On 17 November the CBR published the November issue of its enterprise monitoring commentary. It has run this monthly survey since 1998; 13.0 thousand non-financial enterprises answered this round between 1 and 11 November. Three numbers belong together. The balance of firms reporting rising costs averaged 44.5 points in the third quarter — the highest figure anywhere in the issue's table, which reaches back to 2018. Capacity utilisation fell to 76.3%, the lowest value in that table since 2020. The balance of firms' assessments of credit conditions was −12.6 in November — the third month in a row worse than the one before. The CBR says the tightening came through non-price terms: requirements on borrowers' financial position and collateral. We read the November issue in full this session, together with the October issue, and recomputed every comparison.

Costs: the fastest quarter in the printed series

The cost row of the issue's key-indicator table gives annual averages from 2018 and quarterly averages from the start of 2025. The third quarter of 2026 reads 44.5. Every other value in the row is lower: 39.0 for 2022, the war's first year; 35.5 for 2025; 39.6 for the first quarter of this year; 17.4 for 2018. The balance is the share of firms reporting higher costs minus the share reporting lower ones, and the issue's own methodological note warns how to read it: the series has sat in positive territory for its whole history, so it measures the speed of cost growth, not a level — a record balance means costs rising more widely and faster than in any printed period, not a record cost.

The monthly readings show what happened next. November's balance is 41.4, below the quarterly record, and the commentary says plainly that cost growth slowed for a fourth consecutive month. Firms cite higher prices for raw materials and components, logistics, fuel and lubricants, and utilities. The quarter that just closed was therefore the fastest in the printed series. The months since have come off that peak without returning to anything like the pre-2021 pace: the lowest reading of the autumn, November's 41.4, is still higher than any pre-war annual average in the table.

Credit: tighter on terms, not on price

The credit-conditions row is the one moving against the rate cycle. In the November issue it reads −10.5 for September, −12.1 for October, −12.6 for November: each month worse than the one before. The October issue prints the August reading the November table drops — −7.0 — so the deterioration now runs three months without interruption from the August base. The commentary's own explanation:

“Some tightening of credit conditions was connected with non-price parameters (requirements on financial position, on collateral and so on). The overall balance of enterprises' assessments of credit conditions stood at the level of the average values of the last 10 years.”

Both halves of that quotation matter. The first names the mechanism: firms report that banks are demanding more of borrowers' financial position and collateral. Those requirements bear first on the firms in the weakest position. The second half is the counterweight, and we keep it: the current three-month average, −11.7, sits almost exactly on the decade mean in the issue's chart, −11.6 points for 2016–2025. The autumn's tightening is a direction, not an extreme.

Two details put the reading in context. A footnote records that almost 32% of the surveyed firms had applied for a loan in the preceding month — roughly a third of the sample was seeking credit. In the quarterly block on investment, the cost of borrowed money ranks only fourth among the constraints firms name — behind a shortage of own funds (26.7% of respondents), economic uncertainty (25.8%) and insufficient demand (20.8%) — and the CBR notes that its restraining influence is declining. By the firms' own account, access to credit and a shortage of their own funds constrain them more than the interest rate.

Capacity: the lowest use since 2020

The quarterly questionnaire — answered between 1 and 8 October — puts capacity utilisation in the third quarter at 76.3%, down from 77.3% in the second. The commentary states the bounds itself:

“Capacity utilisation in Q3 2026 fell to 76.3% (Q2 2026: 77.3%) and remained above the values of 2017–2019. Utilisation declined in all industries, with the exception of manufacturing producing intermediate goods.”

In the printed row, every value from 2021 onwards is higher than 76.3%, and only the pre-war years sit below it: 75.2% in 2018, 75.8% in 2019, 74.9% in 2020. After five years in which the table shows utilisation at 78–81%, the rate is now just half a point above its 2019 value — while the cost balance stands at its printed record.

Prices point the other way

One part of the survey moved in the opposite direction, and it is the part closest to our own data. Firms' price expectations fell: the balance of expected price changes dropped to 18.6 points, and the average expected price growth for the next three months, annualised, fell to 4.5% from 6.0% in October — below all three quarterly averages this year: 6.1%, 4.6% and 6.0%. The declines reached nearly every industry except agriculture, with expected growth ranging from 2.4% in mining to 7.6% in retail trade.

In September we read the September round, which reported expected price growth of 7.0% — the highest figure in its row. The November issue prints September at 6.4%: the survey's note says previous periods are recalculated as late questionnaires arrive and seasonal adjustment is re-run. The revision has partly erased the September spike. The direction since is one way: 6.4%, 6.0%, 4.5%.

This is not our inflation indicator and must not be read as one. Ours is Rosstat's annual consumer-price inflation; the 4.5% is surveyed enterprises' expectation for their own selling prices over three months, annualised — a different series from a different publisher. What the survey supports is a direction: the price side of the questionnaire is easing while the cost side stands near its record.

Which month the readings describe

The survey's standing footnote applies: firms assess current conditions for the previous month, while expectations cover three months beginning with the survey month. November's current readings — the −12.6 on credit, the 41.4 on costs — therefore describe October; the quarterly figures for capacity and investment were collected in early October and describe the third quarter. The headline business-climate indicator, 1.1 points against 1.0 in October, holds above zero only through expectations (9.0 points); current assessments are negative at −6.4 and worsened from October's −5.9.

Why a survey of firms is a receivables question

This monitoring is the only regular measure of the corporate sector that does not pass through Rosstat, and this round describes a channel through which firms can start paying each other later. Costs grew in the third quarter at the fastest printed pace since at least 2018; a third of the sample sought credit in the preceding month; and firms report tighter requirements on their financial position. A firm that fails a bank's balance-sheet test still owes its suppliers — but it may settle later. Our overdue receivables series is where that behaviour would eventually surface, at one remove. The survey measures what firms say about conditions, not what they owe; we treat it as context for that series, not as a reading of it.

What this changes in our model

Nothing today — survey balances are not among our eight components, and no indicator moved in this morning's collection. Overdue receivables scores 4 of 100, the worst of the eight. Annual inflation stands at 6.3% year on year and scores 95 of 100, the strongest. The strength index stands at 47.6 of 100, in the stress zone; all eight series are on the monitor.

What we do not know

The balances are qualitative. The issue contains no rouble figure anywhere — no cost level, no loan volume, no price — so nothing in it can be converted into money, and we have not tried. The respondent count, 13.0 thousand against an average of 15.8 thousand in 2025, is the count at publication; the survey's own note says questionnaires keep arriving after the operational cut-off, so the gap supports no conclusion about firms declining to answer. The credit-conditions series is published without seasonal adjustment and is revised between issues — the October issue printed September at −10.4 and October at −12.7, the November issue prints −10.5 and −12.1 — so the three-month run of deterioration is the current table's account, exact to the current vintage. And the record cost balance is a record of the printed series, which begins in 2018: the issue does not publish the monthly history behind the annual averages, so we cannot test any single month before 2025 against the 44.5.