Turkey's $20 Billion Fund Collapse Explained
An in-depth analysis of the massive financial collapse that shook Turkey's investment fund industry in September 2026. The commentary explores how flagship funds reported impossible returns through concentrated stock positions, self-reinforcing valuation loops, and portfolio pumping mechanics. It breaks down how retail money market funds became dangerously exposed through reverse repo loans backed by heavily inflated collateral, the regulatory intervention that halted the cycle, and the subsequent systemic economic damage.
#financialcrisis #investmentfunds #marketmanipulation #turkeyfinance #ponzischeme #macroeconomics #liquidityspiral #financialregulation #assetmanagement #bankingnews
AI Generated·May contain errors
The Turkish financier Emre Tesmin did not believe that anyone could control the stock market.
He told Bloomberg this back in April, saying you cannot control anything 100% in a market.
If that were the case, prices would only go up. Now that is a great point, but for the better part
of three years, prices at Terra, his financial group, did only go up. Its flagship investment
fund reported a return of more than 66,000% over three years, and that's considered really good
performance. Terra, a brokerage firm with around 130 employees, was at times worth more than the
global investment bank Lazard, which employs several thousand people, which seems a bit strange,
but possibly his 130 employees were just way more productive than the slackers at Lazard.
Then, about five months after that interview, everything fell apart. Turkey's capital markets
regulator ordered the liquidation of 131 investment funds holding about $20 billion
and 455,000 people found that their savings were locked inside. Turkey's justice minister
has since described the whole thing as a Ponzi-like scheme, which is never a good sign.
When a fund's returns depend on a steady stream of new money coming in, its collapse is going to
raise the question of who exactly was buying at the top. So, in this video, we'll look at how the
alleged scheme worked, why some Turkish money market funds, the product people treat like cash,
ended up holding the bag, what Bill Hwang and Bernie Kornfield have to do with any of this,
and what lessons the rest of the world might want to take away.
Mostly, though, we're going to try to solve a mystery. As his funds collapsed,
Tesman posted a statement on social media blaming the whole thing on a planned, deliberate,
and organized speculative attack by notorious malevolent forces, the worst kind of malevolent
forces. That is a serious allegation, and I think we should take it seriously. So,
over the course of this video, let's try and work out who these malevolent forces might be.
It would appear that Emery Tesman was rarely troubled by self-doubt. He told Bloomberg that
his ambition was to build the Goldman Sachs of Turkey. We should note that Goldman Sachs took
more than 150 years to build, surviving several financial panics, and grew to employ tens of
thousands of people while building the reputation that Matt Taibbi of Rolling Stone summed up
as a great vampire squid wrapped around the face of humanity.
Terra was a lot younger and smaller, but it had recently signed a deal to sponsor a Turkish
football club, so that's kind of similar. Tesman had a straightforward view of finance too.
I make money, the investor makes money, he told one interviewer, and what does the state lose?
There's no problem here, he said. We'll have to come back to that question though.
So, how did the Goldman Sachs of Turkey manage to so comprehensively outperform
the actual Goldman Sachs? Well, if you want a stock price to go up, you need buyers,
and if you run an investment fund, the most convenient buyer you can find is, of course,
yourself using your customers money. Bloomberg found regulatory filings showing that at one
point in 2023, Terra's flagship fund had 99% of its assets in a single stock,
shares of its own parent company, Terra. Most fund managers would consider that
a fairly concentrated portfolio. Orkan Saka, a lecturer at City St. George's,
University of London, described how this works to Bloomberg back in April, long before any of
it had fallen apart. The fund's purchases, he said, push up affiliated stock prices,
which inflates the fund's net asset value, which makes the fund management business
appear more valuable, which elevates the parent company's earnings and its stock price. As he put
it, this loop is self-reinforcing by construction. That's a very academic way of describing a
perpetual motion machine. Turkey's justice minister later went with the less academic
description of Ponzi-like. Before I dig into how Terra kept that machine running,
let me tell you about this week's video sponsor, Salee.
