No OFZ auction has been held since July: Minfin's results page still lists 15 July as the latest auction, with no entry after the 20th. With the primary debt market shut, the banking sector's July balance sheet shows what the main potential buyer of future issues was doing.
Banks were cutting their OFZ holdings, not building them
In the CBR's condensed analytical balance sheet at 1 August, banks' OFZ holdings stand at 19,189.3 bn roubles, down from 19,357.2 bn a month earlier — a fall of 167.9 bn over July. They remain above the January level of 18,852.9 bn at the start of the year.
This was not a revaluation effect; the file makes that visible. The bonds line before revaluation also fell, from 30,793.2 to 30,766.4 bn, while the revaluation itself became less negative over the month: −507.0 to −499.4. Market prices moved in the banks' favour, yet their holdings still shrank.
Corporate bonds went the other way over the same month, from 6,658.4 to 6,909.1 bn.
State money in banks: down 1.66 trillion since January
The liabilities side carries a line for “state funds” — budget money placed with banks. It stood at 10,148.8 bn on 1 January and 8,491.2 bn on 1 August. Over seven months, that is −1,657.7 bn. In July itself, the line edged up by 49.9 bn.
This is a liquidity buffer separate from the NWF, and our model does not measure it. It also proves nothing by itself. The CBR notes that within July money moved out of the Treasury (−0.2 trn) and into regional budgets (+0.2 trn) — redistribution, not spending. The seven-month fall is consistent with a deficit covered out of balances while the debt market is shut. It does not establish it.
Where July's profit came from
The sector reported net profit of 442.5 bn roubles for July on its balance sheet, rounded to 443 in the text of the CBR's review. The profit did not come from operations: provisioning costs on problem assets fell from 282 to 168 bn, down 40%, and by 103 bn in the corporate segment, down 77%. The CBR puts it this way:
“Individual banks released provisions, among other reasons after some borrowers updated their business plans to reflect rising revenue.”
What our model reads from this
Directly, nothing. We retired OFZ placement from the model on 22 August, and we have never measured state deposits in banks. Indirectly, the story touches the deficit, which scores 65 out of 100 and remains the healthiest of the eight, and the liquid NWF, at 28 out of 100.
The mechanism matters in the scenario where borrowing resumes. The buffer a bank could use to absorb new issues got smaller, not larger: sector balance-sheet capital fell over July from 22,330.8 to 21,826.5 bn. The only line pushing it up was that same profit, made by releasing provisions rather than by earning, and the 0.9 trn of accrued dividends took out more than twice as much. The rest of the indicators are on the monitor.
What we do not know
The file does not show how much of the 167.9 bn is redemptions and how much is sales: there is no split. Nor does it distinguish between systemically important banks and the rest. It is not clear whether the “state funds” line includes regional budget money, which would change how the figure should be read. August data will not appear until late September.