Taiwan futures open interest — a stock of positions, not a direction
The standard misreading here is to read contract counts as a view. This page refuses that. First, open interest is a stock of outstanding contracts, not a measure of conviction or direction. Second, the institutional buckets are account classifications, not intent: hedges and speculative positions are booked under the same label, and public data has no field that separates them, so a sentence like "institutions are bullish" is not something this data can support and we do not write it. Third, this is market-level aggregate data and not a per-stock series. Fourth, contracts and expiry months are not additive, because their specifications and notional values differ.
Fields and granularity
Each row returns the data date, contract type, expiry month, category, long and short open interest, the net and the change versus the prior session. We keep the contract and expiry dimensions rather than pre-aggregating, because aggregation erases the differences that matter.
Non-trading days and unpublished dates return undetermined with the missing day named. We never roll yesterday's stock forward.
A stock, not a strength
Every open contract has a long and a short. Rising open interest means new positions exist and have not been closed; by itself it says nothing about which side is more confident. Translating a change in stock into a direction reads something the data does not contain.
The net figure is simply one count minus another, and it shifts with category definitions, contract choice and expiry mix, so we show the components instead of compressing them into a score.
Categories are not intent
Within one category you may find cash-market hedges, market-making offsets and outright directional positions. Public data carries no field marking the purpose of each contract, so statements like "institutions are bearish" are unsupported by an account classification and we do not publish them.
Related entities may also hold offsetting exposure in other contracts or in the cash market, so reading one contract in isolation is reading a single page of a ledger.
What this page must not be used for
No institutional view is inferred, no rankings, no timing, no price targets, no settlement forecasts, and no summing across contracts or expiry months, whose specifications and notional values differ.
This is market-level data and not a per-stock indicator. The right use is as a starting question: which contract and which expiry moved, is a roll period near, and did other contracts move the other way?