Using EVDS 3 for Turkish Economic Data in 2026: Downloads and Sound Comparisons

[سفير طرابزون]

Using EVDS 3 for Turkish Economic Data in 2026: Downloads and Sound Comparisons

Start with a good question, then choose a chart
EVDS brings together economic time series for following prices, interest rates, exchange rates, credit and other indicators. Choosing a quick chart without reading the data definition can lead to comparing a monthly rate with a weekly balance, a nominal amount with a real index, or millions with billions. Begin with a specific question and the units, frequency and coverage needed to answer it.

This guide explains how to read, download and compare data. Its EVDS 3 sources were checked on 8 October 2026. The numerical examples are fictional and do not represent Türkiye's current statistical releases. We will build a repeatable process: select a series, inspect its metadata, retain the original data, make an appropriate calculation and write a clear conclusion. Readers should understand the indicator choice and be able to reproduce the comparison rather than rely on a screenshot.

The official EVDS 3 address and the 2026 upgrade
The official service is EVDS 3, provided by the Central Bank of the Republic of Türkiye. In its 26 January 2026 announcement, the bank launched the third version in beta, with improvements to search, presentation, themes, metadata and the release calendar. The beta description refers to the announcement date; check the platform's current status when using it instead of assuming that status stays unchanged.

Do not use illustrated EVDS 2 instructions to locate buttons in EVDS 3. If the menus move, retain the underlying steps: identify the data, read its definition, select a period and use the display or download options currently offered. These steps are more reliable than memorising a button's location in a changing interface. When writing instructions for others, state the version and address so readers start in the intended interface.

Public data and optional membership
The platform's official privacy policy explains that public data is available free of charge without membership, while registration is optional for some personalised services. Do not make an account a prerequisite for reading a public series when the page permits access. Distinguish access to data from saving personal settings and reports.

If you need a personalised feature, read its current description and conditions. Do not put account details or service keys in a public file to demonstrate your source. For someone seeking a figure or dataset, recording the data-group name, series code, period and download date is often sufficient. Economic documentation concerns data and its meaning; it does not require revealing membership details or creating accounts for others.

The Istanbul Financial Centre in an archive photograph from September 2023.​

Search by keyword, then read the full name
Start with a specific Turkish or English term for the indicator, then inspect the results and data groups. A broad term such as “interest” can return policy, loan and deposit rates at different frequencies. Instead of taking the first result, read the group name, series name and producing institution. If the question concerns commercial lira loans, a heading saying only “loans” is insufficient.

The EVDS 3 documentation page includes guides to search, all series and report pages. Consult the guide matching your screen, particularly when a result is unexpected. Failing to find an Arabic keyword does not establish that the indicator is unavailable. Try its official name or producing institution, then record the actual name in your file.

Metadata: the most important part of the process
Check the units, frequency, coverage, series start date, seasonal adjustment and base year for an index. Record whether observations are balances at a date, flows over a period or averages. These details matter as much as the value because they determine which calculations have a meaningful interpretation.

For practice, imagine two results labelled “deposits”: one measures end-of-week balances in millions of lira; the other is an annual interest rate on new deposit business. Adding their values or putting them on one unexplained axis is inappropriate. Similar-looking codes can describe different concepts, revealed by their units. Keep a short metadata record for each series before downloading another, so you do not later have to guess what an old column means.

Reference period, publication date and download date
A file can involve three dates: the period described by a value, the date of publication and the date you obtained that version. A September value may be published in October and downloaded a week later. Calling it “today's data” without specifying the date can imply that the figure measures today's activity.

Select a period covering the required comparison and confirm that both observations exist. A year-on-year comparison for a month requires the same month of the previous year; a rolling 12-month calculation requires the complete window. Do not replace a missing month with the latest available observation simply to finish a chart, or display missing data as zero. State the last actual reference period below the table or in the accompanying report.

Select the original series before transforming it
Begin with the original values appropriate to your question and keep transformations separately. If you select a pre-calculated rate of change on the platform, do not calculate another rate of change from it and assume that it measures the underlying economic growth. When using a seasonally adjusted indicator, retain that choice throughout the comparison instead of switching to unadjusted data partway through.

Give columns informative names: original balance, absolute change, percentage change or rebased index. Calling every column “value” makes calculations hard to trace. If two index groups have different base years, check their comparability, methods and revisions before linking them. Rebasing alone does not resolve a change in an index's basket or definition.

Download the file and check the transfer
Use the download or export option in the current interface and select a format suitable for your software from those actually available. Avoid manually retyping a long table when a download is possible. Save the original under a name identifying the indicator, period and download date, then work in a separate calculations copy.

After opening the file, check the first and last dates and three observations spread through it. Make sure a decimal separator has not become a thousands separator and dates have not been interpreted as different months. Compare a platform value with the downloaded value, and inspect missing cells and explanatory symbols. Many errors arise when spreadsheet software changes a column's data type after download, rather than in the economic source itself.

Skyland towers in Istanbul's Seyrantepe district, in an archive photograph from May 2024.​

Matching frequency does not mean adding every observation
Monthly flows, such as activity during a month, can be summed into a quarter when the months and coverage are consistent. Weekly balances should not be added together to create a monthly balance. Usually you need the final observation in the period, or an average if your question concerns a typical level. Explaining the choice is part of reporting the result.

