Reading TOBB’s Company Formation and Closure Statistics in 2026

سفير طرابزون · Thursday at 12:09 PM

Reading TOBB’s Company Formation and Closure Statistics in 2026

A report of rising company formations may seem to measure jobs created immediately or projects already producing goods. Similarly, a closure count may become a bankruptcy count in a headline. Formation and closure statistics are useful, but their meaning depends on the register, unit, period and classification. They do not automatically measure every economic outcome that might follow registration.

This practical guide explains how to read bulletins from the Union of Chambers and Commodity Exchanges of Türkiye (TOBB). Last reviewed on 8 October 2026. Fictional figures illustrate the calculations; no actual company counts for the current month or conclusions about reasons for closure are published here. The aim is to understand what a table measures, how to compare two versions and what requires an additional source before writing a broader report on business, investment or employment.

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

1. Start with the bulletin and its metadata

TOBB provides pages for monthly statistics, metadata and publication dates. The metadata explains that the series is based on announcements in the Turkish Trade Registry Gazette, with data produced by trade registry offices and coordinated through the union. Its starting point is therefore registration and announcement events, rather than a survey asking every shopkeeper whether their business is profitable or operating at full capacity.

Open the table for the period you need and read column headings and footnotes before using the news summary. The same release may include companies, cooperatives, commercial enterprises owned by natural persons and branches. These units should not be combined without explanation. Record the table name and category in your notes, retaining the bulletin link and version. A circulating screenshot may conceal a classification or footnote that changes the meaning of the headline figure.

2. The reference period differs from the publication date

The metadata defines the monthly period as the first through the last day of the month, while publication occurs on a different date under the release calendar. Do not describe a figure published in October as the number of companies formed in October unless that is the period specified. Name the month covered by the data; add the release date where it helps explain how current the information is.

At the calendar review on 8 October 2026, September 2026 data was scheduled for release on 23 October at 10:00. September appearing in the calendar indicates a scheduled release; it does not establish that its data had been published at the review date. This example shows why a schedule must be read differently from a results table. In each subsequent review, check which release is actually available rather than replacing a planned date with a confirmed publication event.

3. Companies, sole proprietorships, cooperatives and branches are different units

Read the original category name, then translate it without erasing the distinction. A count of new companies differs from a count of registered commercial enterprises owned by natural persons, or sole proprietorships, and from a count of cooperatives. The metadata also explains that branch counts concern branches opened or closed by existing companies. Branches may matter when reporting network expansion, but they do not automatically represent independent companies formed for the first time.

If combining categories to present a particular total, state its components and the reason for combining them. Do not label a sum “companies” when it includes units the table does not call companies. Nor should a broad group this month be compared with a narrower category in the previous month. Keeping a reference card of category names and their positions in the bulletin prevents a common error when news summaries change or a cropped table hides a row heading.

4. Separate formation from the start of operations and commercial success

A recorded formation establishes an event within the series' definition; it does not alone identify the first sale, employee or production date. A business may need equipment, licensing or funding after registration, and activity may begin gradually. These stages are not being attributed to every company in the table. They illustrate why moving from an administrative entry to an operational outcome requires additional evidence appropriate to your question.

Conversely, an existing activity may change its legal form or structure, so do not treat every registration as an equal addition to the economy's production capacity. For employment, consult suitable labour statistics. For actual investment, seek a measure of expenditure or implementation. Formation data can indicate recorded activity, but becomes less useful when made to carry conclusions outside its scope.

5. Distinguish closure, liquidation and bankruptcy

Where the bulletin presents closure or liquidation categories, retain those terms and examine the accompanying definitions. Do not turn every recorded closure into a bankruptcy, or a bankruptcy occurring in the month the bulletin was published. The reason an entity leaves the register may require another document, and the date operations actually ceased may differ from the registration date. The count alone provides neither a record of reasons nor grounds for assigning unpublished categories of causes.

Similarly, do not automatically add the liquidation row to the closure row and present the result as distinct entities permanently leaving the market. The rows may reflect different stages or categories; combining them requires understanding potential overlap. If the material does not explain their relationship, present them separately. Retaining the authority's definitions is better than creating an appealing total that may count an event or entity twice, or substitute one stage for another.

6. Compare like months and equivalent cumulative periods

Distinguish a comparison with the previous month, the same month of the previous year and a cumulative period. Name the comparison before giving its percentage. Working days, holiday timing or the pattern of registration may change, but a monthly difference does not automatically prove any of those causes. They are possibilities to investigate if offering an explanation; a descriptive report can simply give the correct comparison without an undocumented cause.

Do not compare eight months with a full year and call the difference annual growth. When adding months, use the same version and categories wherever possible, then check for updates or revisions. Comparing with the equivalent period a year earlier may reduce some seasonal differences, but does not remove every calendar effect or issue with definitions. A calculation answers only the question for which it was constructed; it does not ensure that the question covers every aspect of business activity.

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

7. A fictional example: counts and growth rates

Suppose one consistently defined category records 1,000 formations in a reference month and 1,200 in the corresponding month the next year. The difference is 200 and the increase is 20%. If closures in the same category rise from 100 to 150, the difference is 50 and the increase is 50%. Different starting values explain why the larger percentage can accompany the smaller numerical change. Do not report the percentage alone.

An accurate report of this example gives both counts, the period and definition, and may present the differences alongside the percentages. It does not say that 200 jobs were created or 50 companies went bankrupt, because those meanings are absent from the example's assumptions. Nor does it attribute the increase to a decision or crisis without further evidence. Calculating change is the start of the analysis; interpreting its effects and causes needs other questions and data beyond the arithmetic.

