Daily Economics, Business and Finance Study
October 10, 2026 (KST) · Day 20
Today's Topic: With Investor Flows, Why They Bought Matters More Than Who Bought — Reading Foreign, Institutional, Retail and Program Trading
1. Review of yesterday's key points
Day 19 showed that RSI and MACD are derived indicators that recalculate price, so overbought readings, crossovers and divergences should not be treated as independent predictive signals. Today, we will read the investor-category trading records left behind price and volume, while learning to evaluate why the orders appeared and whether they can persist rather than stopping at the result that “foreign investors bought.”
2. Today's core concept
Investor-flow data aggregates how much each investor category bought and sold over a period and is an after-the-fact record.
\[ Net\ purchases=Purchases-Sales \]
- A positive result means the category bought more than it sold during the period.
- A negative result means net sales.
- Every purchase has a sale on the other side, so total market purchases and sales come in matching pairs.
- Foreign, institutional and retail investors are not single entities united by one investment objective.
- Program trading is closer to how an order was placed than who placed it, so it can overlap with investor classifications.
Today's learning objectives are:
- Distinguish net purchases from gross purchases and sales.
- Understand that foreign, institutional and retail investor classifications and program trading are different classification axes.
- Calculate why the meaning of the same net purchase amount changes with liquidity, market capitalization and order purpose.
- Ask whether a flow came from index inclusion, ETF activity, derivatives hedging, short covering or fundamental buying.
- Use flow data together with price, volume, earnings, valuation and disclosures.
3. Concept explained
3-1. Every Execution Has a Buyer and a Seller at the Same Time
For one share to trade, someone willing to buy and someone willing to sell must meet at the same price. If trading volume is one million shares, someone bought one million shares and someone else sold one million shares.
When those trades are divided by investor category, the following relationship generally holds within the same market, period and classification scope.
\[ Foreign\ net\ purchases+Institutional\ net\ purchases+Retail\ net\ purchases+Other\ net\ purchases\approx0 \]
Displayed totals may not equal exactly zero because of detailed categories, whether off-exchange trades are included, corrections and rounding. The key question is not whether “only buying entered the market,” but which category absorbed the shares supplied by another category.
Net purchases also do not mean that new money entered the company. Unless the trade occurred in the primary market through an issuance such as a rights offering, money paid in the secondary market goes to the investor who sold the shares.
3-2. Net Purchases and Gross Purchases Convey Different Information
If foreign investors buy KRW 100 billion and sell KRW 95 billion in one day, their net purchases are KRW 5 billion.
\[ 100-95=KRW\ 5\ billion \]
They would also have net purchases of KRW 5 billion if they bought KRW 6 billion and sold KRW 1 billion. The trading activity, however, would be different.
| Category | Purchases | Sales | Net purchases | Difference in interpretation |
|---|---|---|---|---|
| Case A | KRW 100 billion | KRW 95 billion | +KRW 5 billion | Both buying and selling were extremely active |
| Case B | KRW 6 billion | KRW 1 billion | +KRW 5 billion | Participation was small, but direction was net buying |
Net purchases alone cannot tell you whether the result came from heavy two-way turnover or one-directional orders. Whenever possible, review gross purchases, gross sales and total trading value together.
3-3. “Foreign Investors” Are Not a Single Investor Sharing One View
The foreign-investor category can contain accounts with very different objectives.
- Long-term fundamental investing
- Passive money tracking global indexes
- Orders arising from ETF creations and redemptions
- Country and sector weight adjustments
- Asset allocation driven by exchange-rate changes
- Cash-equity hedges linked to futures and options
- Arbitrage, high-frequency and short-term strategies
- Short covering that buys back borrowed shares
Therefore, several days of foreign net buying does not mean all of it reflected the same judgment about corporate value. A large net purchase can appear at the index-inclusion or rebalance close and disappear the next day.
Foreign flows are better read alongside the KRW/USD exchange rate, global risk appetite, international sector comparisons, futures positions and index-change schedules.
