A higher household inflation estimate made up 90% of a constructed Russian rate-gap fall

By Cumulant Research

The Bank of Russia announced on [24 July 2026](https://www.cbr.ru/eng/press/pr/?file=24072026_133000key_e.htm) that it would lower its key ratekey rateThe Bank of Russia's main policy interest rate, used to steer very short-term rates between banks and influence wider financial conditions. [Bank of Russia](https://www.cbr.ru/eng/dkp/w_infl/) from 14.25% to 14.00%, effective [27 July](https://www.cbr.ru/eng/). The quarter-point move was small beside a much larger change recorded in a household survey before the meeting.

InFOM's medianmedianThe median is the middle result after responses are arranged from lowest to highest. estimate of inflation over the following year rose from [12.4% in June](https://cbr.ru/Collection/Collection/File/62072/inFOM_26-06.pdf) to [14.7% in July](https://cbr.ru/Collection/Collection/File/62217/inFOM_26-07.pdf). The July fieldwork ran from 6 to 15 July, so it ended nine days before the policy announcement.

Pair June's expectation with the 14.25% rate effective at the end of that month and the constructed gap is 1.85 percentage points. Pair July's expectation with the 14% rate effective at the end of July and it is -0.70 point. The decline is therefore 2.55 points. Of that, 2.30 points come mechanically from the higher inflation estimate and 0.25 point from the lower key rate. The shares are about 90% and 10%.

The calculation is key rate minus expected inflation. It resembles a simple real-rate calculation, but it should be called a proxyproxyA proxy is a simplified measure used to approximate a more complicated concept without claiming to measure it completely. because its inputs do not cover the same time period. The key rate guides short-term financial-market rates, while the survey asks how much prices will rise over the next 12 months. The [Bank's explanation of monetary transmissionmonetary transmissionMonetary transmission is the gradual chain through which a policy decision reaches money-market rates, bank credit, spending and inflation. [Bank of Russia](https://www.cbr.ru/eng/dkp/voc/)](https://www.cbr.ru/eng/dkp/w_infl/) says the key rate moves overnight interbank rates almost immediately, while effects on wider credit conditions and inflation pass through gradually.

Think of the calculation as subtracting a one-year weather forecast from today's thermostat setting. The subtraction can be informative, but the two readings describe different horizons and should not be mistaken for the temperature a household or company actually experiences.

The July estimate is a grouped mediangrouped medianA grouped median is an estimated middle value calculated from response ranges rather than from an exact number supplied by every respondent. [InFOM](https://cbr.ru/Collection/Collection/File/62217/inFOM_26-07.pdf). Respondents selected ranges such as 9%-12% or 13%-16%, and InFOM estimated the middle of the resulting distribution. The [July survey](https://cbr.ru/Collection/Collection/File/62217/inFOM_26-07.pdf) covered at least 2,000 adults through face-to-face interviews in 100 settlements across 54 Russian regions.

There is no uniquely correct clock for this exercise. A month-end comparison is useful for a regular monthly series. A survey-to-survey comparison instead uses the rate respondents faced while interviews were under way. An event comparison freezes expectations and examines only the policy decision.

The month-end and survey-to-survey versions happen to produce the same 90% share because both contain a 0.25-point rate decline and a 2.30-point rise in expected inflation. Comparing the June survey directly with the new 14% rate includes two policy cuts, totaling 0.50 point, so the expectations share falls to about 82%. Examining only the 24 July decision holds expectations unchanged and assigns the entire 0.25-point movement to policy.

This sensitivity is not a defect hidden in the calculation. It is the main lesson: a decomposition can answer only the question built into its start and end dates.

The -0.70-point July snapshot is not a household deposit return, a mortgage rate, a company loan rate or a measure of any borrower's purchasing power. Bank rates also reflect a loan's maturity, the borrower's risk, banks' funding costs, lending standards and any government support. The Bank describes monetary transmission as a [multi-stage process](https://www.cbr.ru/eng/dkp/voc/), not a one-for-one reset of every financial contract.

That distinction is especially important on decision day. The Bank's [release calendar](https://www.cbr.ru/eng/calendar/) schedules June weighted-average rates for non-financial companies on 6 August and June household rates on 10 August. Both July series are scheduled for 8 September. As of 24 July, therefore, no published monthly loan-rate series could show what happened after the decision.

Even July's eventual averages will not isolate the cut. Most of July occurred before the new rate became effective on 27 July, and other financial conditions can change at the same time. Establishing a causal loan-rate effect would require more granular data and a design that separates the announcement from competing influences.

