On 1 September, the CBR’s Research and Forecasting Department published its September bulletin, “What the trends say”. It gives quarterly figures for sectors hit by damaged capacity: in the second quarter, mining fell 1.4% q/q seasonally adjusted and intermediate-stage manufacturing 3.5%, because some refineries and export terminals were idle. In the same section, the department treats repairs to that infrastructure as a source of investment demand in the third quarter.
What the department counted
Section 2.1 is called “Supply shocks temporarily reduce GDP growth”. Its second point reads:
“One-off supply shocks continue to restrain economic growth. In the second quarter, conditions deteriorated in mining (-1.4% q/q SA, Figure 19) and intermediate-stage manufacturing (-3.5% q/q SA) because some refineries and export terminals were idle, while paid services and passenger transport stopped growing as access to fuel and security deteriorated in several regions.”
Two things in that sentence are absent from the CBR’s official releases. The first is the attribution: the numbers are tied to identified physical causes — idle refineries and terminals. The second is the word “security”, given as a reason why paid services and passenger transport stopped growing. The bulletin does not say what damaged the plants or what worsened security, here or anywhere else.
The summary on page 3 states the same boundary in one line:
“Except in the sectors where production and infrastructure were damaged, the economy continued to grow slightly. That growth remains tilted towards producers working on state orders…”
The wider macro picture: preliminary GDP growth for the first half is 0.6% y/y. Rosstat’s preliminary estimate for the second quarter was +1.3% y/y, which the department converts in a footnote to +1.6% q/q SA, after a fall of 1.7% q/q SA in the first quarter. Construction rose 5.0% q/q SA. The department expects the loss of capacity to weigh on growth again in the third quarter.
About a third of July’s price growth
The fuel channel is quantified in section 1.1:
“Higher fuel costs remained a significant driver of July’s sharp price growth: petroleum-product prices rose 5.9% m/m SA and, as in June, contributed 0.3 pp directly to overall price growth (about a third of the total).”
The “about a third” can be checked against the bulletin’s own table. Table 1 gives July price growth for all goods and services at 11.6% m/m SAAR, equivalent to about 0.92% for the month after seasonal adjustment; 0.3 divided by 0.92 is 33%. The same table makes the contrast sharper still. Prices for non-food goods rose 15.1% m/m SAAR in July. Excluding petroleum products, they rose 3.5%.
The department also records that the fuel market began to improve at the end of July — panic buying subsided, imports rose, refining capacity gradually recovered and regulatory measures took effect — but says the situation “in many regions remains tense” as of the end of August. The table marks the August inflation reading as an estimate based on weekly data, not an observed figure.
Repair is demand
The line worth reading twice closes the same section 2.1:
“At the same time, the increased need for the repair and protection of critical infrastructure may support the further expansion of investment demand in the third quarter.”
This is methodologically correct, and it is the reason the site exists. Rebuilding what was destroyed counts as investment; investment enters GDP; measured GDP growth then rises. The department is not concealing anything — it is applying national accounting exactly as written. But the resulting figure describes activity, not condition. Damage creates spending that would not otherwise have been needed, and the accounts record that spending as investment. A quarter that loses capacity and rebuilds it can therefore show more investment than a quarter in which nothing was damaged — while the economy ends up back where it started.
That is the same distinction the site draws between a level and a rate. We measure oil and gas revenue against 2021 rather than against last year for this reason: a rate recovers by definition once the base is low enough, while the level says whether the money is back.
Budget uncertainty, priced in basis points
Section 3.1 is titled “The key rate down, bond yields up”. After July’s cut of the key rate to 14%, the fall in OFZ yields was brief:
“From the beginning of August they turned upwards amid investor pessimism because of the aggravation of the geopolitical situation, tension on the fuel market and uncertainty regarding the parameters of the budget over the medium-term horizon.”
Medium- and long-dated OFZ yields rose 16–32 bp from their post-decision levels despite the pause in Minfin’s placements. Ten-year yields again exceeded 16% a year, and their spread over two-year paper widened to 159 bp. Yesterday we wrote that the CBR’s draft monetary guidelines were prepared without an agreed path for the deficit. That absence is the third of the department’s three reasons, and the market has put a price on it.
What this changes in our model
Nothing today. The bulletin contains no data point that our parser ingests: every figure is either from Rosstat, which we already read, or is the department’s own estimate, which is not a published series.
It does explain a seemingly contradictory pattern on the monitor. Inflation scores 96 out of 100 — the bulletin’s Table 1 gives the same 6.0% y/y for July as our series — while oil and gas revenue scores 14 and rail freight scores 33. Prices look almost normal against the pre-war anchor while the physical economy behind them does not. The department’s fuel arithmetic is one reason: a third of July’s price growth came through a supply failure, not an excess of money.
The bulletin itself comes with a caveat. Its title page states that the conclusions and recommendations it contains “may not coincide with the official position of the Bank of Russia”. This is a department speaking, not the board.
What we do not know
The bulletin never says how much capacity is idle — not the number of refineries, the volume of refining or which terminals. The phrase “production and infrastructure were damaged” carries no detail at all. There is no split showing how much of the −3.5% q/q SA in intermediate-stage manufacturing came from refining and how much from everything else. The August inflation reading is an estimate based on weekly data. The bulletin was prepared using data as of 28 August, so it contains nothing after that date. Nor does it once mention the digital rouble, which started for the public on 1 September, the day the bulletin appeared. Rosstat, which would provide physical refining volumes, has been unreachable from our container for seven days.