Where banks sit in the Turkish financial system

The lecture now leaves the general institution types and examines a particular financial system. The cover identifies the 2026–2027 Management of Financial Institutions course, Assoc. Prof. Dr. R. Can Akkay, Istanbul University, and source material labelled BRSA March 2026 and KPMG 2025. Treat these as the lecture’s dated observations, not a live dashboard. A figure about March 2026, an end-2025 ratio and a 2024 digital-customer count do not become observations from the same date just because they appear on one slide.

The system map separates four supervisory branches. Under BDDK / BRSA the bank box contains deposit banks, participation banks, and development and investment banks. Deposit banks collect deposits and lend. Participation banks collect participation funds on an interest-free basis. Development and investment banks do not accept deposits in this diagram. The outlined deposit-taking area includes only the first two. A financial institution can therefore provide finance without funding itself through public deposits.

Read the other BDDK boxes by the financial problem they solve. Leasing companies finance tangible assets: the source distinguishes the lessor’s ownership from the lessee’s risks and rewards. Factoring companies advance cash against unpaid invoices and provide collection services, bringing forward money otherwise received later. Finance companies lend to individuals and firms but do not take deposits. Savings finance companies are marked as falling under Law No. 6361 since 2021. Financial holding companies hold a bank and non-bank affiliates. Asset management companies in this branch buy banks’ non-performing receivables. That last activity is different from managing an investor’s securities portfolio.

Under SEDDK / IPRSA, insurance and reinsurance companies pool and price risk for premiums and invest premium income in long-term assets; private pension companies operate the private pension system, including BES and auto-enrolment. Under SPK / CMB, brokerage houses intermediate securities and derivatives trading and underwrite new issues; portfolio management companies manage investment funds and discretionary portfolios; investment funds and trusts include mutual funds, real-estate investment trusts (GYO), and venture-capital trusts (GSYO). Under TCMB / CBRT, payment and e-money institutions process payments and issue e-money; the source expressly says they cannot take deposits or extend credit.

The common thread is function, funding source, and claim held by the customer. A payment balance, a pension contribution, a mutual-fund share and a bank deposit are not interchangeable contracts. Likewise, “asset management company” in the distressed-receivables branch and “portfolio management company” in the capital-markets branch are not the same box. The diagram is a teaching map of institutional functions, not a complete statement of overlapping legal powers or current regulatory requirements.

Banks, assets and concentration (March 2026)

The next slide describes the system’s scale in three different ways: how many institutions exist, how large their combined balance sheet is, and how unevenly that balance sheet is distributed. These are different questions. The source gives 68 banks and TL 49.7 trillion total assets. A count does not tell you that every bank is equally large; the concentration figure immediately shows why that would be misleading.

The activity and ownership rows classify the same 68 banks along different dimensions; adding the two sets would double-count institutions. Likewise, total assets are a stock measured at a date, whereas GDP is a flow over a period. An assets/GDP ratio of 0.74 means assets are 74% of the stated annual GDP base; it does not mean banking produced 74% of GDP.

Approximately 43% of assets in three banks indicates concentration. The handwritten comment beneath it appears to question the risk-management implications. A careful interpretation is that the distribution matters: a disruption at a very large bank can affect many claims and customers at once. It does not follow from concentration alone that these named banks are unsafe, or that concentration proves poor risk management. Portfolio quality, capital, liquidity and interconnections would also be needed.

The ownership-group loan figures sum to TL 25.0 tn, whereas the printed total is TL 24.9 tn. Treat the one-decimal figures as rounded, not as an exact reconciliation. Public banks hold the largest stated loan amount. Branches, personnel and digital customer counts describe delivery channels, but the digital count is not necessarily a count of distinct people across all institutions: the source does not provide a deduplication rule. The defensible lesson is the system’s large scale and digital reach, not an invented population share.

Sector balance sheet (March 2026)

A bank’s balance sheet is easier to understand if you read it as two questions. The asset side asks where has the funding been used? The liabilities-and-equity side asks where did the funding come from? A customer’s loan is the bank’s asset because the bank has a claim on repayment. A customer’s deposit is the bank’s liability because the customer has a claim on the bank. Keeping that perspective prevents the most common reversal.