Mobile roaming is one of the last industries still run on the hotel mini-bar pricing model.
The product is perfectly fine, but the price is revealed later, and it's always considerably more
expensive than you expected. You land somewhere, check a map, send a few emails, and a month later
you discover that finding directions to your hotel cost about as much as the hotel did.
Salee is an eSIM app that solves this problem. You pick a prepaid data plan for wherever you're
going, and they offer affordable plans in over 200 destinations. Because it's prepaid, you know
exactly what you'll pay, so there are no surprise roaming fees and no need for angry calls to your
company. What I like most is the convenience. You install the Salee eSIM once and that's it.
There's no need to install a new eSIM for every country you visit. You just choose a plan in the
app and it activates automatically when you land. If anything does go wrong, their live chat support
is available 24-7, which is useful when you're in a different time zone and your problem doesn't
care what time it is back home. Salee also blocks malicious sites, trackers, and ads while you
browse, which has the pleasant side effect of saving you data. They now offer app-based phone
numbers too, so you can hand out a Salee number to shops and delivery apps and keep your real one to
yourself. To get an exclusive 15% discount on Salee data plans, download the Salee app and use
code BOIL at checkout or click the link in the description. So anyhow, a mild version of this
sort of thing happens everywhere. Academics call it portfolio pumping. In a 2002 paper called
Leaning for the Tape, Carhartt, Kanyal, Musto, and Reed showed that some American mutual fund
managers buy extra shares of stocks that they already own at the very end of the quarter,
nudging up the closing prices their performance is measured on. The prices then drift back down
the following day. The American version is worth a few basis points of performance a few times a year
in time for the quarterly rankings. Terra's fund was buying the shares of its own affiliated
companies all year long, and it reported 66,000% returns, which just shows you what you can achieve
with a bit of determination. To keep a machine like this running, you need a steady supply of
new money, and in Turkey, that wasn't very hard to find. Inflation had averaged about 50% a year
for three years, eating away at people's savings, so a fund reporting more than 1,500% in a single
year was going to get a lot of attention. And with inflation at 50% a year, a big number on a
fund statement probably seems a bit less crazy than it would somewhere else. Back in 1998,
the economists Siri and Tufano showed that investors pile into top-performing funds
far more eagerly than they pull money out of bad ones. Terra's brokerage also took companies public
– after all, they were trying to be the Goldman Sachs of Turkey – and its fund would then buy
their shares. Now, a small free float, the portion of a company's shares actually available to trade,
isn't that unusual on its own. But when the underwriter's own fund is one of the main buyers
of the few shares that do trade, it doesn't take much to move the price. Take Vizni Madinçek,
a mining company that I'm fairly sure I mispronounced there, which Terra took public
in February 2025, putting 21% of its shares in public hands. By the end of July, the stock was up
more than 2,100%. That same month, Terra's fund and brokerage sold the shares that they had bought
in the IPO. In August, the stock fell 75%. The Capital Markets Board fined two of Terra's fund
managers 8.9 million lira each, which is about $199,000, for creating a misleading perception
on the price, supply and demand of the shares. Keep the size of that fine in mind, we will come
back to it later. Then there was Destek Finance Factoring, an invoice factoring company that
took public in 2025. Its shares rose more than 4,700%, making it Turkey's third largest listed
company. Even after the market turmoil started, it was still trading at about 132 times earnings.
JP Morgan trades at about 15. And after the crash wiped out about $4 billion of his paper fortune,
Destek's controlling shareholder was left roughly as rich as the founder of Chobani Yogurt.
Invoice factoring is a perfectly respectable business, but it doesn't usually produce
Greek yogurt levels of wealth. It's usually just Greek yogurt that produces that kind of wealth.
