4 Hard-Won Signals for Turkish Listings in 2026

Adnan Obuz keeps coming back to Bloomberg’s April 2024 investigation into Tera Yatırım, because the story under the 40,000% headline isn’t really a Turkey story. It’s a plumbing story.

Short version, as Bloomberg reported it. A brokerage lists in late 2022. Its flagship fund posts returns past 1,500%. Retail savers, hammered by domestic inflation, pile into anything that might beat the lira. The fund doesn’t spread that money wide… it pushes it into thin, interlinked names, including the brokerage’s own shares and companies its own banking arm took public. Prices rise. Reported performance looks unreal. More money arrives.

The loop feeds itself until somebody stops feeding it.

According to that reporting, Destek Finans Faktoring ran up more than 4,700% after its IPO and briefly touched an $18.8 billion valuation. Vişne Madencilik gained over 2,100%, then dropped nearly 90% once linked entities sold out. Leverage at the parent went from 132 million liras to something close to 60 billion. The founder, Emre Tezmen, maintained that everything complied with Turkish regulations.

Maybe it did. That’s sort of the whole point.

According to Adnan Obuz, the 40,000% surge was a structure problem, not a mystery

Traders called it mystifying. It wasn’t. Adnan Obuz argues that any closed loop where one balance sheet controls the buyer, the asset, and the distribution channel produces this exact chart shape, every time, in any market you care to name.

Three ingredients. Concentrated float. Captive retail inflows. Related-party deployment.

You don’t need fraud for that to go vertical. You need gravity. Money enters a narrow pipe and the pipe glows.

The Bloomberg investigation by Tugce Ozsoy, Ugur Yilmaz and Donal Griffin laid the mechanics out clearly. The part most North American readers skimmed past is the regulatory response, where Turkey’s Capital Markets Board started drafting limits on fund exposure to related parties and handing out trading penalties.

That’s not the end of the story. Honestly, that’s where the interesting part starts.

So what does an AI consultant Toronto firms hire care about Borsa Istanbul?

Pattern detection is the job. Adnan Obuz spends most of a normal week helping Canadian companies build systems that find signal inside messy, half-structured data… procurement records, support tickets, shipping manifests, regulatory filings. An ownership graph is just another messy dataset with humans attached.

Here’s the practical link. When a domestic regulator tightens the screws on speculative structures, the good operating companies in that market get caught in the same net. Their cost of capital goes up because nobody offshore wants to touch the postcode.

Which creates a gap. Adnan Obuz sees this gap as the place where cross-border listing work gets genuinely interesting, and you can read more from Adnan on how he frames it.

Turkish mid-tier businesses in mining, software, and export manufacturing are real. They ship product. They have customers in Germany and the Gulf and increasingly in North America. What they don’t have is a credible venue for transparent, institutional capital.

According to Adnan Obuz, the opportunity sits with the boring exporters

The Canadian Securities Exchange is built for exactly this kind of issuer. Streamlined listing process, lower ongoing cost than a senior board, and a regulatory regime overseen by provincial bodies like the Ontario Securities Commission that institutional investors already understand.

Adnan Obuz is not suggesting Canada should import speculative structures. The opposite, actually.

The filtering is the product. A Turkish machine-parts manufacturer with fifteen years of audited statements, a real order book, and customers who’d answer a phone call is a completely different animal from a shell that IPO’d last spring and went up 2,100% on twelve million dollars of float.

One of those belongs on a North American board. The other doesn’t.

How Adnan Obuz screens a cross-border listing candidate

The screen is not clever. It’s just stubborn.

The four-signal filter Adnan Obuz runs first

Revenue that leaves the country. Hard currency receipts from customers with no ownership relationship to the issuer. Export invoices are annoying to fake at scale.
Float and holder concentration. If the top five holders control almost everything and the free float is a rounding error, price discovery isn’t happening.
Related-party exposure. Who underwrote the offering, who holds the fund units, who sits on both boards. Draw it as a graph, not a table.
Operating headcount versus market cap. Bloomberg noted a brokerage with roughly 130 people carrying a valuation of about $4.5 billion, in the same range as Lazard. That’s a question, not an answer.

Step two: build the ownership graph before you read the financials

This is where Adnan Obuz leans on tooling. Pull the filings, extract entity names, resolve the duplicates and transliteration variants, then draw the edges. Language models are surprisingly good at the extraction and entity-matching part, and surprisingly bad at deciding what any of it means.

