Turkey's fund crisis: fund outflows and bank deposit growth

₺444.74 billion left the 937 funds in my TEFAS-open selection between September 16 and October 23, 2026, a net outflow equivalent to 10.46% of their starting assets. The pattern varied considerably: some smaller managers saw much larger relative withdrawals than several large bank-affiliated managers. Meanwhile, TCMB’s weekly data shows growth in bank deposit balances.

Seeing redemptions blocked can prompt investors to reassess where they keep accessible cash. Some may prefer a familiar bank-affiliated manager; others may choose a bank deposit account (mevduat). I wanted to see how well the available data fits that interpretation.

The Bloomberg story about Norway’s sovereign wealth fund prompted me to look beyond the frozen funds. I wanted to see what happened where investors could still buy and sell. The SPK dates the affected funds’ TEFAS closure to September 17, so I used September 16 as the starting reference. SPK clarification.

These are net withdrawals from the selected funds, after offsetting inflows. They do not tell us how much money left Turkey.

Withdrawals reached every fund category

Every şemsiye, or umbrella category, saw net outflows. Para Piyasası had the largest outflow relative to starting assets; Serbest had the largest lira amount.

The 42 funds under the Para Piyasası umbrella saw ₺81.41 billion in net outflows, or 22.32% of starting assets. Serbest accounted for approximately 61.5% of the selection’s total net outflow.

Money-market strategies also appear under Serbest and Katılım. Counting funds whose KAP names contain “Para Piyasası” gives 92 funds with ₺204.90 billion in net outflows, or 20.14% of starting assets. This overlapping slice cannot be added to the umbrella totals.

PPFs still had reasons to appeal

Money-market funds (para piyasası fonları, or PPFs) were a substantial place to park cash: that 92-fund slice held ₺1.017 trillion at the start, almost 24% of this selection’s assets. Daily liquidity and attractive lira returns help explain their appeal.

Some older individual holdings also retained sıfır stopaj: eligible ordinary fund units bought December 23, 2020–April 30, 2024 inclusive carried 0% withholding on redemption gains. Serbest and several other fund types were excluded. Yapı Kredi’s purchase-date rules.

Yet money-market funds saw substantial withdrawals. We cannot identify which tax lots were redeemed, but the availability of that advantage makes the broader outflow worth examining.

Take BGP, Ak Portföy Üçüncü Para Piyasası (TL) Fonu. Over the year to September 30, it returned 46.61% before investor withholding, compared with 17.89% for USD/TRY and 29.73% TÜİK CPI inflation. It is one fund example, not a category average.

Even illustrative 17.5% withholding on BGP’s gain leaves 38.46%, or 6.73% after CPI inflation. Strong past returns clearly coexisted with withdrawals: BGP itself had ₺2.53 billion in net outflows during the shorter flow window.

Which managers saw the largest relative outflows?

The chart covers only each manager’s funds in this TEFAS-open selection. Switch between percentages and lira amounts to see how the ranking changes.

Small bases matter: Destek’s result covers two funds and Trive’s seven. These percentages measure withdrawals relative to starting assets, not investment losses. Deniz had the largest lira outflow, ₺95.90 billion, but other managers saw much larger relative withdrawals.

Familiar bank names may have offered reassurance

Several large bank-affiliated managers saw milder relative withdrawals than many smaller managers. These examples are taken from the same 53-manager chart; the links document their affiliations.

Allbatross’s selected funds started with ₺7.75bn, compared with ₺543.00bn for Yapı Kredi. These examples illustrate the difference in scale as well as withdrawals. They do not form a complete classification of bank and boutique managers.

My reading is that more cautious investors may have preferred familiar bank names. The bank-affiliated managers above still had net outflows, so their figures show better retention, without establishing that money redeemed from smaller firms arrived in their funds. Client mix, strategies and distribution could also explain the differences.

There are clear exceptions. Ziraat Portföy’s selected funds saw 16.57% net outflow relative to starting assets. Its ZBJ money-market fund saw 49.98%, compared with 4.56% for BGP. Bank affiliation did not prevent large withdrawals everywhere.

KKL gives a boutique-manager example, but its strategy differs: Allbatross Portföy Kısa Vadeli Katılım Serbest (TL) Fon is a short-term participation Serbest fund for qualified investors, outside the Para Piyasası umbrella. Official fund description.

What happened to mevduat and döviz?

The report of a ₺631.3bn weekly rise in bank deposits offers another part of the picture. I checked it against TCMB’s original weekly tables and EVDS series. The increase was ₺631.29bn in October 9–16 alone, taking total deposits at domestic bank branches to ₺34.42tn. That headline includes multiple depositor groups and currencies.

For a broader view, I used September 11–October 16: a Friday before the closures through the latest published banking week available when I checked. This differs from the September 16–October 23 fund-flow window.

Over those five weeks, residents’ TL deposits rose ₺743.73bn, including a ₺318.23bn rise in household TL term deposits (vadeli mevduat). Resident FX deposits, excluding precious-metal accounts, increased $3.62bn in dollar-equivalent balance. The TL figures overlap and should not be added.

The separate vadeli mevduat series matters here: bank-branded funds and deposit accounts are different products. Growth in term-deposit balances is consistent with deposits being one alternative for investors reassessing their cash holdings. These balances also include interest and other banking activity, so the increase cannot be attributed entirely to fund redemptions.

Döviz also needs a careful definition. TCMB’s broad YP series includes precious metals. After excluding those accounts and summing TCMB’s currency-adjusted weekly changes, the increase is $5.44bn over the same five weeks. The $3.62bn balance increase above still contains exchange-rate translation effects. Neither number identifies purchases by former fund investors.

How I read the figures

My interpretation is a more selective approach to where cash is held. Some investors may have preferred larger bank-affiliated fund managers, while others may have chosen an actual bank deposit account. The milder relative withdrawals at several large managers and the growth in deposit balances fit that reading, although they cannot establish individual transfers.

There is also a practical route. Someone with money trapped in a frozen fund still has bills, living expenses or business payments to make. They may redeem another fund to get spendable cash, even if they remain confident in it.

These possibilities can overlap. The aggregate data cannot show how many investors switched products, which bank received their money, or how much was spent.