Anyhow, back to our search for the notorious malevolent forces. Our first suspect is Ramazan
Basak, the former head of MASAK, Turkey's financial crimes agency. Mr. Basak says that
he repeatedly warned the authorities about suspicious transactions and that his warnings
were ignored. According to the FT, he was detained in April for disseminating misleading information
in two social media posts, one of which questioned whether a fund could really have returned 10,128%,
which I suppose makes sense. If people start doubting 10,000% returns, where does it end?
Still, Mr. Basak's attack consisted mainly of asking questions on social media, which isn't
usually enough to bring down a $20 billion fund industry. So we'll have to keep looking.
How did Terra pay for all of this? A fund that keeps buying its own affiliated stocks
needs a steady supply of cash, and so Terra borrowed a lot of it. According to Bloomberg,
its brokerage's borrowing went from 132 million lira at the end of 2024 to 59.8 billion lira a
year later. That's about 450 times as much in just one year. And a convenient place to find cash like
this is in a money market fund. People think of money market funds as more or less the same as
bank accounts. They're supposed to hold boring, safe, liquid things and earn an interest rate.
But Bloomberg reports that some of the firms at the center of this were lending money and taking
hugely inflated stocks as collateral. This type of loan is known as reverse repo, where the fund
makes a loan taking securities as collateral, which it can sell if the borrower doesn't pay it
back. That should be very safe, as long as the collateral is good. But in this case, the
collateral was pumped up shares in the thinly traded companies that the affiliated funds were
busy buying. It's a bit like taking a $10 watch to a pawnbroker, explaining that it's actually
worth $10,000 because you just sold one to yourself for that price an hour ago, and walking
out with a $9,000 loan. And as Hilmi Guvenal points out on his substack, the money market funds
booked these loans at face value, so their prices stayed perfectly stable, right up until the
collateral had to be sold. If this sounds familiar, it's because it's similar to what
Bill Hwang did at Archegos. Hwang used swaps from global banks to build enormous leverage
positions in stocks like ViacomCBS. The Turkish funds did their borrowing at home, and financial
experts who have looked at the trades say they borrowed against the inflated stocks themselves.
When Hwang's stocks fell in 2021, his banks made margin calls. When Terra's prices stopped rising,
it faced a run from retail savers. Both collapses happened very quickly, and economists have a name
for why this happens. Brunemeyer and Peterson called it a liquidity spiral. When prices fall,
the collateral is worth less, so you can borrow less, which forces you to sell, which pushes
prices down further. And the faster you're forced to sell, the less anyone wants to pay.
And that brings us to our second suspect, and I feel we're maybe getting closer. Bloomberg spoke
to Mehmet Kervanky, a 28-year-old who had put his savings into a high-flying Turkish stock fund
to help pay for his upcoming wedding. Two weeks later, he was frantically hitting the sell button
on his phone's trading app. Another investor, a 30-year-old trader, had put in the gold he
received as wedding gifts, hoping to use it towards a deposit on a home. So it's possible
that the planned, deliberate, and organized speculative attack on Terra was carried out
by a 28-year-old trying to pay for a wedding. But I don't want to jump to conclusions, so
we should probably keep searching. If you push the stock up a few thousand percent,
you only really make money if you can sell it at the high. And to sell at the top, you need someone
who is willing to buy at the top, which is really the core question here. Who was buying? Hilmi
Guvenal, writing on his substack, which I'll link to in the video description because it's one of
the clearest explanations of this whole mess that I found, points to one example, a shipbuilder
called Ozada. According to its June financial statements, Ozada was losing money and had
equity of around 4.5 billion lira. By September, its share price had reached 4,482 lira,
giving it a market value of about 320 billion lira or 71 times book value, which is a lot for
any shipbuilder, but especially so for a loss-making shipbuilder. At that point, one of Terra's hedge
funds owned about 15% of the company. Valued at the September price, that's roughly 48 billion
lira of investors' money in a single loss-making shipyard. Turkey's capital markets regulator has
since filed criminal complaints against 11 people over-trading in Ozada's shares.
