Foreign Flow Indicators: How Smart Money Predicts Asian Markets (And When It Doesn't)

Every major Asian exchange publishes a daily breakdown of turnover by investor type — typically local institutional, local retail, and foreign institutional. The foreign bucket is watched closely because it proxies the movement of cross-border money in and out of the market. A persistent outflow is often read as a bearish signal; a sustained inflow is read as institutional accumulation.

Both readings contain signal. Both also break in predictable ways. This guide covers the mechanics and the failure modes.

What 'foreign net' actually means

Each exchange defines 'foreign' differently. Some examples from the six Phase-1 markets we track:

• TWSE (Taiwan) combines foreign institutional investors with mainland Chinese investors and explicitly excludes foreign dealers — a tiny bucket that usually trades at 0 net anyway.
• NSE (India) separates Foreign Portfolio Investors (FPIs) from Domestic Institutional Investors. The reported figure is the FPI cash-market net for the day.
• Bursa Malaysia reports 'Foreign Institution' as a single aggregate against Local Institution and Retail.

In every case, the figure is cash-equity-market net. Block trades and derivatives positions are not included. This matters because large players often express views via futures or OTC swaps, which don't show up in the daily foreign-net number.

When it leads price

The most reliable pattern: sustained inflows/outflows — weeks of one-sided prints — often precede directional moves by 2–4 weeks, particularly in Taiwan and India where foreign ownership is a large share of free float. The mechanism is straightforward: when institutional allocators decide a country is cheap (or expensive), they don't enter or exit in one session. The daily prints pick up the footprint.

When it misleads

Three documented failure modes:

• Benchmark rebalances. MSCI and FTSE index changes drive mechanical foreign buying or selling unrelated to fundamentals. The headline print looks bullish, then reverses the next month as the rebalance completes.
• Currency-driven selling. A weakening local currency forces USD-benchmarked foreign funds to reduce positions even when the underlying market is attractive in local-currency terms. Foreign selling in IDR terms can coincide with IDX rallying in local currency.
• Sector concentration. In markets dominated by one or two names (TSMC for Taiwan, Reliance and TCS for India), 'foreign flow' is really a vote on those names. Extrapolating to the broader market is risky.

What we'll do with the data

Once we have 6–8 weeks of accumulated daily history across all six tracked markets, two derived metrics activate:

• Smart Money Score (per stock, Taiwan only to start) — 0–100 composite weighting consistency of foreign buying, volume share, trend duration, and divergence from retail.
• Divergence alerts — flag tickers where retail flows are buying while foreign institutions are selling (or vice versa). Historically these divergences have a measurable hit rate, which we'll publish honestly once our own dataset is long enough to measure it.

Until then, the site shows only raw daily aggregates with a Beta badge. That's the honest presentation — a signal is only a signal after enough observations to compute a prior.