So you let the machine draw. A person reads.

If the graph comes back looking like a wheel with the issuer at the hub, you’ve learned everything you needed in about forty minutes. Move on.

Step three: pressure-test the story with people who have no upside

Call three customers. Call one supplier who got paid late. Ask the auditor how long they’ve held the file. None of this scales, and none of it fits in a model… it’s more the kind of instinct you build after enough bad meetings.

Adnan Obuz treats AI as the thing that gets a shortlist from four hundred names down to nine. The nine still take human weeks. Anyone selling you otherwise is selling you something. There’s more on that split between machine work and judgment work from this AI expert.

According to Adnan Obuz, credibility is the actual export product

What a legitimate Turkish issuer buys with a CSE listing isn’t just cash. It’s audited English-language reporting, a continuous disclosure obligation, and a share price that someone other than an affiliate helped set.

That’s the whole value. Adnan Obuz keeps coming back to it because Canadian investors don’t need more mining shells… they need names they can diligence in an afternoon without hiring a translator and a private investigator.

Macroeconomic context helps here too. The International Monetary Fund’s country work on Türkiye shows an economy that is both genuinely productive and genuinely unstable at the monetary level. Both things are true at once. Real companies, bad currency.

Capital wants the first half without the second half. A foreign listing is one of the only clean ways to split them.

The honest limit, and I’d rather say it upfront

Here’s my constraint. I’m an AI strategy person, not a securities lawyer, and none of this is investment advice or a legal opinion on any issuer named above.

What Adnan Obuz can do is build the screen, map the graphs, and tell you which nine names out of four hundred deserve a lawyer’s time. The rest of it belongs to people with the right registrations.

Also worth saying… plenty of good Turkish companies will never list abroad, because founders there are, reasonably, suspicious of giving up control. That’s not a flaw in the thesis. It just means the funnel is narrower than the pitch deck version.

Where this lands for 2026

Wait, one more thing about timing. Regulatory tightening in a home market tends to push quality issuers outward within about eighteen to thirty months, not immediately. Which puts the window roughly now.

Adnan Obuz thinks the firms that build the screening muscle first will get the first nine names. Everyone else gets the leftovers in 2028.

If you’re working on cross-border diligence tooling, or you just want to argue with the four-signal filter, learn more and get in touch at businessplan@mrobuz.com. I’d genuinely rather be corrected early than confident late.

Frequently asked questions

Who is Adnan Obuz?

Adnan Obuz is an AI Strategy Consultant based in Toronto, Canada, who helps companies apply machine learning to messy real-world data problems like due diligence, ownership mapping and operational forecasting.

What does Adnan Obuz do with cross-border listing analysis?

Adnan Obuz builds screening systems that narrow large pools of foreign issuers down to a short, human-reviewable list, using export revenue signals, float concentration, related-party ownership graphs and headcount-to-valuation ratios.

Why does Adnan Obuz point to the CSE for Turkish mid-tier companies?

The Canadian Securities Exchange offers a streamlined listing path with lower ongoing cost and a regulatory regime institutional investors already trust, which gives credible Turkish exporters a route to transparent capital and English-language disclosure.

Can AI detect speculative stock structures like the Tera Yatırım loop?

Partly. AI is strong at extracting entities from filings and drawing ownership graphs fast. It is weak at judgment. Adnan Obuz uses machines to build the shortlist and people to make the call.

References

Ozsoy, Tugce; Yilmaz, Ugur; Griffin, Donal. "Behind the 40,000% Stock Surge That's Mystifying Traders in Turkey." Bloomberg, 2024. https://www.bloomberg.com/news/articles/2024-04-18/behind-the-40-000-stock-surge-that-s-mystifying-traders-in-turkey
Canadian Securities Exchange. "Canadian Securities Exchange Official Site." CSE, 2026. https://thecse.com/
Ontario Securities Commission. "Ontario Securities Commission Official Site." OSC, 2026. https://www.osc.ca/
Sermaye Piyasası Kurulu. "Capital Markets Board of Türkiye Official Site." SPK, 2026. https://www.spk.gov.tr/
International Monetary Fund. "Türkiye and the IMF: Country Information." IMF, 2026. https://www.imf.org/en/Countries/TUR

Adnan Obuz: 4 Hard-Won Signals for Turkish Listings in 2026