Yelo Bank and Expressbank accounted for disproportionately large shares of complaints received by the Central Bank of Azerbaijan in September 2026 relative to their shares of the measured customer base. Yelo Bank’s complaint share was 3.49 times its customer share, while Expressbank’s was 2.44 times. These were the highest readings among the 17 banks in the regulator’s table.
For both banks, this is an unfavourable signal about customer complaints. A large customer base alone cannot explain the readings because the methodology already adjusts for its share. Unibank followed with an index of 2.20.
Two banks in the upper complaint category
The central bank placed Yelo Bank and Expressbank in its upper category. Unibank, Azər-Türk Bank, Rabitəbank and Bank of Baku were in the middle category. The other 11 banks in the table were in the lower category. Bank VTB Azərbaycan, PASHA Bank and Xalq Bank had the lowest readings at 0.11, 0.16 and 0.31.
ABB’s index was 0.90, Kapital Bank’s 0.79 and AccessBank’s 0.78. The document lists 17 banks. The regulator states that it received no complaints during the month concerning banks absent from the table.
All banks in the report
| № | Bank | Complaint index |
|---|---|---|
| 1 | Yelo Bank | 3.49 |
| 2 | Expressbank | 2.44 |
| 3 | Unibank | 2.20 |
| 4 | Azər-Türk Bank | 1.81 |
| 5 | Rabitəbank | 1.77 |
| 6 | Bank of Baku | 1.54 |
| 7 | TuranBank | 1.02 |
| 8 | Bank Respublika | 0.92 |
| 9 | AFB Bank | 0.91 |
| 10 | ABB | 0.90 |
| 11 | Kapital Bank | 0.79 |
| 12 | AccessBank | 0.78 |
| 13 | Premium Bank | 0.70 |
| 14 | Yapı Kredi Bank Azərbaycan | 0.61 |
| 15 | Xalq Bank | 0.31 |
| 16 | PASHA Bank | 0.16 |
| 17 | Bank VTB Azərbaycan | 0.11 |
Why Yelo Bank’s reading stands out
The index divides a bank’s share of complaints by its share of the relevant customer base. A reading of 3.49 means the first share is 3.49 times the second. It does not mean that 3.49% of customers are dissatisfied, nor is it a count of complaints per customer.
A reading of 1 means the two shares are equal. The regulator’s colour categories are separate: TuranBank, for example, is in the lower category with a reading of 1.02. The boundary between categories should therefore not simply be drawn at 1.
Which complaints and customers count
The calculation includes oral and written complaints and complaints submitted during citizen reception appointments at the central bank over the month. Repeat complaints are excluded, as are requests for credit concessions, requests to obtain credit and matters outside the regulator’s remit.
The customer-base measure uses customers with current-account balances of at least AZN 100 and at least one transaction in the previous three months. It is therefore a defined statistical base rather than any headline customer total reported by a bank.
The answers customers still need
The September results raise concrete questions for the banks: which problems lead customers to the regulator, how many complaints are upheld, and what corrective steps have been taken? Without that information, a high index is a reason to scrutinise service quality rather than an explanation of what went wrong. The central bank document does not include responses from the banks.
The central bank explicitly states that the measure does not indicate threats to banks’ financial stability or continuity of operations. The table does not disclose absolute complaint counts, their causes or the outcome of investigations. It cannot establish the share of complaints found to be justified.
Readings and categories can change monthly as complaint volumes and customer bases change. Assessing a persistent service trend requires a sequence of observations, not a single September snapshot.
