An index in the older sense of the word: a catalogue of record. 385 records covering 310 Iranian technology companies across 23 sectors, each resolved to a canonical entity and built from public sources. This page is the standard it is held to.
A catalogue of record, not a benchmark: no index level is published, because a level has to be reproducible from a published formula and we do not have one yet. Coverage standard →
Until 30 August 2026 this page published a benchmark level of 100.00, a base, an inception date and a promise that a performance series would accrue from day one. There was no formula behind it. Nothing could move that number, and nothing was scheduled to. We have removed it.
A benchmark level has to be reproducible: someone outside this desk must be able to take the published inputs, apply the published rule, and arrive at the same number we did. That requires a defined formula, weights, a treatment for missing data, a rebalancing rule and a comparable time series across constituents. We do not have those yet, and publishing a level without them is the kind of claim this platform exists to be the opposite of.
We are building the inputs in the open: a daily free-market FX series, a two-store application index, and quotes on the listed constituents. When those support a level that genuinely moves, we will publish the formula first and the level second. Until then this is a registry, and it is described as one.
385 published records across 23 sectors, of which 310 are companies. The remainder are the objects a real registry has to hold in order to be usable: 55 products, 11 brands, 5 holding entities and 4 records for businesses that have ceased trading.
A record is published once the entity itself is resolved, meaning we can say with confidence which legal or commercial object it is, that it is not a duplicate, and which kind of object it is. Until 30 August 2026 this page said the registry contained no brands and no products. It did contain them, 65 of them, and the honest fix was to type every record and publish the breakdown rather than to quietly drop the sentence.
A further 30 entities are identified but deliberately not covered: banks and financial groups named as the parent of something we do cover. They carry a canonical name and, where obtainable, a national ID, and nothing else. They are never counted as records, because identifying an entity and covering it are different claims.
Entity resolution is the entry bar. Field completeness is not, and the two used to be conflated here: this page previously claimed that unverified records were excluded, while publishing a universe that includes them. Records enter thin and deepen. Coverage depth is disclosed on every record and in the directory legend rather than hidden behind a single count, because a reader deciding whether to rely on a company page needs to know which of these it is.
91 of those records are at the shallowest tier, carrying a resolved entity and little else. They are in the registry and counted in it. They are not represented as researched profiles, and no aggregate figure on this site is computed as though they were.
Constituents are grouped by sector, and the distribution above recomputes as the registry grows. Read it as a map of where coverage currently is, not as a weighting scheme: there are no weights, because there is no level for weights to feed.
Every figure carries a public source and a review date. Where a fact cannot be verified it is left out, never estimated in. Corrections are published with a timestamp rather than edited silently, and the live state of the dataset is exposed at /audit and in machine-readable form at /audit/manifest.json, computed from the same queries this page renders from.