The largest use is lending and the largest source is deposits. Securities on the asset side are investments, such as the bonds identified by the handwritten note. Securities issued on the other side are instruments the bank itself sells to obtain funding. Similarly, due to banks means money the bank owes to other banks, not loans it has made to them. Repo funding raises funds against securities under a repurchase arrangement. Required reserves and cash are not interchangeable categories simply because both may support settlement or regulatory needs.

The handwritten equity reminder points to the owners’ funding. Equity is a residual claim and a loss buffer, not an amount contractually owed to depositors. The accounting relationship is A = L + E. The source’s right-hand percentages use total liabilities and equity as the balancing total. Its later description of deposits as “57% of liabilities” is therefore best read with this fuller denominator in mind, rather than excluding equity and still expecting the same percentage.

Turn a percentage into an amount without inventing precision

Using total assets of TL 49.7 tn, the loan bar implies 49.7 × 0.50 = TL 24.85 tn, which rounds to the separately reported 24.9 tn. Deposits imply 49.7 × 0.57 = TL 28.329 tn, close to 28.3 tn. Equity implies 49.7 × 0.09 = TL 4.473 tn. These are calculations from rounded graph shares, not independently measured precise balances. In particular, the 9% equity share is E/A; it is not the capital adequacy ratio, whose denominator later becomes risk-weighted assets.

Deposits by maturity (March 2026)

The lecture next asks how long the bank can rely on the funding it has obtained. Contractual deposit maturity is not simply a descriptive customer statistic: it determines when the institution may have to repay, renew or reprice a funding contract. Demand deposits can be withdrawn without waiting for a fixed maturity. A one-month deposit provides a different contractual promise from a three-year business loan, even if both initially equal the same amount of money.

The prominent “approximately 90%” is reproducible: 38 + 19 + 33 = 90%. Demand deposits belong in this short-funding grouping because they are immediately accessible, not because they literally mature after three months. The remaining buckets give 6 + 2 + 2 = 10%; all six sum to 100%. Only 2% is at least one year. The continuation on PDF 9 reports total deposits and participation funds of TL 28.3 tn, compared with TL 24.9 tn of loans.

Why should a manager care? If a bank promises short access to depositors while holding longer, illiquid loans, it must manage the timing difference. At renewal, the interest rate needed to retain funding may rise while a fixed-rate loan continues earning the old rate. If depositors withdraw, the bank needs cash before its loans repay. The same funding structure can therefore create interest-rate risk and liquidity risk through different mechanisms.

Do not calculate a precise asset–liability maturity gap from this chart alone. The source has not supplied the matching distribution of loan cash flows. Nor does the contractual maturity distribution tell you exactly how long customers actually keep rolling their deposits over. Behavioural stability can differ from contract maturity. The disciplined conclusion is that the displayed funding is contractually short; the size of an actual mismatch needs the asset side and a stated behavioural assumption.

Net income and returns

The source first shows net income amounts, then return ratios. An amount answers “how much profit?” A ratio asks “how much profit relative to the resources used?” Neither can replace the other. The plotted amounts are 216, 206, 248, 270 and 289 billion TL for March 2025, June 2025, September 2025, December 2025 and March 2026 respectively. The chart calls them “quarterly cumulative”; because the March-to-June value falls, it should not be silently treated as an ordinary year-to-date accumulating series without more explanation.

The latest bar is 289, and the ownership amounts—public 106, private 101, foreign 82 billion TL—sum to 289. Yet the bottom banner says Q 1 2026 net profit reached 288 billion TL. Both are visible in the teacher’s source. We preserve the discrepancy rather than selecting one as secretly authoritative. It could be a rounding, timing or drafting difference; the supplied slide does not settle the cause.

The handwritten expansions say return on assets and return on equity. The displayed sector ROA is 3.1% at March 2026. ROE is 28.0%, with comparators 30.5% in 2024 and 42.7% in 2023. The source does not give all averaging and annualisation conventions, so these are not instructions to divide one quarterly bar by a single date’s assets and expect precisely the published ratio.