3-4. Institutions Are Not One Entity Either, and Their Capital Has Different Characteristics
The institutional category can include securities firms, asset managers and investment trusts, pension funds, insurers, banks and private funds. Detailed names and coverage can differ by data provider and market.
| Example institution type | Factors that may generate trades |
|---|---|
| Financial investment | Derivatives hedging, liquidity provision and proprietary trading |
| Investment trusts | Fund subscriptions and redemptions, benchmark tracking and manager decisions |
| Pension funds | Long-term asset allocation, target-weight adjustment and periodic rebalancing |
| Insurers and banks | Asset-liability management, regulatory capital and interest-rate outlook |
| Private and other funds | Absolute-return strategies, event investing and arbitrage |
Even when institutions are net buyers, persistence and meaning depend on which subcategory led. A pension fund's long-term increase in allocation and a securities firm's futures hedge are both “institutional net buying,” but they should be interpreted differently.
3-5. Program Trading Is Closer to an Order Method Than an Investor Identity
Program trading refers to trades in which computer systems place orders under predefined conditions, often across multiple securities or through linked cash and derivatives positions. Korea Exchange statistics distinguish program trades between arbitrage and non-arbitrage activity.
- Arbitrage trading: Trading designed to exploit price differences between linked assets such as cash securities and futures
- Non-arbitrage trading: Program orders other than arbitrage, potentially including index-tracking baskets, rebalancing and large portfolio adjustments
Program trading is not a fourth independent investor group to add to foreign, institutional and retail investors. If a foreign account buys shares through a program order, the transaction can be recorded as both foreign buying and program buying.
\[ Investor\ classification=Who\ traded \]
\[ Program\ classification=How\ the\ trade\ was\ placed \]
Therefore, simply adding “KRW 100 billion of foreign net purchases + KRW 70 billion of program net purchases = KRW 170 billion from separate buyers” may double count the same trades.
3-6. Price Is Moved Not Only by the Net Purchase Figure but Also by Order Aggressiveness
Purchases and sales are equal in every execution, yet prices change. The difference is which side is more urgent and accepts the other side's quoted prices.
- Price can rise when buyers repeatedly lift higher sell quotes.
- Price can fall when sellers repeatedly hit lower buy quotes.
- A stock with a thin order book can move sharply on a small net purchase amount.
- A highly liquid large-cap stock may experience less price impact from the same amount.
Index-rebalance orders entered in the closing auction can be large but may represent mechanical weight matching rather than an information-driven revaluation. Conversely, even a small net purchase can have a large impact if buyers aggressively cross a thin set of offers.
3-7. Normalize Absolute Amounts by Trading Value and Market Capitalization
The meaning of KRW 10 billion in net purchases differs across stocks. The same KRW 10 billion entering a small or mid-cap stock with KRW 20 billion of trading value does not have the same influence as it does in a large-cap stock trading KRW 2 trillion.
Two useful supplementary ratios are as follows.
\[ Net\ purchase\ to\ trading\ value = \frac{Net\ purchase\ amount}{Total\ trading\ value}\times100 \]
\[ Net\ purchase\ to\ market\ capitalization = \frac{Net\ purchase\ amount}{Market\ capitalization}\times100 \]
The first ratio shows the relative size of the net purchase within that day's trading, while the second shows its size relative to the company's total market value. Neither reveals why the buying occurred or what future returns will be, but both reduce scale differences across stocks.
3-8. Examine Accumulation, Persistence and Breadth Rather Than One Day
A large single-day flow may arise from a block trade, index change, derivatives expiration, ETF rebalance or one-off large order. Interpretation improves when you also examine the following.
- Five-day and 20-day cumulative net purchases rather than one day
- Number of net-buying and net-selling days
- Whether flows are concentrated in one stock or spread across the sector
- Whether buying occurred throughout the session or was concentrated in the closing auction
- Whether the investor chased gains by buying on up days and selling on down days
- Whether the investor continues buying through price pullbacks
- Whether earnings estimates and guidance are being raised at the same time
Persistent net buying is not automatically positive. Buyers can be wrong, and they can lose money after chasing a high price. Flow data observes participant behavior; it does not determine which participant is smarter.
3-9. Co-Movement Between Flows and Price Does Not Prove Causation
Several explanations are possible when foreign net buying and a price increase occur on the same day.
- Aggressive foreign buying caused the price to rise.
- Positive earnings news caused both the price increase and foreign buying.
- Trend-following capital bought after observing the price rise.
- Index inclusion and passive buying lifted the closing price.
- Short covering was recorded as buying.
Therefore, a simple correlation should not be reduced to the single claim that “the price rose because foreigners bought.” Review disclosure timing, earnings expectations, volume, exchange rates, index schedules, futures and options expiration and short-selling data together.
