Korea vs Taiwan — picking your Asia tech exposure as an Indian investor

For an Indian investor who wants Asia technology exposure beyond the US, two markets dominate the conversation: South Korea and Taiwan. They look interchangeable from a distance — both are semiconductor-heavy, both are export economies plugged into the same AI-hardware boom, both have one or two giant companies that swing the whole index. But they are not the same bet. Korea is, at its core, a memory-chip and broad-industrial play; Taiwan is a foundry monoculture built around a single company. The access routes differ, the tax treatment differs, and the concentration risk differs. Picking the wrong one for your thesis means you bought something other than what you thought.

This guide compares the two markets head to head — what each actually gives you, how an Indian resident accesses and is taxed on each, and how to decide between them (or, often, how to own both deliberately). It builds on our dedicated South Korea and Taiwan market hubs and the detailed semiconductor exposure: Taiwan, Korea, US compared breakdown.

Why this comparison even comes up for Indian investors

Most Indian investors already hold Indian equity and, increasingly, US equity. The question of Korea versus Taiwan arises specifically when someone wants Asia ex-China technology exposure that isn't already in their portfolio. A broad US ETF gives you Nvidia, AMD, and the chip designers and equipment names, but it does not give you the manufacturing side of the semiconductor supply chain — the memory makers and the leading-edge foundry. That manufacturing layer is overwhelmingly Korean and Taiwanese. So the choice is really "how do I add the chip-manufacturing leg of the AI trade," and Korea and Taiwan are the two doors.

It also comes up because both are accessible, treaty-friendly markets for Indians, unlike, say, mainland China (capital controls, Stock Connect quotas) or markets with punishing withholding. Korea's post-2023 opening and Taiwan's clean TSM ADR mean an Indian investor genuinely can express this view — which is why it's worth thinking it through rather than defaulting to whichever ticker is most familiar.

The core difference: memory vs foundry

Strip everything else away and this is the distinction that matters.

Korea is a memory bet with an industrial tail. Samsung Electronics and SK Hynix are the world's two dominant makers of DRAM and, increasingly, the high-bandwidth memory (HBM) that AI accelerators depend on. Together they have recently been around 42% of the KOSPI, and SK Hynix specifically has ridden the HBM wave to enormous gains. But Korea also has Hyundai (autos), shipbuilders, battery makers (the LG and Samsung SDI complex), and consumer names. So a Korea allocation is concentrated in memory chips but diversified beyond them in a way Taiwan is not.

Taiwan is a foundry monoculture. Taiwan Semiconductor Manufacturing Company (TSMC) makes the leading-edge logic chips that nearly every advanced processor in the world is fabricated on. TSMC alone is over 40% of the TAIEX and is Asia's most valuable company. Taiwan's market has become increasingly tied to TSMC and to global semiconductor demand, to the point of being detached from its own domestic economy. Owning Taiwan is, to a first approximation, owning TSMC with a thin wrapper of other names.

	South Korea	Taiwan
The bet	Memory (DRAM/HBM) + industrials	Leading-edge foundry (logic)
Anchor companies	Samsung Electronics, SK Hynix	TSMC
Index concentration	Samsung + Hynix ~42% of KOSPI	TSMC alone over 40% of TAIEX
Breadth beyond chips	Meaningful (autos, batteries, ships)	Thin
Cyclicality	Memory is famously boom-bust	Foundry is steadier but still cyclical

The investment implication: memory is the more cyclical, higher-beta side of the chip industry — DRAM/HBM pricing swings hard, which is why SK Hynix can double in a year. Foundry, TSMC's business, is structurally steadier (long customer relationships, technology lead, pricing power) but still rides the overall chip cycle. If you want the more aggressive, higher-variance chip exposure, that's Korea's memory tilt. If you want the closest thing to a "toll booth on advanced computing," that's TSMC.

How an Indian investor accesses each

The access stories are genuinely different, and this often decides the question before tax even enters.

South Korea — direct access got easier in 2023

Korea abolished its Investor Registration Certificate (IRC) requirement in December 2023, so foreign individuals can now open Korean accounts on a passport basis (no prior FSS registration). Routes for an Indian investor:

• Direct on the KRX (e.g. Samsung's 005930) via Interactive Brokers or a Korean local broker — needs a KRW account.
• London GDR (5SMSN.L) for Samsung specifically — USD, no Korean custody.
• EWY, the US-listed iShares MSCI South Korea ETF — buyable in any US brokerage, but a US-situs asset.
• Local KODEX 200 / TIGER 200 ETFs on the KRX.

The full Korea route map is in how to buy Samsung from India and KOSPI ETFs for Indian investors.

