The KOSPI Fear & Greed Index

Why a Korean sentiment page shows trade data. The KOSPI is, to an unusual degree, a basket of exporters. Samsung Electronics is around a fifth of the index by itself; add SK Hynix, Hyundai, Kia, POSCO and LG and the index largely tracks the earnings of companies that sell abroad. Export figures are therefore close to a direct read on the revenue line of the index's largest constituents — which is why they are watched by people who never trade Korean equities.

And they are fast. Korea publishes provisional figures three times a month — the 1st to 10th, the 1st to 20th, and the full month — while most large economies report trade once a month, four to six weeks after the fact. Because Korea sits early in the supply chain, shipping chips, displays and components that become other countries' finished goods, its export line is widely used as an early read on global demand, not only Korean demand.

What to look at. Direction and the working-day count, more than the level. A ten-day window can contain eight working days or eleven depending on where weekends and holidays fall — Chuseok alone can move it by two. Korea Customs publishes the working-day count for both the current and the year-ago window, which is what lets this panel state whether a comparison is like-for-like instead of assuming it. When the counts differ, the headline year-on-year figure is distorted and the per-working-day figure is the one to read.

Mind the base. Year-on-year changes on a series this volatile partly reflect how strong or weak the same ten days were a year ago. A large number is not self-explanatory. Read it alongside the year-to-date line, which smooths that out, and against the direction of recent periods.

What this is not:

• Not part of the Fear & Greed score. The index has four components and this is not one of them. It is context for reading the score, not an input to it.
• Not semiconductor-specific. This is total exports — autos, petrochemicals, ships, machinery and steel alongside chips.
• Not company revenue. These are customs clearance values in US dollars. A shipment clearing customs is not the same as revenue recognised, and currency moves affect the year-on-year comparison.
• Not final. Provisional figures are revised in the confirmed monthly statistics published later.

Source: Korea Customs Service trade statistics portal. Free and public, no key required.

It's a daily reading of Korean stock market sentiment on a 0–100 scale. A score near 0 means Extreme Fear — investors are risk-averse and selling. A score near 100 means Extreme Greed — investors are confident and buying. It's a contrarian reference point that describes the market's current mood, not a forecast of where prices go next.

It combines four components, each weighted equally at 25%: Market Momentum (the KOSPI versus its 125-day moving average), Volatility (downside deviation of daily KOSPI returns), Safe-Haven Demand (the USD/KRW exchange rate), and Foreign Investor Flow (net buying or selling by overseas investors). Each is normalized to a 0–100 score, then averaged.

Overseas investors are one of the three main forces in the Korean market, alongside domestic institutions and individuals, and their net buying or selling often sets the market's direction. Sustained foreign selling has historically pressured the KOSPI; foreign buying has supported rallies. We show foreign flow as context for understanding who is moving the market — it's informative, not a trading signal.

Extreme Fear (score 0–20) means sentiment is highly risk-averse: the KOSPI is weak relative to its trend, volatility is elevated, the won is under pressure, and foreign investors are typically net sellers. Extreme readings are often discussed as contrarian reference points — but markets can stay fearful for extended periods.

Extreme Greed (score 80–100) means sentiment is very confident: the KOSPI is well above its trend, volatility is low, the won is stable or strong, and foreign investors are typically net buyers. Markets can remain greedy for a long time, so a high reading describes mood — it isn't a sell signal.

KOSPI index levels, price history, and the USD/KRW exchange rate come from public market data. Foreign investor net-flow figures come from publicly reported Korean market data. The index is recomputed each Korean trading day after the market close.

No. It's a sentiment thermometer — a normalized, easy-to-read snapshot of market mood. It is not a forecast and should not be used on its own to make investment decisions. It's intended as one reference point among many. Always consult a qualified financial professional before investing.

It's recalculated each Korean trading day after the KOSPI closes. The page shows the most recent completed reading, plus the previous close, one week, one month, and one year prior for trend context.

Because the KOSPI is largely a basket of exporters. Samsung Electronics is around a fifth of the index on its own, and SK Hynix, Hyundai, Kia, POSCO and LG follow — so Korea's export figures are close to a direct read on the revenue of the index's biggest constituents. Korea also publishes them three times a month rather than once, which makes them one of the earliest hard numbers available anywhere on global demand. They are shown as context alongside the index, not as one of its four components.

Because without it a ten-day export figure cannot honestly be compared year on year. A ten-day window might contain eight working days this year and eleven last year depending on weekends and holidays — Chuseok alone can shift it by two — which would move the headline change dramatically for reasons that have nothing to do with demand. Korea Customs publishes the working-day count for both windows, so the comparison can be verified rather than assumed. When the two counts match, the change is like-for-like. When they differ, the panel says so and shows the per-working-day figure instead.

No. This is Korea's total goods exports — semiconductors alongside cars, ships, petrochemicals, steel and machinery. Chips are the largest single category, so the total does move with the memory cycle, but a change here cannot be attributed to semiconductors alone. Treat it as an economy-wide number.