4. A simple numerical example
The following is hypothetical one-day flow data for educational purposes for Companies A and B. Amounts are in KRW 100 million, and we assume that only regular-session trades on the same day are included.
① Company A's Trading by Investor Type
| Investor | Purchases | Sales | Net purchases |
|---|---|---|---|
| Foreign Investors | 320 | 170 | +150 |
| Institutional Investors | 180 | 230 | -50 |
| Retail Investors | 200 | 300 | -100 |
| Total | 700 | 700 | 0 |
Foreign net purchases are as follows.
\[ 320-170=KRW\ 15\ billion \]
Institutional and retail net sales sum to the following.
\[ -50+(-100)=-KRW\ 15\ billion \]
Foreign investors' KRW 15 billion in net purchases absorbed the KRW 15 billion in net sales from institutions and retail investors.
② Normalize by Trading Value and Market Capitalization
Assume that Company A's total trading value that day was KRW 50 billion and its market capitalization was KRW 1 trillion.
\[ Foreign\ net\ purchase\ to\ trading\ value = \frac{150}{500}\times100 =30\% \]
Because KRW 1 trillion equals KRW 1,000 billion, the calculation is as follows.
\[ Foreign\ net\ purchase\ to\ market\ capitalization = \frac{150}{10,000}\times100 =1.5\% \]
③ The Same KRW 15 Billion Has a Different Meaning for Company B
Assume Company B also had KRW 15 billion of foreign net purchases but KRW 500 billion of total trading value and a KRW 10 trillion market capitalization.
\[ Net\ purchase\ to\ trading\ value = \frac{150}{5,000}\times100 =3\% \]
\[ Net\ purchase\ to\ market\ capitalization = \frac{150}{100,000}\times100 =0.15\% \]
The absolute foreign net purchase amount is the same for Companies A and B, but its relative size is ten times larger for Company A. All else equal, flow pressure could feel stronger in Company A.
④ Do Not Add Program Trading Again
Assume that KRW 12 billion of Company A's KRW 32 billion in foreign purchases was executed through program orders. That KRW 12 billion is already included in the KRW 32 billion of foreign purchases.
Therefore, do not add KRW 15 billion of foreign net purchases and KRW 12 billion of program purchases and conclude that “KRW 27 billion came from separate buyers.” Program data should be treated as a clue about order method that overlaps with investor classification.
5. Practical use in equity and company analysis
① Identify the Event Behind the Price Move Before Examining Flows
First look for an event capable of moving the price.
- Earnings releases and guidance changes
- Supply contracts, orders, new products and regulatory changes
- Rights offerings, convertible bonds, treasury shares and major-shareholder trades
- Index additions, deletions and periodic rebalancing
- Changes in rates, exchange rates, commodities and industry data
Starting with the flow figure without an event review makes it easy to mistake mechanical activity for information-driven trading.
② Read Flows Together With Price, Volume and the Closing Location
Price response matters even when net purchases are the same.
| Flow and price combination | First question to ask |
|---|---|
| Large net purchases + strong gain + close near the high | Was there aggressive buying and an information-driven revaluation? |
| Large net purchases + flat or falling price | Are buyers absorbing even larger supply, or is distribution occurring? |
| Net sales + rising price | Was there more aggressive buying from another group or short covering? |
| Flows concentrated in the closing auction | Was this a mechanical index or ETF rebalance order? |
If price does not rise despite large net purchases, sell-side supply may have been heavy or the information may already have been priced in. Conversely, if price rises despite foreign net selling, another group such as retail or institutions may have accepted higher prices more aggressively.
③ Record One-Day, Five-Day and 20-Day Data in the Same Units
A table like the following can be useful for each stock.
| Period | Foreign net purchases | Institutional net purchases | Retail net purchases | Total trading value | Net purchases / trading value |
|---|---|---|---|---|---|
| Day 1 | |||||
| 5-day cumulative | |||||
| 20-day cumulative |
Do not mix amount and share-count data; use the same market, period and units. When prices changed substantially over the period, share-count and amount measures may create different impressions, so you can review both.
④ Ask Questions to Separate Mechanical Flows From Judgment-Driven Flows
On the following dates, flow may be driven by something other than a corporate-value judgment.