Taiwan — direct access stays paperwork-heavy

Taiwan never had Korea's 2023 liberalisation. Direct TWSE access for foreign individuals still runs through the FIDI / FINI registration regime plus a Taiwan custody account — paperwork-heavy enough that most retail foreigners skip it. The practical routes:

• TSM ADR on the NYSE — the dominant, cleanest offshore route for TSMC, buyable in any US brokerage.
• EWT, the US-listed iShares MSCI Taiwan ETF.
• Local Taiwan 50 (0050.TW) ETFs, if you do set up TWSE access.

The Taiwan routes are detailed on the Taiwan hub and the semiconductor exposure comparison.

The access verdict: Taiwan's saving grace is that TSMC has a clean US ADR, so getting TSMC is arguably easier than getting Samsung (which has no US ADR). But Korea's direct-market access is now broadly easier than Taiwan's. So: easier to get the single company in Taiwan (TSM ADR); easier to get the whole market directly in Korea (post-IRC).

Tax: side by side for an Indian resident

Both markets withhold at source on dividends, and India taxes you on worldwide income with a treaty credit. The rates differ.

	South Korea	Taiwan
Statutory dividend WHT (non-resident)	22% (20% + 2% local)	21%
India treaty (DTAA) dividend rate	15%	12.5%
Local capital-gains tax on listed shares	Generally exempt for small foreign holders	0% on listed-share CG
Major-shareholder CGT trigger	Yes — KRW 5bn per company (a 2025 plan to cut to KRW 1bn was dropped)	n/a in the same form

A few things stand out. Taiwan's treaty dividend rate (12.5%) is lower than Korea's (15%) — a small edge to Taiwan on dividend friction, though both are far below the US's 25% treaty rate. Both markets are friendly on capital gains for ordinary foreign retail holders. Korea's major-shareholder CGT threshold (currently KRW 5 billion per company; a 2025 plan to lower it to KRW 1 billion was dropped in September 2025, though the figure has moved around) is something to watch only if you build a very large single-stock position.

The Korea dividend mechanics — claiming 15% at source, the 2026 beneficial-ownership documentation, and the Form 67 (being renumbered Form 44 from TY2026-27) credit in India — are covered in full in Korea dividend WHT and DTAA. Taiwan's equivalent is on the Taiwan hub.

The estate-tax wildcard

Here is a subtlety that cuts the same way in both markets but bites hardest on the easy routes. The most convenient way to own either market from India is a US-listed wrapper — EWY for Korea, EWT or the TSM ADR for Taiwan. All of these are US-situs assets, which means they sit inside the US estate-tax $60,000 trap: everything above $60,000 of US-situs assets is exposed to up to 40% US estate tax, with no India–US estate treaty to soften it.

So the convenience routes for both Korea and Taiwan quietly add to your US-situs pile. The estate-tax-clean alternatives are the local listings — KODEX 200 / direct KRX for Korea, local 0050.TW / direct TWSE for Taiwan — and, for Korea specifically, the London GDR. This is a genuine point in favour of going direct/local once your positions are large, in either market. It's covered fully in the US estate-tax guide.

Indian-side tax (identical for both)

On the Indian side, the treatment is the same regardless of which market: shares or ETF units bought under the LRS are foreign assets, taxed at 12.5% LTCG without indexation after 24 months or slab-rate STCG within 24 months, with no Rs 1.25 lakh exemption. See how foreign / US stocks are taxed in India. Both must be disclosed on Schedule FA every year, and both consume LRS headroom — model the remittance and TCS with the LRS and TCS calculator.

Concentration and correlation — the risk you're really taking

Neither market is "diversified Asia tech." Both are single-name-dominated, and that is the central risk.

• In Taiwan, the single-name risk is starker. TSMC is over 40% of the index and the market trades almost as a TSMC proxy. If TSMC stumbles — a geopolitical shock, a technology misstep, a customer concentration issue — Taiwan-market exposure has very little to cushion it.
• In Korea, the single-name risk is real but cushioned. Samsung-plus-Hynix is ~42% of the KOSPI, but the remaining ~58% includes autos, batteries, shipbuilding, and consumer names, so a Samsung wobble is partly offset by the rest.

And critically, Korea and Taiwan are not a diversification pair. Both are leveraged to the same global semiconductor and AI-hardware cycle. Owning both does not hedge you against a chip downturn — it doubles down on it. If you want diversification, the counterweight to Asia chips is non-tech and non-Asia exposure (a broad US or global allocation), not the other chip market. The geopolitical risk also rhymes: both sit in a region where US–China tension and, for Taiwan especially, cross-strait risk are persistent overhangs.

The macro and geopolitical backdrop

The two markets share a region but not a risk profile, and the macro context shapes both.

