📚 Korean Financials Capital-Buyback Compounding Series — Part 1/N. Subsequent posts will track quarterly payout ratios, share-buyback-and-cancel disclosures, and the time-evolution of the ROE-PBR matrix across the cohort.

The repricing has already happened. Korean financials are no longer a “low-PBR discount asset” trade — the market now evaluates them through the matrix of ROE, payout ratio, and EPS growth. This piece is not about discovering that change. It is about what the landscape looks like after the change has happened. Meritz sits at the peak of the new standard; the rest of the cohort is following at different speeds. The peak’s price already reflects the recognition; what’s still investable is the time gap the rest of the cohort takes to catch up.

Executive Summary

• The new standard is already in place. Korean financials are no longer in the “cheap because dividends are weak” bucket. KB Financial 2025 payout 52.4%, Hana 46.8%, Meritz 61.7%. Share-buyback-and-cancel has moved from exception to standard practice. The recognition shift is no longer in progress — it is already substantially in the price.
• So the question changes. It is no longer “which financial will start returning capital?” — that question is answered. The new question is “which firm can sustain this model the longest and the most deeply?” ROE durability and capital-allocation consistency are the two variables that decide the answer.
• Meritz is the peak of that answer. ROE 22.7% (2025), EPS +13.6% growth (2026E), BPS +20.2% growth (2026E). In a year when revenue fell -24.3% and operating profit -9.9%, per-share value still compounded. The company’s identity is now capital allocation, not top line.
• PBR 1.6–1.9× already prices the recognition shift. With cost of equity 8.5% × sustainable ROE 16.5%, justified PBR sits near 1.94×; at cost of equity 10%, near 1.65×. Forward PBR 1.5–1.6× is already inside the justified range. The next leg of returns is EPS compounding under a continuing model, not multiple expansion.
• Relative time gaps are the new alpha. On the same matrix, Korea Investment Holdings (securities, ROE 16.8%, PBR <1.0×), DB Insurance (insurance, ROE 16.6%, PBR 1.0×), and Hana Financial (banking, ROE 10.5%, PBR 0.7×) are catching up to the Meritz model at different speeds. The pace at which they raise payout ratios determines the size of that time-gap alpha.

1. The Recognition Shift Has Already Happened — A New Starting Point

1.1 The Landscape Has Changed

A few years ago, the standard read on Korean financials was simple: “cheap, but capital return is weak.” Bank holdcos at PBR ~0.5×, dividend payouts in the 25% range, occasional share buybacks but rarely with cancellation. That is what made “low-PBR discount asset” a usable label.

That label no longer fits. 2025 closing-year payout ratios across the major financials look like this:

The implication is clean. Payout ratios in the 40–60% range and total yields of 5–7% are now the standard across Korean financials. The era of “the dividend story” has been replaced by the era of “ROE × capital allocation.” The market has already absorbed this shift — bank-holdco PBRs moving from 0.5× to 0.7–0.9× is the proof.

1.2 The Question Has Changed

Two consequences follow from this landscape shift.

First, the discovery era is over. “KB Financial is starting capital return” is no longer alpha. The market knows; some of it is in the price. The same applies to “Hana raising payout.” At the macro level, the recognition shift is complete.

Second, the era of speed-and-sustainability has begun. Once capital return is the standard, the operative question is straightforward: who can sustain this model the longest and most deeply? Two variables decide: ROE durability, and capital-allocation algorithmic consistency.

The company furthest along on both is Meritz Financial Holdings. This piece accepts that position as the starting condition rather than the conclusion.

2. Meritz — The Peak of the Standard

2.1 Top Line Shrinks; Per-Share Value Compounds

Meritz’s 2025 top-line numbers look weak on the surface:

Revenue            = ₩35.26T   (YoY -24.3%)
Operating profit   = ₩2.87T    (YoY -9.9%)
Net income         = ₩2.35T    (YoY +0.7%)
ROE                = 22.7%     (held vs prior year)

Arithmetic check: operating margin = 2.87 / 35.26 = 8.14%. Net margin = 2.35 / 35.26 = 6.66%. Holding ROE at 22.7% in a year when revenue falls means capital efficiency was preserved.

The real picture, however, lives in per-share metrics:

Arithmetic checks:

• 2026E EPS growth = 15,330 / 13,494 − 1 = 13.61% ≈ +13.6% ✓
• 2026E BPS growth = 72,803 / 60,553 − 1 = 20.23% ≈ +20.2% ✓
• Gap between net-income growth (+7.8%) and EPS growth (+13.6%) = 5.8 percentage points — exactly the wedge created by share-buyback-and-cancel

Single-line takeaway: net income grows ~7–8% but EPS grows ~12–14%. That gap is the accounting-level definition of “capital-buyback compounding.” When the top line is flat, per-share value still moves.

