margin steady ~13–14%
The squeeze wins
The 2028 reform sharply cuts the repeat-exam cycle and the online-pass model; demographics outrun per-student spend; adult and edutech fail to offset.
The high-school core turns down; contraction, not crisis
Three demand regimes for the Korean high-school-led private-education market to 2030 — market-environment constructs, not a company forecast, target or valuation. Each is defined by whether per-student spend and the new naesin/offline demand offset the demographic decline and the 2028-reform repeat-exam cliff; the revenue path is illustrative, anchored to the leader's segment trajectory and the sector's blended dynamics.
“A market that grows as its customers disappear is a bet on intensity, not on the birth rate. The decisive question is never how many students remain — it is whether a shrinking cohort keeps competing hard enough to spend more per head, and whether the policy machinery that built that intensity leaves it standing.”
— Nathan Research Group, Korean Education Series N°01
The lesson generalizes past education. Any market where the customer count falls while the value extracted per customer climbs — premium healthcare for aging populations, status goods, credential competition itself — rewards the same discipline: model it off intensity per head, not off the headcount curve, and watch the policy or structural variable that could break the intensity rather than the demographic that everyone already sees. Megastudy is a clean case because it monetized the paradox in plain sight — doubling through a decline, then defending a steady margin through mix and price — and because its single largest forward risk is not the shrinking cohort but a scheduled change in admission rules. Our full brief on the Korean private-education market — the spend-versus-cohort mechanics, the medical-quota and repeat-exam cycle, the 2028-reform swing, the competitive map and the three-regime outlook — is available to download with this article.
与 NRG 合作的方式
与 Nathan Research Group 合作。
Statistics-Korea surveys and DART filings establish the shape of this market — the record spend, the collapsing cohort, the repeat-exam counts, the segment splits. What they cannot show is how admission offices will actually weight school records once the 2028 reform lands and the repeat-exam engine cools, whether a focused premium challenger's instructor-poaching and boarding build-out is durably taking the highest-value students, or whether the adult and edutech lines can scale into a real demographic hedge. That intelligence lives with the educators, admissions specialists and boarding operators who built, taught in and competed with these businesses — and reaching them, compliantly, is what we do.
韩国首家专家网络 — 首尔，2013年至今
我们会请谁到场
Admissions & policy (CSAT / naesin / 2028 reform)
Admission-office and policy specialists who read how the 2028 overhaul will really re-weight school records versus the national exam — and thus the true size of the repeat-exam cliff the market is bracing for.
Boarding-academy (repeat-prep) economics
Operators of premium repeat-prep boarding academies on the real unit economics — instructor contracts, capacity, occupancy — of the sector's highest-ARPU product and most contested battleground.
Online subscription & ASP (the all-pass model)
Product leads behind all-pass lecture subscriptions on how far prices can rise on a captive, high-intent base before intensity-driven price-insensitivity finally breaks.
Megastudy consolidated revenue and operating margin, FY2019–FY2025 (DART, rcept_no 20260318000849). Revenue nearly doubled to a KRW 942.2bn peak in 2024 while Korea's student population fell, then eased to KRW 884.9bn in 2025; operating margin held in the low-to-mid teens throughout (2020's dip was COVID disruption to in-person prep). Shown as market evidence — the demographic paradox monetized — not as a forecast.
Medical-track admissions & the N-su cycle
Specialists on the med-school quota and the repeat-test-taker cycle — whether the 21-year-high intensity persists through the transition or fades, the swing that sets the high-school market's trajectory.
Adult, professional & university-transfer education
Contacts across the transfer, licensing and professional-exam markets — the counter-cyclical, demographically-insulated lines that decide whether the leader can diversify off the policy-driven high-school cycle.
Edutech & AI-tutoring delivery
Specialists on AI-personalized learning and the government's AI digital-textbook initiative — the new delivery layer that is at once a cost lever for incumbents and an entry ramp for challengers.
项目怎么推进
1范围界定
把您的投资逻辑，转成具体的专家画像和问题清单，对准您需要落定的那几个决定。
2寻访与核实
每一位专家在访谈前都要过一遍：课题相关性、经验的时效性，以及有没有利益冲突。
3组织与归纳
按您的时间安排访谈；需要的话，另附一份按问题逐条对应的书面归纳。
如果您的团队正在评估 Megastudy、the Korean private-education market，乃至范围更大的 education sector，请告诉我们您眼下要做的是什么决定。