Korea is a broad industrial export economy — it sells chips, but also cars, ships, batteries, steel, and consumer electronics into global markets. Its equity market has historically traded at a valuation discount (the so-called "Korea discount") attributed to governance concerns, chaebol structures, and geopolitical proximity to North Korea. Ongoing corporate-governance and shareholder-return reforms have been a slow-burning re-rating story. The KOSPI's strength through 2025 was driven overwhelmingly by the memory names riding AI demand, but the underlying economy is more diversified than the index's chip concentration suggests.

Taiwan is, economically and in market terms, increasingly the TSMC story. The market has become tightly coupled to global semiconductor demand and decoupled from Taiwan's domestic economy. The dominant overhang is cross-strait geopolitical risk — the possibility of China–Taiwan tension disrupting the world's most important chip foundry. That tail risk is precisely why some investors prefer to take their leading-edge-logic exposure through TSMC's diversified global customer base rather than betting on Taiwan-the-market, and why others avoid concentrated Taiwan exposure altogether.

For an Indian investor, both markets also carry the standard emerging-Asia overlay: sensitivity to global risk appetite, USD strength, and the broader semiconductor capex cycle. Neither is a defensive holding.

A worked allocation example

Suppose you have decided you want Asia-tech exposure as a satellite — say 10% of an international equity sleeve — and you're choosing how to split it. Three reasonable stances:

Stance	Korea	Taiwan	Rationale
Memory-cycle conviction	70%	30%	You believe HBM/DRAM is the highest-growth chip segment
Foundry conviction	30%	70%	You believe TSMC's leading-edge logic is the indispensable bet
Balanced chip-cycle	50%	50%	You want broad Asia-chip exposure without picking a winner

Two caveats on any of these. First, this is not diversification — both sleeves rise and fall with the same chip cycle, so the "balanced" split reduces single-name risk, not chip-cycle risk. Your actual diversifier is the other 90% of the international sleeve in non-Asia, non-chip assets. Second, at this size, mind the wrapper: if the 10% sleeve is large in absolute terms and you're using US-listed EWY/EWT/TSM, you're stacking US-situs exposure — consider the local/direct routes to keep the estate-tax line clean.

So which should you pick?

The honest framework:

Pick Korea if you want the higher-beta, more cyclical memory/HBM exposure and you value having some non-chip ballast (autos, batteries) in the same allocation. Korea also rewards you with genuinely easier direct-market access post-2023 if you want to own the underlying rather than a US wrapper. The trade-off is a slightly higher treaty dividend rate (15% vs 12.5%) and Samsung's lack of a US ADR.

Pick Taiwan if your thesis is specifically "leading-edge logic / TSMC is the indispensable company of the AI era." TSMC's clean US ADR makes the single-name bet the easiest of any Asia-tech name to own, and the treaty dividend rate is marginally lower. The trade-off is brutal single-name concentration and a market that is essentially one company.

Own both — deliberately — if you want full Asia-chip-cycle exposure and understand you are concentrating, not diversifying. If you do, weight them by conviction (memory vs foundry), keep the combined Asia-tech sleeve modest as a satellite to a broad core, and prefer the local/direct routes at scale to keep US estate-tax exposure down.

Pick neither, or keep it tiny, if you already get incidental Korea/Taiwan exposure through a broad emerging-market or global ETF (most do) and don't have a specific chip-cycle view. Concentrated single-country EM tilts are for expressing a thesis, not for default allocation.

A note on the US route — and its hidden cost

It bears repeating because it's the choice most Indians actually make: the easiest way to own either market is a US-listed instrument (EWY for Korea; EWT or the TSM ADR for Taiwan), bought in a Vested, INDmoney, or Interactive Brokers account. That convenience is real and, at small position sizes, entirely reasonable.

But all three are US-situs assets, and they accumulate alongside any other US holdings (your VTI, your individual US stocks) toward the $60,000 US estate-tax threshold. An Indian investor who builds a meaningful US-direct portfolio plus a US-wrapped Asia-tech sleeve can cross that line faster than they realise — and the estate-tax exposure is up to 40% with no India–US treaty relief. The clean alternatives are local: KODEX 200 / direct KRX for Korea, local 0050.TW / direct TWSE for Taiwan, or Samsung's London GDR. This is not a reason to avoid the US route at small scale — it's a reason to switch to local/direct as the position grows, in either market. The full mechanics and workarounds are in the US estate-tax guide.

Whatever you choose, treat it as the high-conviction satellite it is, not a core holding. Read the route-specific guides before committing — Samsung from India, KOSPI ETFs, and the Korea dividend WHT and DTAA mechanics on the Korea side; the Taiwan hub and semiconductor exposure comparison on the Taiwan side. For the wider context, the South Korea hub and the full markets index.

---

This is general information, not tax or investment advice. Single-country, single-sector exposure is highly concentrated and volatile, and geopolitical risk is material in both markets. Tax rates, treaty terms, and access rules change. Figures reflect data and rules as understood in early 2026 — verify current details with your broker and a qualified cross-border tax advisor before acting.