For the same income, period and consistent denominators, ROE = ROA × A/E. Equity is smaller than assets when debt finances part of the balance sheet, so a given asset return can produce a larger equity return. That is leverage, not free efficiency. Leverage also magnifies losses relative to owners’ capital. For example, the earlier 9% equity share implies A/E ≈ 11.11, but multiplying it by this ROA need not reproduce 28% because the source’s ratio definitions may use different timing or average bases. The useful habit is to identify the numerator, denominator and period before comparing performance.

Foreign-currency share and FX position (March 2026)

A foreign-currency share tells you how much of a category is denominated in foreign currency. A net foreign-currency position asks whether foreign-currency claims and obligations offset one another. The source gives FX shares of 39% of assets, 43% of liabilities, 36% of loans and 41% of deposits. These percentages do not all have the same denominator, so subtracting 39% from 43% does not by itself reproduce the currency position in lira.

The balance-sheet FX position is a deficit of TL 1,685 billion, while the off-balance-sheet position is a surplus of TL 1,702 billion. A derivative contract can create an offsetting currency exposure without appearing as the same gross asset or liability as a conventional loan. The lesson’s near-flat net label is approximately +TL 16 billion. The displayed rounded inputs give −1,685 + 1,702 = +TL 17 billion. Keep both results visible: the underlying unrounded dataset is not supplied, so the one-billion difference cannot be definitively reconciled here.

The small sector net does not mean every bank has the same exposure. Positive and negative ownership-group positions may offset in an aggregate. Nor does a close currency hedge remove every risk: the two contracts may have different maturities, collateral requirements, counterparties or access to cash. Net currency direction and funding liquidity are different questions.

The slide’s footer continues at the very top of PDF 10, before the securities heading. It reports due to banks: TRY TL 0.7 tn / FX TL 4.6 tn; and securities issued: TRY TL 0.1 tn / FX TL 1.8 tn. All four are lira-equivalent amounts; TRY versus FX identifies the denomination, not different units being added. These two funding channels are heavily foreign-currency denominated in the source. They help explain why a bank must consider the currencies of its financing as well as the currencies of its loans. Their rounded totals need not exactly match amounts inferred from an earlier rounded percentage chart.

Securities portfolio (March 2026)

The asset-side securities bar is now opened into its components. The chart title rounds the portfolio to TL 7.5 tn; the accompanying number is TL 7.47 tn, compared with TL 5.77 tn in March 2025. These are a rounded and a more precise description of the same current portfolio, not two amounts to add. Public banks hold TL 3.9 tn, the largest stated ownership-group book.

The four shares sum to 100%. A large government-related component means many institutions may be exposed to common public-sector interest-rate or credit factors, but the categories are not economically identical. A domestic-currency government bond and a foreign-currency Treasury eurobond can respond differently to exchange rates and funding conditions. Their maturity and coupon structures also matter. The source does not supply enough detail to calculate the portfolio’s price sensitivity from these shares alone.

Sukuk is a separate source category, not a reason to assume the same contract as every conventional bond. What the lecture needs here is the classification and the concentration of securities holdings. The next risk chapters will explain why a security can lose market value even when its issuer keeps making the promised payments: changing market yields alter the present value of those payments. Thus “a security rather than a loan” does not mean “no risk.”

NPL ratios by segment, March 2025–March 2026

A non-performing loan (NPL) is a loan classified as not performing as required; the lecture uses the label to discuss credit quality. The NPL ratio compares the non-performing loan stock with the total loan stock. It is therefore a ratio of two balance-sheet amounts, not the probability that every new borrower will default and not the amount of loss that will certainly occur.

These are the explicitly written footer values. Some plotted line endpoints appear higher than these printed numbers. The chart’s intermediate dates—June, September and December 2025—show a rising pattern but do not give exact point labels. We use the explicit numbers for arithmetic and do not manufacture precise quarterly data by estimating pixels. The handwritten definitions identify SME as small and medium-sized enterprises, commercial lending here as lending to large companies, and “sector” as the banking-sector aggregate.

