TradeAlmanac
Sign in

CIR: what counts as costs, what counts as income, and why the denominator is the more contested part

· 6 min · beginner

Automated material · TradeAlmanac editorial deskDraft prepared by a language model from our stored data; not reviewed by an editor.

34 views
CIR: what counts as costs, what counts as income, and why the denominator is the more contested part — Investing basics

The cost-to-income ratio (CIR) is a bank's operating expenses divided by its operating income before provisions. The metric answers one question: how many kopecks the bank spends on running itself in order to earn a rouble of revenue. The lower the value, the more income reaches the profit line without being eaten up by salaries, rent, IT and marketing. CIR is a working tool specifically for the financial sector: a bank has no cost of goods sold in the industrial sense, so efficiency has to be measured not by margin but by the share of its own costs in what it has earned.

What exactly goes into the numerator

The numerator holds operating expenses, which in IFRS statements is the operating expenses line: staff costs and related taxes, depreciation and amortisation, rent and premises maintenance, IT and communications, advertising, consulting, and other administrative items. Crucially, the numerator does not include provisions for credit losses or interest expense. Interest expense is the price of the funds the bank has raised, which makes it part of the business itself rather than a cost of servicing it; provisions reflect the quality of risk, not the efficiency of the operating machine. If provisions end up in the numerator, the metric stops distinguishing between "an expensive bank" and "a bank with a bad portfolio", and those are two entirely different diagnoses.

Why the denominator is more contested than the numerator

Almost everyone calculates expenses the same way. Income is another matter, and this is where most of the discrepancies hide. The canonical denominator is operating income before provisions: net interest income plus net fee and commission income plus other operating income. The question is always about the "other" part: the trading result, revaluation of securities, income from foreign exchange operations, and non-recurring gains from the sale of assets and subsidiaries.

Such items are volatile and sometimes negative. A good quarter in the market inflates the denominator and mechanically improves CIR, even though not one process in the bank has become any cheaper. A bad quarter does the opposite. That is why an analyst comparing periods usually recalculates the metric on a "clean" denominator made up of interest and fee income, and shows the non-recurring items separately. If a research note does not say which denominator was used, such values cannot be compared across sources.

How to read the trend

A falling CIR is not an achievement in itself. Work out what exactly has changed:

Step 1 — see whether the numerator has moved. A genuine gain in efficiency shows up as restrained growth in expenses while income is rising: economies of scale are at work, and a new client is served almost for free because the infrastructure has already been paid for.

Step 2 — check the denominator for non-recurring items. The sale of a non-core asset improves the ratio one time and then stays in the comparison base forever.

Step 3 — take the phase of the investment cycle into account. A bank that is building an ecosystem or rewriting its IT landscape worsens its CIR temporarily and deliberately: the expenses come now, the income later. A persistently high value without an investment programme is a different signal, and a worse one.

Step 4 — keep cost inflation in mind. Salaries of IT specialists and spending on replacing imported software weigh on the numerator regardless of the quality of management, so in such periods it is more accurate to compare banks with each other rather than a bank with its own past.

What CIR is not

It is not profitability. A bank can have a low CIR and a weak ROE — for example, if it operates cheaply but loses a lot on provisions. And the reverse is true: an expensive operating model is compatible with high profitability if the bank occupies a niche with a wide margin. CIR describes only one link in the chain: from operating income to profit before provisions. For the full picture it is read together with the net interest margin, the cost of risk and the return on equity.

Nor is it a universal metric. Carrying it over to retail, steelmaking or telecoms makes no sense — those sectors have a cost of goods sold and the familiar margin indicators. CIR belongs to the family of ratios in which the meaning lies not in the absolute figure but in the basis of comparison — just like the debt-to-GDP ratio, which says nothing without the country and the maturity of the debt.

Where to get the data

The platform's database has no ready-made metric code for CIR, so the value is calculated from primary reporting: the banks' publications are gathered in the issuer reports section, and the release dates are in the events calendar. In their investor presentations banks, as a rule, disclose CIR themselves and state their own methodology — and it is the methodology that matters more than the figure. The pages for the sector's securities are in the Russian stocks section; for the largest issuer this is {{instrument:SBER}}, and the market's valuation of its earnings is 3,77.

For an investor, CIR matters as a leading indicator of earnings quality. Operating expenses are sticky: they cannot be cut by the next report without closing branches and laying people off. A bank with a persistently low cost-to-income ratio therefore gets through a margin squeeze more easily — it has a gap between revenue and costs that does not need to be defended immediately. A bank with a high CIR in the same cycle slides into a loss before provisions more quickly, and so loses its ability to pay sooner. If the dividend policy is tied to profit, the CIR trend points to the sustainability of payouts long before the board of directors announces its recommendation — the nearest decisions in the sector can be seen in the dividend calendar.

The metric should be treated as part of a set, not as a verdict. A low value is a reason to ask what produced it and whether that can be repeated in the next period; a high value is a reason to ask whether it is being paid for by future income. Definitions of related indicators are collected in the platform's glossary.

Related instruments

Share
Was this useful?
How this material was prepared

Draft prepared by a language model from our stored data; not reviewed by an editor.

Model: claude-opus-5

How we use language models

Similar articles