Ozada has also made its way into Turkish politics. An opposition spokesman alleged
that a deputy chair of the ruling AK party bought about 63 million lira of Ozada's shares in April
and sold them in September for about 1.34 billion. He said the sales were completed before the
turmoil hit the stock market on September 16th and he made similar claims about the trading of
her husband. Bloomberg notes that none of this has been independently confirmed. She stepped down
from her party post saying that she wanted the claims clarified without leaving any room for doubt.
She hasn't been charged with anything and neither she nor her husband were among the 11 people named
by the regulator. To be clear, nobody has said who bought any of those shares. Guvenal notes
that the answer exists since every trade is logged day by day. But his broader point is the more
important one. When a fund is one of the biggest buyers of an inflated stock, it can become the
exit door for anyone selling at the top. And the people paying for that exit are the fund's
less informed investors. Economists have a name for insiders moving money out at the expense of
outside investors. In a 2000 paper, a group of researchers including the Nobel laureate Simon
Johnson called it tunneling. The classic example is a controlling shareholder selling something
to their own company at an inflated price. The insider gets the cash and the minority shareholders
pay for it. A fund can work the same way. If it pays an inflated price for shares, the seller
walks away with real money and the fund's investors are left holding the bag. So how did a machine
that was running so smoothly seize up? On August 29, the Capital Markets Board, Turkey's security
regulator, introduced a new rule capping how much of a company's free float a single fund could hold.
And as Hemi Guvenal points out, the rule was surprisingly gentle. Funds that were already
over the limit were given until 2029 to comply. So no one had to dump their shares right away.
They just couldn't buy any more. And a system that depends on someone buying every single day
has a problem when the buying stops. So that brings us to our third suspect for the malevolent
forces – the regulator. John Paul Rathbone of the FT said on the Unhedged podcast that the regulator
had described some of these fund prices as fictitious in an internal memo back in December.
It then spent eight months tightening things around the edges. And when the big rule finally
arrived, all it did was stop funds from adding to their positions, which it turns out was enough.
With nobody left to keep buying, the prices had nothing holding them up. In mid-September,
Terra announced that it was buying Pusala, another fund manager. The next day,
Pusala said it couldn't meet redemption requests. This was not reassuring. Investors pulled about
4.7 billion dollars out of Turkish investment funds in just over two weeks. And on September
16, Terra said that it too was unable to pay clients who wanted their money back.
The football club that Terra had sponsored played its next match without the Terra logo on its shirts
and the club's own shares lost about a third of their value that week. Now I would have expected
the lack of writing on their shirts to be viewed positively, but I guess they probably needed the
sponsorship money. Runs like this tend to feed on themselves, and there's a good reason for that.
If you have to sell your house in the next hour instead of in the next month,
you're going to have to cut the price, however nice the house might be.
A fund facing a wave of redemptions has the same problem. It pays the first people out
by selling whatever is easiest to sell, and the people who stay are left owning the thinly traded
stocks which can only be sold in a hurry at a discount. A well-known 2010 study of American
mutual funds found exactly this. Funds holding illiquid assets see much bigger outflows after
bad news because investors understand that waiting costs the money. So the people rushing for the
exit that week were actually just being shrewd. The ones who waited were the ones who ended up
paying for the discount. For the people who didn't get out on time, the results have been devastating.
One investor told Bloomberg that he had moved $2.4 billion into two Terra equity funds at the
end of August. My nerves are wrecked, he said. I can't sleep. And as an entrepreneur who lost 90%
of his investment in Pusala funds put it, these were all regulator approved funds.