Governor Elvira Nabiullina attributed the July surge to developments in the fuel market. She said higher fuel prices had begun feeding into a broad range of prices and described petrol as unusually prominent in households' regular purchases and companies' costs. She also said expectations could retreat as fuel conditions stabilised. Those are the Bank's assessment and forecast, not conclusions established by the household survey itself. [Governor's statement](https://www.cbr.ru/eng/press/event/?id=32727)

The Bank's June research note provides a reason to treat the level cautiously. It says survey estimates of expected and observed inflation in Russia and abroad almost always exceed measured inflation because people pay more attention to rising prices than to stable or falling ones. The note nevertheless argues that changes in the survey and comparisons with its own history can contain useful information. [Bank research](https://www.cbr.ru/Collection/Collection/File/62101/Infl_exp_26-06_e.pdf)

The July estimate of 14.7% sits above the Bank's 8.4%-11.2% reference band, which contains 90% of household observations from 2017-2019. The Bank chose that period because measured inflation was stably low and close to 4%. The comparison identifies an elevated reading, but it does not prove which event caused it.

The overall median also conceals a large divide. July expectations were 13.0% among respondents with savings and 16.5% among those without savings, a difference of 3.5 percentage points. This is a descriptive subgroup comparison. It does not show that lacking savings causes higher expectations.

The 24 July decision was not simply an easing message. Alongside the immediate 0.25-point cut, the Bank raised its baseline forecast for average key rates in both 2026 and 2027. It also raised its 2026 year-end inflation forecast and lowered its 2026 GDPGDPGross domestic product is the value of final goods and services produced within a country during a specified period. [IMF](https://www.imf.org/en/about/glossary)-growth forecast. [July forecast](https://www.cbr.ru/Content/Document/File/194313/forecast_260724_e.pdf)

The midpoint of the projected 2027 average-rate range rose from 9.0% in April to 11.5% in July, even though the rate applying from 27 July fell to 14%. This is not contradictory. One figure is today's setting; the other is a forecast for the average across a future calendar year.

The Bank linked the higher path to elevated inflation expectationsinflation expectationsEstimates of how quickly prices will rise in the future, which can differ across households, businesses, investors and economists. [Bank of Russia](https://www.cbr.ru/Collection/Collection/File/62101/Infl_exp_26-06_e.pdf), possible second-round effects from supply disruptions and a more expansionary fiscal outlook. These are components of the Bank's baseline assessment, not independently identified causal estimates. [Decision](https://www.cbr.ru/eng/press/pr/?file=24072026_133000key_e.htm)

Calling the household subtraction the real rate without qualification would be misleading. A [Bank of Russia survey of 31 economists](https://www.cbr.ru/eng/statistics/ddkp/mo_br/) produced a positive 8.3-point real key-rate estimate for 2026, while the household snapshot is negative 0.7 point.

Both calculations are arithmetically valid, but they answer different questions. The economists' version subtracts a 6.2% December-to-December CPI forecast from a 14.5% annual-average key-rate forecast. The household version subtracts a 14.7% one-year survey estimate from the 14% rate on one date. Their opposite signs mainly expose different inputs and horizons, not an error in either subtraction.

The Bank's temporary-fuel explanation makes a testable prediction: household expectations should retreat if fuel conditions stabilise and broader price pressure remains contained. Persistence would look different, repeated survey readings near July's level, wider pass-throughpass-throughPass-through describes an increase in one cost, such as fuel, spreading into the prices of other goods and services. into other prices, or an increase in measures of underlying inflationunderlying inflationUnderlying inflation refers to measures intended to identify persistent price pressure while looking through unusually volatile or temporary movements. [Bank of Russia](https://www.cbr.ru/eng/press/pr/?file=24072026_133000key_e.htm). [Governor's statement](https://www.cbr.ru/eng/press/event/?id=32727)

• The full July inflation-expectations commentary is scheduled for 28 July. [Calendar](https://www.cbr.ru/eng/calendar/)
• June company loan and deposit rates are scheduled for 6 August. [Calendar](https://www.cbr.ru/eng/calendar/)
• June household loan and deposit rates are scheduled for 10 August. [Calendar](https://www.cbr.ru/eng/calendar/)
• July company and household rate series are scheduled for 8 September. [Calendar](https://www.cbr.ru/eng/calendar/)
• The Bank's next scheduled key-rate meeting is 11 September. [Decision](https://www.cbr.ru/eng/press/pr/?file=24072026_133000key_e.htm)

Until those releases arrive, the defensible conclusion is narrow. Household expectations dominated one clearly defined subtraction. They did not cause a measured fall in borrowing costs, and the arithmetic says nothing by itself about the eventual economic effect of the rate decision.