Suppose weekly balances are 10, 12, 11 and 13. Their sum, 46, is not the month-end balance. The final observation is 13 and the weekly average is 11.5. Either can serve a different question, but must be labelled. Likewise, do not simply add monthly inflation rates and call the result cumulative inflation: an index or compounded change factors provides a more appropriate measure of accumulation.

A quick reading table

Data type	Treatment explained here	Required label
Monthly flow	Sum consistent months into a quarter.	The period's flow and coverage.
Weekly balance	Use the final balance or an average, depending on the question.	Period-end balance or average.
Interest rate	Distinguish it from balances and monetary units.	Interest rate and its unit.
Rebased index	Divide each value by the starting value and multiply by 100.	Common base and relative change.
Missing observation	Retain it as missing.	Identify any separately produced estimate.

An example of comparisons across time
Imagine a monthly index at 120 in the current month, 115 in the previous month and 100 in the same month a year earlier. The month-on-month change is approximately 4.35%, while the year-on-year change is 20%. Neither is inherently more correct; each answers a question about a different interval.

If using a quarterly average, add the three monthly levels and divide by three when averaging is appropriate for that indicator. Do not compare this average with the final day's level in the previous quarter and present the result as a consistent quarter-on-quarter comparison. For seasonally adjusted data, read the producer's definition before choosing a growth measure. In economic reporting, “annual” may refer to year-on-year or year-to-date, a distinction that needs to be explicit.

Aligning units: balances, loans and interest rates
Suppose you download loan balances in millions of lira and deposit balances in billions. Convert one unit before calculating the loan-to-deposit ratio. Fictional loans of 10,000 million and deposits of 20 billion mean deposits of 20,000 million and a ratio of 50%. Dividing 10,000 directly by 20 produces a meaningless comparison because the units differ.

Now add a fictional interest rate of 40%. Do not add 40 to the loan or deposit balances or present it in billions. You can use separate charts, or carefully labelled axes for a descriptive comparison. Often the clearest option is to chart the balances together and the rate separately. Clear units matter more than fitting every series into one crowded panel.

To compare trends more clearly, you can rebase two series to 100 at a common date by dividing each observation by its starting value and multiplying by 100. This removes size differences from the chart, but does not turn loans and interest rates into one economic concept or establish causation. State that the chart compares relative change from a chosen base, and retain the original values to show scale when needed.

Revisions, missing values and changes in definition
An old observation may change after a source or methodology update. Keep your download date and read revision notices and metadata when the start date or latest observations change. Do not combine two releases without understanding their differences, or assume that an unbroken chart proves the definition stayed constant throughout.

The official reserves metadata offers an example of how a methodology document records units, frequency and revisions. Apply that approach to each indicator instead of assuming the platform gives every dataset the same meaning. Leave a missing value missing and record the reason when available. If an analytical estimate is needed, distinguish and label it explicitly so it does not circulate as an official published figure.

Checks before turning the file into a news report
Record the question, series names and codes, original producer, units, frequency, transformation method and comparison period. Check that dates align and observations have not been counted twice, and manually verify three calculations. A spreadsheet formula running without an error does not mean it uses the correct column.

In the report, distinguish the producer's published figure from a rate you calculated. Give the source link, period and method briefly, then discuss what the comparison shows. Do not turn simultaneous movements into proven causation, or declare an economic decline from one indicator whose coverage you have not established. Retain the original file, calculations and definition notes together so another reader can check both the result and its basis.

Try this exercise with the saved file: choose two series measuring the same type of balance, in the same units, at a common date, and retain the original observations in an unchanged worksheet. In another sheet, record the base date and value, then calculate the change at each date. Ask another reader to recalculate the first and last percentage without seeing your report. If results differ, check cell types, separators and dates before changing the conclusion. Label the chart as showing period-end balances or averages. At the next update, keep the previous file so you can explain revisions to older observations instead of losing the record of the change.

Frequently asked questions
Do I need membership to obtain public data?
The platform's policy states that it is free without membership. Some personalised features may require optional registration, depending on the current service.

Should I use EVDS 2 instructions for EVDS 3?
Use the third version's guides and official address. The calculation principles remain relevant, but interface labels and locations can differ.

Does adding weekly balances give the monthly balance?
No. Select the final observation or an average according to the question, and explain your choice. Adding balances does not give a period-end level.

Why does a value differ from an older news report?
The period, units, release or method may differ. Check these before deciding that either number is wrong.

Should a blank cell be replaced with zero?
A missing value is not an economic zero. Retain it as missing until you establish the reason or produce a clearly identified separate estimate.

Does rebasing to 100 make series equivalent?
It helps display relative change, but does not align their coverage or economic meaning, or repair an undocumented definition change.

Does a simultaneous rise in two series prove that one caused the other?
A chart alone may show movements occurring together. Causation requires an analytical design and information beyond comparing lines.

Related Topics
Following Economy and Business News in Türkiye
Trade and Logistics in Türkiye: Exports, Imports, Ports and Customs
Industry and Manufacturing in Türkiye: Production, Supply Chains and Local Impact
Companies and Deals in Türkiye: Reporting Results, Mergers and Acquisitions
Markets in Türkiye: Following Borsa Istanbul, Gold and the Lira
Taxes, Fees and Regulations in Türkiye: Reading Changes That Affect Households and Businesses
Banking and Monetary Policy in Türkiye: Understanding CBRT Decisions
Türkiye’s Economic Indicators: A Practical Weekly Reading Guide
​