How did this go on for so long then? Well, as Robert Armstrong put it on the FT's Unhedged
podcast, the problems were something of an open secret. Quite a lot of people could see what was
going on. In February 2025, Istanbul's chief prosecutor wrote to the capital markets regulator
asking it to look into unusual share trading by the chairman of Pusala. In November 2025,
Turkey's finance minister publicly stated that the government knew stock price manipulation was
taking place through certain funds. In June 2026, MSCI warned about recurring instances
of possible coordinated trading behavior. And yet the big rule change didn't come until the end of
August. Part of the reason this went on for so long may be that getting caught just wasn't
very expensive. The economist Gary Becker, who later won a Nobel prize, argued back in 1968
that criminals approach crime much like the way businessmen approach any other business decision.
They weigh the expected gains against the chance of getting caught multiplied by the penalty.
If stock market manipulation can make you hundreds of millions of dollars, and a fine if you get
caught is just $199,000, the math isn't very complicated. A fine like that is just a small
operating expense. Deterrence needs more than just fines. In 2001, the legal scholar Bernard Black
analyzed the institutions a country needs to have strong stock markets. His list includes
honest regulators, independent courts, auditors that can be held liable, and a free press.
Turkey has all of these things on paper. In practice, as the economist has noted,
officials were slow to act on the warnings. The former financial crimes official we met earlier
says that he was detained twice after raising concerns about terror, and terror threatened
When people don't trust the institutions, they often then rely on connections instead.
As the FT noted, the boards of companies caught up in this scandal included a former ambassador
to the OECD and a university rector, and Terra's financial holding company was run by a former
deputy governor of the central bank. Americans may remember that Terranos had two former
secretaries of state on its board, Henry Kissinger and George Shultz, which was comforting to
investors even if neither of them had any background in science or blood testing.
In a famous 2001 paper, the economist Raymond Fisman put a number on this sort of thing.
He looked at Indonesia under Suharto and found that whenever rumors spread about Suharto's health,
the stocks of companies closely connected to him fell much further than those of less well-connected
firms. Connections can be very valuable for as long as they last. Our fourth suspect for the
notorious malevolent forces has to be MSCI, a foreign index provider based in New York.
The Turkish government has spent years promoting Istanbul as a global financial center,
so I imagine they were excited to read MSCI's June 2026 market classification review.
It opened with good news for Bulgaria, which is being promoted up to frontier market status,
and closed with a reminder that Greece is being promoted to developed market status.
In between, it warned Turkey and Indonesia about coordinated trading activity on their stock
markets. Turkey was told that MSCI may launch a consultation on the appropriate treatment for
Turkey, which is possibly not a great review. As the economist read it, Turkey was being warned
that it might pass Bulgaria on the way down, and S&P Dow Jones has since put Turkey under review
for exactly that downgrade. They point out that frontier status would put Turkey in the
same category as much smaller markets like Tunisia and Romania. This isn't what you want to hear
when you're trying to build a global financial center. John Paul Rathbone suggested on the
Unhedged podcast that the threat of a downgrade in November was probably what finally pushed
the regulator to act. So MSCI may have set the whole thing in motion, although calling a press
release a speculative attack does seem a bit of a stretch. Still, we should probably keep them on
our list of suspects. The alternative would be to start looking at the people who were running the
funds and that hardly seems fair. It would be easy to assume that a scandal like this only affects
the people who lost money in the funds, but unfortunately the damage spreads a lot further
than that. Back in 1970, the economist George Akerlof, who later won a Nobel prize, explained
why dishonesty is so harmful, using the market for used cars as an example. When buyers can't tell
the good cars from the bad ones, the lemons as he called them, they'll only pay an average price for
a used car. So the owners of good cars stop selling and the market fills up with lemons.
Turkish stocks now have a market for lemons problem. Turkey's finance minister has been
trying to get companies to raise money on the stock market rather than relying on expensive
bank loans and it was working. According to the FT, 34 companies raised 82.4 billion lira
through IPOs in the first eight months of this year, far more than in all of 2025.
But if investors can't tell which small companies are good businesses and which are propped up by
a friendly fund, they'll either stay away from the stock market altogether or demand a discount
on all of them. Honest companies end up paying a higher cost of capital in this situation.