- Major index review, addition and deletion dates
- Futures and options expiration dates
- Large ETF creation and redemption dates
- Block trades and after-hours large trades
- Listing dates for rights-offering shares
- Lockup expiration dates
The categories cannot be separated perfectly, but checking schedules and disclosures reduces misinterpretation.
⑤ Ask Whether Flows Confirm the Fundamental Thesis
For example, if a company with rising earnings estimates sees 20 days of cumulative foreign and institutional net buying, increasing volume on up days and limited selling during pullbacks, you may investigate whether a market revaluation is under way.
Be cautious, however, in the following situations.
- Earnings estimates are falling while flows are only temporarily positive
- Large net purchases on the index-inclusion date immediately reverse to net selling
- Net buying persists, but price continues to form lower highs and lower lows
- Chasing net purchases surge at an expensive valuation
- Volume and price collapse as the dominant flow category changes
Flow data is not an input for calculating fair value. It is supporting evidence for examining the price-formation process and whether a catalyst can persist.
⑥ Take Care When Backtesting the Rule “The Stock Rises When Foreigners Buy”
Avoid the following errors.
- Assuming you bought at the same close at which end-of-day flow data became final
- Survivorship bias from using only a later index membership list
- Ignoring execution feasibility and market impact in low-turnover stocks
- Treating index changes, expirations and block trades as ordinary buying
- Selecting only the best-performing accumulation window after testing many windows
- Interpreting only cash equities while excluding short sales, securities lending and derivatives positions
Respect when flow data became public and the next point at which an order could actually have been placed.
6. What to look for in filings
Investor-category net purchases and program trading are not financial-statement items. Flows come from exchange market data, while corporate actions and ownership changes that may explain those flows come from Filing for these items.
Korean Market: KRX, DART and KIND
KRX Data Marketplace, Financial Supervisory Service DART, Korea Exchange KIND to distinguish the following.
| Location | What to check |
|---|---|
| KRX trading activity by investor type | Purchase, sale and net purchase amounts and share counts for foreign, institutional, retail and other investors |
| KRX program trading activity | Purchases, sales and net purchases from arbitrage and non-arbitrage programs |
| KRX short-selling and securities-lending statistics | Supporting clues that net buying may relate to short covering |
| DART and KIND ownership filings | Changes in holdings of controlling shareholders, executives and major shareholders |
| Treasury-share acquisitions and disposals | Plans and execution for company repurchases or supply of held shares |
| Rights offerings, CBs, BWs and exchangeable bonds | Potential share supply, dilution and hedging activity |
| Block-trade and after-hours large-trade disclosures | Whether a one-off transfer of a large position occurred |
| Mergers, demergers, stock swaps and tender offers | Basis for index, arbitrage and event-driven flows |
When downloading KRX data, record the market, security, period, share-count or amount basis, detailed investor category and whether the scope includes regular-session or after-hours trading. Adding both the displayed institutional total and its subcategories can double count the same activity.
U.S. Market: SEC EDGAR and Market Data
SEC EDGAR for the following materials.
- Form 13F: Quarter-end holdings of institutional investment managers meeting specified requirements. It is not daily flow data, and the filing delay after quarter-end means the position may have changed.
- Schedule 13D·13G: Beneficial ownership above the applicable level and changes in holdings or purpose.
- Form 4: Transactions by executives, directors and major insiders.
- 10-K·10-Q: Shares outstanding, treasury shares, stock-based compensation, dilutive securities and capital structure.
- 8-K·S-3·424B: Buyback plans, securities issuance, acquisitions, divestitures and other material corporate events.
Form 13F is a holdings snapshot and must not be used as proof that an institution was a net buyer on a specific day. Daily investor flows in the United States are not aggregated and disclosed under the same categories and methodology used in Korea, so always verify the data provider's definitions.
Market data needed outside the filings
| Data | Basis to Check |
|---|---|
| Purchases, sales and net purchases by investor type | Amount or share count, market, period and classification rules |
| Program arbitrage and non-arbitrage activity | Overlap with investor classifications and concentration at the close |
| Daily OHLC, volume and trading value | Regular or after-hours session, adjusted prices and currency |
| Market capitalization and free float | As-of date and treatment of treasury and locked-up shares |
| Index inclusion and exclusion schedules | Announcement date, effective date and closing-price methodology |
| ETF creations, redemptions and holdings | Reference time and lag before actual trading |
| Futures, options and basis | Potential hedging and arbitrage activity |
| Short Selling, Securities Lending and Margin Trading | Opposing positions and potential short covering |
| Exchange rate, market and sector returns | Separating foreign allocation changes from common factors |
An analysis record should state the as-of date, market, period, unit and classification together, for example: “October 10, 2026; KRX regular session; amount basis; five-day cumulative foreign net purchases; ratio to total trading value.”