8. Formations minus closures are not the stock of operating companies

Subtracting 150 closures from 1,200 formations in the example gives an arithmetic difference of 1,050 between the recorded flows. That description is valid if the categories match, but it is not the number of companies operating at month-end. Establishing a stock requires its starting value, the relationships linking flows to it, coverage and other relevant changes. The calculation alone does not prove that the actual stock increased by that amount.

Nor is the ratio of closures to formations a failure rate for new companies. Companies closing this month may have been formed in different years; they are not necessarily part of this month's formation cohort. Studying survival requires following a defined formation cohort through time under a clear definition. Distinguishing two separate flows from tracking the same group prevents a headline claiming that a particular share of new companies failed when the data did not follow them.

Comparing the concepts explained in this guide

Concept or stage	What the measure establishes
Reference period	The monthly period from the first through the last day of the month.
Publication date	A separate date under the release calendar; it does not automatically identify the data month.
Recorded formation	Does not alone establish the first sale, employee or production date.
Branches	Branches opened or closed by existing companies do not automatically represent independent companies formed for the first time.
Formations minus closures	Is not the stock of operating companies.

9. Registered capital does not equal foreign direct investment

Where a table reports capital, check whether it describes the amount registered at formation or another category, and read the currency and units. Do not turn the total into cash entering the economy during the period without explaining the source. Nominal capital, paid-in capital, an acquisition price, actual fixed investment and international flows are different concepts. A series counting companies with foreign participation cannot replace a financial series measuring investment flows.

The metadata defines the category of foreign-capital companies using elements including a partner's nationality or country of residence. The count measures entities within that definition; alone, it does not establish the amount of funding, the degree of foreign control or the effect on employment. When comparing with another count, ensure the category and period match. Do not infer that a company with a foreign shareholder is wholly foreign-owned unless the source explicitly gives that ownership percentage.

10. Read the notes on sector and location

Formation counts by activity or province may help identify where registered activity is occurring. Read the activity classification and how location is assigned; do not turn a table heading into a map of every place where companies actually operate. A company registered in one province may have activities or branches elsewhere: the register is not a map of every investment. If the notes do not explain the matter relevant to your question, make the limits of interpretation visible in your report.

When comparing two sectors, an absolute formation count alone does not establish which is stronger. Sector size, the nature of businesses and registration requirements may differ; assessing a rate or performance requires additional data. A sector heading may also cover several activities within the classification, so do not replace the group with the name of a narrower activity. Use the table's terminology and explain its meaning, especially where the everyday name is narrower than the published scope.

11. Versions and new classifications affect comparisons

TOBB states that revisions are announced on its official statistics page. Retain the version date of the tables used. When recalculating, obtain the comparison periods from a compatible source rather than mixing an old saved value with a current one. If you find a difference, examine the announcement, notes and method; do not automatically treat it as a contradiction or an economic change that occurred in the past. Data or presentation coverage may change as documented by the organisation.

In 2026, TOBB announced the addition of tables on women and young entrepreneurs beginning with January data. A new table opens a useful line of enquiry, but comparisons with earlier years require checking that compatible definitions are available. Do not fill the historical gap with zero: an unpublished value does not mean the phenomenon did not exist. When presentation changes, read the group's full definition before naming what the count measures or attributing it to all business owners.

12. Checklist and common mistakes

Identify the bulletin, reference period, publication date and version.
Separate companies, sole proprietorships, cooperatives and branches.
Review closure and liquidation definitions before combining or interpreting them.
State whether the comparison is monthly, annual or cumulative.
Present the starting count, difference and percentage where useful.
Do not turn the difference between flows into a stock of operating companies.
Do not treat the closure-to-formation ratio as a survival or failure rate.
Check sector, location, capital and new definitions.

Common errors include confusing the data month with the publication month, extending one category to a broader total, treating registered capital as actual expenditure and attributing every closure to bankruptcy. Official figures may be correct while a headline misrepresents them. First write a descriptive sentence reflecting the table's definition, then add analysis you can support. Leave undocumented causes as questions for investigation rather than stated facts.

13. Frequently asked questions

Do new company counts measure new jobs?
They measure what the registration series defines, rather than establishing employee numbers or the start of operations. Use employment data or appropriate operational evidence to answer that question, retaining formation counts as registration activity within the published period and category.

Has every closed company gone bankrupt?
A closure count alone cannot establish that. Read the term and definition, then find the source needed to identify the reason for exit. The discontinuation of a registration or activity in a particular document does not justify the same financial judgement about every case.

Do net formations give the number of existing companies?
They are a difference between flows where categories match, whereas a stock requires its starting value and relevant changes. Do not use one month's difference as the number of existing companies or aggregate it over periods without checking compatibility with the question and method.

Does dividing closures by formations measure the failure of new companies?
No. The two counts may concern companies of different ages. Measuring survival requires tracking a cohort formed in a defined period. You may present the ratio under its proper name, but do not rename it a failure rate without suitable follow-up data.

Why might last month's data be unavailable at the start of this month?
The series has a release calendar distinct from its reference period. Check the official date and the release actually available. Do not treat a scheduled date as published results or a gap in a limited search as proof that the data was never produced.

How should I handle a saved figure that differs from the current table?
Compare category definitions, periods, notes and version dates, and look for a revision announcement. Recalculate from compatible periods where necessary. State that you used the version available on a particular date, rather than combining versions that may measure different details.

Related topics

Reading TOBB’s Company Formation and Closure Statistics in 2026

سفير طرابزون

Active member