On 21 May 2026, in a Moscow hall belonging to the Moscow Exchange's Nikitsky Club, the chief economist of Russia's state development corporation VEB.RF stood up and said his country would not win a long war against the West. On 16 August, The Bell reported that Andrei Klepach had been dismissed over that speech — after twelve years in the job.
What he actually said
We read the transcript, not the summaries. It runs to 35 pages and was published by the Nikitsky Club as issue 145 of its series "Russia in a global context", under the title "The Russian economy and geopolitical challenges".
Three sentences are worth quoting exactly, because they are more precise — and more damaging — than the headlines about them:
"In this war of attrition we will not win the competition. We have an illusion that everything over there will collapse. It has not collapsed and it will not."
"We are losing both the technological and the economic competition in the world … in some respects we are losing it to Ukraine, however unpleasant that is for me too."
"In my opinion the economy will hold, but a social crisis may arise — precisely when nobody particularly expects it."
Note what he did not say. Klepach did not forecast an economic collapse; he explicitly rejected one — "economically we will not fall apart". His claim was narrower and harder to dismiss: that Russia loses the attrition contest anyway, because the other side is being financed faster than Russia can burn its own reserves. By his own arithmetic, Western aid to Ukraine amounts to roughly half of Russia's entire federal budget: "we finance the world, and the world finances Ukraine."
The same discussion produced a second remark that belongs on this site. Oleg Buklemishev of Moscow State University, speaking about the reserve fund that our own third chain link tracks, put it flatly: "Forget about the National Wealth Fund as an instrument of growth. It simply is not there any more, and will not be."
What this does to our numbers
Nothing. That is the point.
Not one figure on this monitor moved because of this. The runway, the burning link, the verdict — all of them come from the same primary series they came from yesterday, and a speech does not enter the model. We publish this because it is the rarest kind of corroboration available: the people inside the system, with access to the real numbers, describing the same strain our indicators measure from the outside — and being punished for saying it out loud.
What the firing does change is the supply of future evidence. Klepach's forecasts at VEB were, in the words of Carnegie's Alexandra Prokopenko quoted by The Bell, "based on assessments of reality rather than a desire to please someone, and often gloomier than the official figures." Removing that voice makes Russia's published economic picture slightly smoother — and slightly less true. This monitor exists because that smoothing is a permanent condition, not an accident: our data rules treat a series going dark as a signal in itself, never as good news.
Why trust this
- The dismissal is reported by The Bell, citing two acquaintances of Klepach; one of them says VEB's head Igor Shuvalov acted "on a call from above". VEB has been sent a request for comment. We are relaying that reporting, not independently confirming it — and we say so.
- The quotes above are not from that report. They are from the published transcript of the session itself, which we read in full.
- Everything else on this site is different in kind: it comes from primary statistical series, with the raw fragment behind every number. See how the model works.
About The Fuse
The Fuse tracks one question weekly: how long can Russia pay for its war? Seven links of the war-financing chain — oil money, borrowing, reserves, deficit, bankruptcies, prices, physical activity — each measured from primary sources against thresholds registered in advance. Occasionally, as here, we publish context that no indicator can capture. It never moves the numbers; it only tells you who else is looking at them. Forecasts of Russian collapse have historically fired too early, which is why we publish ranges rather than dates and keep our own month-by-month record on the track-record page.