The scandal also hits Turkish asset managers. A recent working paper found that money was already
leaving independent fund managers with no connection to the scandal and heading for
funds owned by banks. None of this is uniquely Turkish of course. Financial history is mostly
just different people discovering the same trick a few decades apart and calling it innovation.
In the 1960s, Bernie Cornfield, who happened to have been born in Istanbul, built Investors
Overseas Services into one of the biggest fund businesses in the world, selling funds
door to door across Europe. His famous recruiting line was, do you sincerely want to be rich?
A question which I imagine most people answer, I sincerely do. One of his funds owned oil and
gas exploration permits in the Canadian Arctic, which it had bought from an oil man named John
King. According to a later U.S. court ruling, in December 1969, King arranged the sale of a
small slice of his own group's Arctic permits at a much higher price. That same night, a Friday,
the fund used that price to revalue its entire holding upwards. A few weeks later, the fund
effectively reimbursed King by selling him an identical slice on the same terms. It was the
same trick, a tiny trade, a big revaluation and a price that nobody could sell at. Most American
savers avoided all of this for a fairly boring reason. Back in 1967, the SEC had forced iOS to
stop selling its funds to Americans. The institutions in that case did their job.
Then there's the example of Bill Hwang. His lawyers argued that buying shares on the open
market at whatever price the market is offering can't be a legal manipulation. It's an idea
legal scholars have debated for decades. A jury in New York disagreed and in 2024,
Hwang was convicted of fraud, market manipulation and racketeering and sentenced to 18 years in
prison. He's free on bail while he appeals and according to Bloomberg, he also has applied for
a presidential pardon. So what can the rest of the world take away from all of this? Well, one thing
is to be wary of any price that's mostly set by the people who already own large positions in the
asset, whether it's a company that keeps raising money to buy more of the same thing or a hot new
stock where only a few percent of the shares ever trade. The price on the screen can tell you very
little in these cases about what the asset would be worth if it traded freely in a deep liquid
market. And when bad actors can pump stocks without paying much of a price, honest businesses
eventually stop trying to compete. So where do things stand now? According to Turkey's
Justice Minister, 217 suspects have been targeted in the probe so far, and 56 people have been
jailed pending trial. The Capital Markets Regulator has given Isbank and the state-run
Zirat Bank up to six months to liquidate the 131 frozen funds. But as Hilmi Guvenal points out,
anyone whose sell order was cancelled when the funds were frozen is no longer owed money.
They're simply entitled to their share of whatever the banks manage to raise. And when you try and
sell billions of lira of inflated factoring companies and loss-making shipyards all at once,
you run into problems. Murat Gülkan, the CEO of OMG Capital Advisors, estimates that the assets
would fetch pennies on the dollar. So who is now stepping in as the buyer of last resort?
Well, the Turkish Sovereign Wealth Fund has reportedly been buying blue-chip stocks to
prevent the wider market from crashing. A state-owned Islamic Bank is in talks to take over
two savings finance companies from the Pusala Group. The Deposit Insurance Fund is set to
receive any assets the courts decide are the proceeds of crime. And the country's vice
president has been put in charge of the council overseeing the whole liquidation.
Which brings us back to Emery Tesman's philosophy of finance – I make money,
the investor makes money, and what does the state lose? It appears that the state is now
about to find out. It also brings us to the end of our search for the notorious malevolent forces.
After ruling out whistleblowers, regulators, New York index providers and 28-year-olds trying to
pay for a wedding, we finally have our answer. The planned, deliberate and organized speculative
attack that brought down the Goldman Sachs of Turkey was carried out by sellers. Tesman
had built a perfect financial machine, its only design flaw was that it required an economy in
which absolutely nobody ever pressed the sell button. If you find this video interesting,
you should watch my video on the crypto infinite money glitch next. Don't forget to check out our
salee using the link in the description. And see you in the next video. Bye

