AVARDA — Avarda Bank AB (publ)

Deep-value analysis — AVARDA (Avarda Bank AB (publ), formerly TF Bank AB), Nasdaq Stockholm

August 24, 2026 · reference price SEK … (close 2026-08-24, yfinance) · 64,650,000 shares · market cap SEK 12,962 mil. · reporting and trading currency: SEK (no conversion risk)

Primary sources used: Interim Report January–June 2026 (published 07/10/2026, unaudited), Interim Report January–March 2026 (04/14/2026), Interim Report January–September 2025 (TF Bank), Annual Report 2025 (TF Bank AB, audited by Öhrlings PricewaterhouseCoopers, responsible auditor Frida Main, published 03/20/2026), the July 10, 2026 investor conference call transcript, the data-pack-AVARDA-20260824.md data pack and the research-AVARDA-20260824.md research brief. No SEC filing exists: the issuer isn’t SEC-registered, it’s supervised by Finansinspektionen (Sweden). deep_data_pack.py correctly resolved the symbol collision and intentionally skipped EDGAR.

Executive summary

The thesis in short: an excellent niche bank at an exceptional-bank price, with a distributable flow three times smaller than book profit, because all the growth is consumed by regulatory capital.

Avarda Bank is the former TF Bank, renamed on April 9, 2026 (unchanged registration number, 556158-1041; the ticker switched from TFBANK to AVARDA on April 10). It’s a Swedish consumer bank with a full banking license, which has transformed over recent years from a Nordic consumer lender into a credit-card and checkout-payment issuer, with Germany as its dominant market. At June 30, 2026 the loan portfolio was SEK 26,297 mil. (… y/y), deposits SEK 28,342 mil., total assets SEK 32,737 mil., equity attributable to ordinary shareholders SEK 3,114 mil.

Operating results are, objectively, very good. Over the last 12 months (Q3 2025 – Q2 2026): operating income SEK 3,269 mil., operating expenses SEK 1,146 mil., net credit losses SEK 1,149 mil., operating profit SEK 973 mil., adjusted profit attributable to ordinary shareholders SEK 736.5 mil. (adjusted EPS SEK 11.39). The cost/income ratio fell from 37.7% (Q2 2025) to 34.6% (Q2 2026) — the best level in the listed company’s history. Adjusted ROE 24.7% in H1 2026. The risk-adjusted margin (income margin minus loss rate) is stable at 8.8… for four quarters, meaning growth isn’t being bought with worse risk, but with a better mix.

What doesn’t show in the P&L and completely changes the verdict: the bank cannot distribute what it earns. The data pack reports “simple FCF (CFO − capex) SEK 2,164 mil., yield 16.7%.” The figure is a pure accounting artifact: at a bank, operating cash flow contains the growth in deposits (+SEK 6,140 mil. over the last 12 months) minus the growth in loans (−SEK 4,626 mil.). The “free cash flow” of 2.15 billion is the difference between two financing lines, not shareholder money. The real flow to the ordinary shareholder is bounded by the CET1 capital the bank must retain to support the growth in risk-weighted assets: at the stated capital target (regulatory requirement + 2.5 percentage points, i.e. CET1 11.3%) and a 12% RWA growth rate, owner earnings are SEK 435 mil., i.e. a 3.36% yield on the current market cap — not 16.7%. The SEK 5.00/share dividend (SEK 323 mil.) already consumes 74% of this figure.

The triangulated valuation gives a wide but clearly negatively skewed range: DCF on FCFE with r=11% → SEK …/share (…); Greenwald EPV with no growth → SEK 103.6 (…); a two-stage regulatory-capital model → SEK 114.4 (…); five-year historical multiples (average P/E 13.2x, average P/B 2.88x) → SEK … (…); the bull scenario, with the cost of capital at 9.5% and 15% growth for five years, barely reaches SEK 183.4 (…). The Monte Carlo simulation across 20,000 scenarios gives a median of … and a probability of undervaluation of only …%.

Estimated value: SEK 105–185/share, central point SEK 133. The price of SEK … sits above the upper end of the central range and is supported only by the scenario where the cost of capital is ~9.5% and 15% growth holds for five years.

Verdict: AVOID at the current price. It isn’t a bad company — it’s a good company at a price that assumes everything goes right. The multiple of …x adjusted profit and 4.29x tangible book value are the highest of the five years of available history (a band of 9.3x–17.3x P/E, 1.84x–3.81x P/B). The stock rose from SEK 54.31 (end of 2023) to SEK … — a 3.7x multiplication in 32 months, of which multiple re-rating contributed more than profit growth. On top of that, balance sheet quality deteriorated exactly during the period margins improved: Stage 3 loans went from 1.37% of gross portfolio (Dec. 2024) to 7.63% (Jun. 2026), and 72% of Q2 2026’s credit losses come from a “change in provision for sold overdue receivables” line the CFO couldn’t explain live to the DNB Carnegie analyst. The level at which I’d become interested again: below SEK 145, i.e. an adjusted P/E below 13x, at the company’s own historical median.