For the sector, 2.65 − 1.95 = 0.70 percentage points. Relative to the old ratio, 0.70 / 1.95 × 100 ≈ 35.9%. These are different descriptions of the same ratio change. Neither tells us that the amount of non-performing loans rose by exactly 35.9%, because total loans in the denominator may also have changed. To calculate amount growth, both dates’ loan totals would be required.

The source also reports approximately 75% provision coverage. Provisions recognise expected or identified credit losses under the applicable accounting framework; coverage compares the stated provision buffer with the problem-loan base. A 75% coverage ratio does not mean that 75% of every loan will be collected, or that the final loss must be 25%. Recoveries, collateral, timing and further deterioration matter. The causal link to the bank is that impaired loans can reduce interest receipts, require loss recognition, lower profit and ultimately consume equity.

Capital adequacy, liquidity coverage and risk-weighted assets

The lecture now asks whether banks have buffers against adverse outcomes. There are two distinct buffers here. Capital absorbs losses that reduce net worth. Liquidity allows obligations to be paid when cash is needed. An institution can have valuable assets but still struggle to obtain cash on time, so a high capital ratio is not a substitute for a liquidity measure.

Risk-weighted assets (RWA) are not simply total accounting assets. The denominator reflects the capital framework’s risk measures. Regulatory capital can also differ from accounting equity. A CAR of 16.5% therefore means 16.5 units of regulatory capital per 100 units of RWA, not 16.5 units of cash for every 100 deposits. By ownership, the source gives CAR of 15.8% public, 16.4% private and 17.9% foreign. RWA composition is 87% credit risk, 10% operational risk, 3% market risk. These are shares of the measured risk-weighted base, not probabilities of loss or a forecast that exactly 87% of future losses will be credit losses.

HQLA means high-quality liquid assets. Total LCR of 142% expresses a stock of qualifying liquid assets equal to 1.42 times the modelled 30-day stressed net outflow. FX LCR of 280% expresses 2.8 times the corresponding foreign-currency outflow. The total and FX measures have different scopes and cannot be added into a 422% “combined buffer.” Nor do they promise survival under every imaginable stress.

The footer matters: CAR is for the banking sector; LCR is for deposit banks, both dated March 2026. This is a concrete scope difference, not merely a generic warning to “check denominators.” The 8%, 100% and 80% thresholds are those printed as legal minima in the lecture. They are retained for studying this source, not independently verified as the complete current set of regulatory requirements.

Strengths & weaknesses

The final sector slide places favourable buffers beside unfavourable trends. This is not a contradiction. A bank can begin with substantial capital while experiencing deteriorating credit quality; one variable describes present capacity and the other a pressure on that capacity. Read each SWOT point as a causal claim that should connect to an earlier chart.

The source lists six strengths: capital structure with CAR 16.5% above its labelled minimum; liquidity buffers with LCR above the shown limits; experience operating through turbulent environments; foreign participation and a successful track record; technology and digital banking, referring to more than 119 million digital customers; and regulation supporting the sector in stress periods. The first two are linked to ratios. The others are qualitative judgements in the lecture rather than independently measured guarantees of future performance.

The six weaknesses are macroeconomic uncertainty; rising NPLs, including sector 2.65% and consumer 4.27%; pressure on profitability, with ROE falling from 42.7% to approximately 28% and labelled below inflation; high funding costs squeezing margins; credit quality sensitive to macroeconomic turbulence; and geopolitical tensions raising risk pricing. The source does not provide a matched inflation series, so we cannot calculate an exact real ROE from this slide alone.

Connect the mechanisms rather than memorising two columns. Short-maturity deposits can make elevated funding costs transmit quickly to interest expense. A weaker economy can impair borrowers’ cash flows, increasing credit losses and lowering profits. Geopolitical risk can raise the required return on funding or securities. Digital systems can lower unit operating cost while also making technology resilience important. Strong capital and liquidity may help absorb these pressures, but they do not make the pressures disappear. This is the bridge from describing the sector to managing the risks in the next chapter.