How is it possible that the Israeli economy has been at war for almost three years, many businesses are still struggling to return to full activity, private consumption is recovering slowly, only about one-tenth of employees work in high-tech — and yet growth keeps surprising on the upside, state tax revenues repeatedly beat forecasts, and the deficit has been falling faster than expected for eight consecutive months?
The answer may be one company: Nvidia, the world’s largest company, which has significant operations in Israel. Last week, Alex Zabezhinsky, chief economist at Meitav, published an unusual review in which he argued that Israel’s growth figures are “less representative of the state of the economy,” because they are significantly affected by the activity of one company, or a very small number of companies. Two days ago, Leader Capital Markets chief economist Yonatan Katz opened his weekly review with a similar conclusion.
Nvidia CEO Jensen Huang
(Photo: AFP)
According to Katz, the activity of Mellanox, Nvidia’s Israeli subsidiary, explains a large part of the surprise in tax collection, is putting pressure on the shekel to appreciate and is even allowing the government to raise less debt than planned. When two of the leading macroeconomists point within one week to the same phenomenon, it is no longer an anecdote, but a macroeconomic perspective that deserves attention.
The figures are indeed exceptional. According to Katz, exports of goods that do not cross Israel’s borders — activity associated mainly with Mellanox — reached nearly $10 billion in 2025, and in the first quarter of 2026 were already approaching an annual pace of about $20 billion. He also estimates that this activity alone contributed at least $1.5 billion to state tax revenues and helped raise the revenue forecast by about NIS 7 billion.
Zabezhinsky also presents dramatic figures: According to his calculations, without this exceptional activity, Israel’s growth in 2025 would have amounted to only 1.6%, instead of 2.9%, and in the first quarter of 2026 the economy would have contracted by 10.5%, instead of 3.8%. In other words, a large share of the economy’s impressive performance over the past year stems from the global activity of one company, or at most a handful of companies.
Nvidia does not manufacture its chips in Israel. Production takes place outside the country. However, Mellanox’s development centers, management and intellectual property are registered here, so the added value of the activity is recorded as Israeli exports, even if the goods themselves never cross the country’s borders.
From there, an entire economic chain begins: Exports rise, GDP grows, company profits increase, the state collects more corporate tax, the deficit shrinks, the need to raise debt declines and pressure for the shekel to strengthen increases. It is important to clarify: This is not statistical manipulation. This is the correct measurement under the International Monetary Fund’s rules for national accounting, and all countries in the world operate this way. But it raises a different question: Do these figures still describe the real state of the Israeli economy?
Ultimately, the important macroeconomic question is simple: What is the problem with this? The answer is straightforward, and economic literature provides a clear, though uncomfortable, response.
The first problem is that this dynamic creates a “macro illusion.” When the government sees high growth, tax revenues that positively surprise and a deficit falling faster than expected, it is easy to conclude that the economy as a whole is in “excellent” shape, as Finance Minister Bezalel Smotrich often explains, and boasts. But in practice, most sectors of the economy may be recovering at a much slower pace. After all, only about 10% of employees work in high-tech, and an even smaller share are employed by Nvidia or directly benefit from its success. If one company pulls the economy’s average upward, that average may conceal the condition of the majority.
The second danger is fiscal risk. State tax revenues jumped sharply, and they also explain a large part of the decline in the deficit. But if part of that surge is tied to an exceptional business cycle in the artificial intelligence sector, or to one company, it must not be assumed that it will continue forever.
If the government translates exceptional revenues into permanent spending, it may discover in a few years that the revenues have disappeared — but the commitments remain. This is not a theoretical concern: It is exactly what happened last week, when Smotrich boasted that he had lowered the deficit even as he continues to increase spending and debt, while the deficit has stood at more than 5% for three consecutive years.
The third danger is macroeconomic concentration. Economic history is very familiar with such situations. Finland benefited for years from Nokia’s enormous success. At its peak, the company accounted for a significant share of the country’s exports, growth and research and development investment. When Nokia collapsed at the beginning of the previous decade, the Finnish economy also entered a long period of weakness. In South Korea, Samsung still accounts for a huge share of exports, and the government has been working for years to reduce dependence on it. In Taiwan, a similar debate is taking place around TSMC. In all these cases, the success of a giant company is a tremendous asset, but also a source of risk.
The fourth danger is the illusion of measurement. When one company is so large, averages stop being representative. This is exactly the phenomenon observed in the U.S. stock market over the past two years. Stock indexes broke records mainly thanks to the “Magnificent Seven,” while many other companies showed much more moderate performance.
In Israel, too, macroeconomic data may be painting a better picture than the one experienced by small businesses, manufacturers, retailers or service providers. There is another aspect as well: The more an economy becomes dependent on a small number of global companies, the more exposed it becomes to factors it cannot control — decisions made by company management far from Israel, international tax changes, trade wars, technological shifts or simply a slowdown in the global artificial intelligence cycle. Any such event could affect not only Nvidia’s stock, but also Israel’s growth figures, tax collection and deficit.
This is not criticism of Nvidia. On the contrary, Israel needs more companies like Nvidia, not fewer. Its success is a tremendous achievement for Israeli high-tech and the Israeli economy. The problem begins when one company becomes so significant that it changes the way we interpret the country’s macroeconomic data.
And that may be the most important conclusion from this new phenomenon, as reflected in the reviews published a week apart. The question is no longer how successful Nvidia is. That is clear to everyone. The question is whether the numbers we see still tell the story of the entire Israeli economy, or whether they are beginning to tell mainly the story of one extraordinary company.