Here is a strange fact about one of the most-quoted numbers in Israeli finance: the TA-35 index level scrolls across every business broadcast in the country, but if you want to know what the index is actually made of — which companies, at what weights — there is no free, official, machine-readable answer. The Tel Aviv Stock Exchange publishes the constituent list and the index rules, and licenses the weights through commercial data products. For a research site trying to answer basic questions about Israeli public-market exposure, that gap was untenable. So we did it the slow way: we reconstructed the weights ourselves, from company market capitalizations, free-float disclosures, and the exchange's published capping rules, and we ship the result as an inspectable dataset in our TASE Index Explorer.
What the reconstruction shows
The headline finding is concentration. Five companies sit pinned at the TA-35's 7 percent single-constituent cap simultaneously: Teva Pharmaceutical Industries, Elbit Systems, Bank Leumi, Bank Hapoalim, and Tower Semiconductor. That is 35 percent of the index in five names — the mathematical maximum the cap allows — and it means the cap is not a theoretical safeguard but an actively binding constraint reshaping the index every rebalance. Extend to the top ten (adding Azrieli Group and Mizrahi Tefahot at roughly 4.8 percent each, Nova at 4.5, Phoenix Financial at 4.0, and Israel Discount Bank at 3.4) and you are holding 56.4 percent of the flagship index in ten companies.
The sector picture is more uncomfortable for anyone who reaches for the TA-35 as an "Israeli innovation" proxy. Finance carries 37.3 percent of the index — four of the country's five large banks plus insurers. Technology, in the broadest construction we can defend, carries 23.9 percent, and a substantial slice of that is Tower Semiconductor's capped 7 percent. Energy and real estate together add roughly 25 percent. Healthcare is effectively one position: Teva, at the cap. The index tracking the self-described startup nation allocates more weight to real estate than to all of its healthcare, communications, and consumer names combined.
Two structural observations follow. First, nine of the thirty-five constituents are dual-listed, mostly in New York, so a meaningful share of "Tel Aviv" price formation actually happens on U.S. exchanges in dollars. Second, the technology story Israeli private markets are famous for — cybersecurity, defense autonomy, AI infrastructure — is barely investable through the index at all: those companies either listed on Nasdaq directly, were acquired before listing anywhere, or remain private. The TA-35 is a fine banking-and-industrials index. It is a poor window into the part of the Israeli economy the world is most curious about.
How we curated the weights — and what broke
Reconstruction sounds mechanical until you audit it. Our July 2026 refresh, which produced the current dataset version, removed three companies that had crept into earlier drafts through exactly the kind of error this project exists to catch. Check Point Software — reflexively assumed by nearly everyone to anchor the Israeli index — has never been TASE-listed; it trades only on Nasdaq. Magic Software and Sapiens had both delisted from Tel Aviv and needed to come out. Five smaller constituents were added in their place, and every weight was re-derived from current market data. The build now enforces the exchange's capping rules in code: if any normalized weight exceeds 7 percent in the TA-35 or 5 percent in the TA-125, the site fails to deploy until the seed data is corrected. We would rather break our own build than publish a weight the index rules make impossible.
We are explicit about what the result is: cap-constrained approximations derived from public market-capitalization data, not licensed official weights. They will drift from the official figures between rebalances, and the dataset says so on its face, with an as-of date and a warning that it is not a live quote service. The claim we stand behind is narrower and more useful — the structure of the index (who is capped, how concentrated the top is, which sectors dominate) is robust to the small errors this method introduces, and that structure is what almost every research question actually needs.
What to do with this
For a reader weighing Israeli public-market exposure, the reconstruction reframes the question. Buying the TA-35 is a decision to hold concentrated Israeli financials with a technology sleeve, not a decision to hold the Israeli technology sector — that distinction should drive the choice among index exposure, dual-listed single names, and U.S.-listed Israeli companies, a comparison our exposure-routes guide walks through. For researchers, the meta-lesson is about data supply chains: some of the most-cited numbers in a market can rest on the least-inspectable foundations, and the fix is not to trust harder but to reconstruct and publish the working. Every constituent, weight, membership, and theme classification behind this briefing is browsable in the explorer, and the methodology explains how to challenge anything you find there. As always: educational research, not investment advice, and certainly not a suggestion about what the index will do next.