7. Key takeaways
- Net purchases are an after-the-fact total calculated as purchases minus sales, and every execution has a buyer and a seller at the same time.
- Foreign, institutional and retail are investor categories, while program trading describes an order method; the classifications can overlap.
- Foreign and institutional categories contain passive, hedging, long-term and short-term strategies with different objectives.
- The same net purchase amount can have greater relative impact in a stock with lower trading value and market capitalization.
- Price impact depends not only on net purchase amount but also on order-book depth and order aggressiveness.
- Review five-day and 20-day accumulation, persistence, sector breadth and closing-auction concentration rather than one day alone.
- Index changes, ETF creations and redemptions, derivatives hedging, block trades and short covering can resemble fundamental buying.
- Co-movement between flows and price does not automatically prove causation.
- Flow data does not calculate fair value; it supports analysis of the price-formation process and whether a catalyst can persist.
- In practice, review disclosures, earnings estimates, volume, valuation, exchange rates and index and derivatives schedules together.
8. Three review questions
Question 1
In one stock, foreign investors bought KRW 48 billion and sold KRW 31 billion, while institutions bought KRW 22 billion and sold KRW 30 billion. Assume retail investors traded the remaining opposite amount and there are no other categories. Calculate foreign, institutional and retail net purchases.
Question 2
Companies C and D each have KRW 20 billion of foreign net purchases. Company C's daily trading value is KRW 80 billion and its market capitalization is KRW 2 trillion. Company D's daily trading value is KRW 800 billion and its market capitalization is KRW 20 trillion. Calculate each company's net-purchase-to-trading-value and net-purchase-to-market-capitalization ratios, and compare relative size.
Question 3
At the closing auction on an index-inclusion effective date, foreign and program net purchases in one stock surged and the stock rose 7%. Explain why this cannot automatically be treated as long-term fundamental buying and what to examine beginning with the next trading day.
Answers and explanations
Answer 1
Foreign net purchases are as follows.
\[ 48-31=+KRW\ 17\ billion \]
Institutional net purchases are as follows.
\[ 22-30=-KRW\ 8\ billion \]
Foreign and institutional net purchases sum to (+KRW 9 billion), so retail net purchases must be (-KRW 9 billion) to make the market-wide total equal zero.
\[ +170-80-90=0 \]
Answer 2
Company C's net-purchase-to-trading-value ratio is as follows.
\[ \frac{200}{800}\times100=25\% \]
Because Company C's KRW 2 trillion market capitalization equals KRW 2,000 billion, the calculation is as follows.
\[ \frac{200}{20,000}\times100=1\% \]
Company D's net-purchase-to-trading-value ratio is as follows.
\[ \frac{200}{8,000}\times100=2.5\% \]
Because Company D's KRW 20 trillion market capitalization equals KRW 20,000 billion, the calculation is as follows.
\[ \frac{200}{200,000}\times100=0.1\% \]
The absolute amount is the same, but both ratios are ten times larger for Company C. All else equal, this may represent stronger relative flow pressure in Company C. Actual price impact still depends on order-book depth and timing.
Answer 3
On an index-inclusion effective date, passive funds and ETFs tracking the index may buy mechanically at the designated close to match required weights. If foreign accounts used program orders, the two statistics can also display the same orders under overlapping classifications. Therefore, the net buying and 7% gain do not prove that investors confirmed a long-term corporate-value thesis.
Beginning with the next trading day, examine the following.
- Whether foreign and institutional net buying persists over five and 20 days
- Whether non-program demand remains after excluding program orders
- Whether price holds above the inclusion-date close after volume normalizes
- Whether earnings estimates and guidance are being raised at the same time
- Whether PER, FCF yield and DCF remain reasonable at the new price
- Whether short covering, derivatives hedging or ETF creations and redemptions affected the flow
Next Lesson Preview — Day 21:
“Short Selling Is Not Only a Bearish Bet — Reading Securities Lending, Short-Sale Trading Value and Short Covering”