2.2 The Algorithm — Lower PER Means Higher Capital Efficiency

Meritz pays no cash dividend. It buys back stock and cancels it. The structure is not a policy preference — it is a mathematically rational capital-allocation algorithm.

Theoretical earnings yield from a buyback = 1 / PER
At PER 7.2× → 1 / 7.2 = 13.9%

That 13.9% is the headline yield Meritz cites in IR material. As long as buyback earnings yield (13.9%) exceeds cost of equity (8.5–10%), the buyback is rational. The lower the PER, the higher the yield — and the more aggressive the algorithm should be.

Outputs of the algorithm:

2025 payout ratio                = 61.7%
2025 buyback                     = ₩1.45T
2026E capital return             = ₩1.55T
2026E total yield                = 1.55 / 23 = 6.74% ≈ 6.7% ✓
2026E payout ratio               = 1.55 / 2.48 = 62.5%

Two consecutive years at 50–60% payout and 6–7% total yield is sustained delivery — and Meritz is essentially the only Korean financial that has confirmed both at this level. That is what permits “algorithm” rather than “exception” as the right word.

2.3 At the Peak of the Standard, the Type of Alpha Changes

Meritz being at the peak means the type of alpha available from this name has changed. Pre-recognition shift, the alpha was “buy what the market doesn’t see.” That stage is over. Three forms of return remain.

First — EPS compounding itself. While the model holds, EPS keeps printing +12–14% per year. Even with a flat price, forward PER falls automatically as the denominator grows; BPS rises +18–20%. The same price becomes more attractive simply as a function of time.

Second — model-durability validation as its own value variable. A company that has shown 2 consecutive years of 50%+ payout and 22%+ ROE; if it shows the same pattern for years 3 and 4, the market multiple can step up another notch. That is a different kind of alpha than top-line growth alpha.

Third — accumulated 6–7% total-yield base. Even with 0% price appreciation, the buyback-and-cancel mechanism returns roughly 6–7% per year to per-share value. Five years of that is ~35% accumulated; ten years more than 80%. That is the long-term return architecture of a “capital-buyback compounder,” distinct from a “dividend stock.”

2.4 PBR 1.6–1.9× — A Price Already Embedding the Recognition Shift

Putting that picture into the multiple:

Reference price (April 30, 2026) = ₩111,700
2026E BPS                        = ₩72,803
2026E forward PBR                = 111,700 / 72,803 = 1.534×
2027E BPS                        = ₩85,960
2027E forward PBR                = 111,700 / 85,960 = 1.299×

Arithmetic check: 111,700 / 72,803 ≈ 1.53×; 111,700 / 85,960 ≈ 1.30×. ✓

The same price prints as PBR 1.3× by 2027 — the arithmetic consequence of BPS compounding +18–20%.

Justified PBR ranges:

Read: trailing PBR ~1.8–1.9× sits within conservative-to-neutral assumptions. Forward PBR 1.5–1.6× sits comfortably even under conservative assumptions. The current price is the price after recognition, not before. Provided the model holds, it is hard to call this expensive — but the era of “discovering it cheap” is gone.

3. Different Positions on the Same Matrix — The Time-Gap Landscape

Here the post’s center of gravity shifts. If Meritz is the peak, how is the same standard reflected — at what speed and what depth — in the rest of the cohort?

3.1 The ROE × PBR Matrix

The simplest framing is an ROE-PBR scatterplot. If theoretically PBR ≈ ROE / cost of equity, then under the same cost of equity assumption, a name with double the ROE should command roughly double the PBR.

Arithmetic checks:

• Meritz: 22.4 / 1.6 = 14.0% ✓
• Korea Investment Holdings: 16.8 / 0.94 = 17.87% ≈ 17.9% ✓
• DB Insurance: 16.6 / 1.0 = 16.6% ✓
• Hana Financial: 10.5 / 0.7 = 15.0% ✓

Observation. Ranked by ROE/PBR earnings-yield proxy, Korea Investment Holdings (17.9%) > DB Insurance (16.6%) > Hana (15.0%) > Meritz (14.0%). Meritz is the peak of the standard, but ranked by raw price efficiency three names price more efficiently than Meritz — not because Meritz is overpriced, but because the same matrix makes the time-gap distribution visible.

3.2 Korea Investment Holdings — The Same Model in a Different Industry

Securities is among the slowest sub-sectors to absorb the Meritz model. Earnings volatility (trading volume, IB, prop-trading P&L) makes payout ratios harder to anchor. Still, Korea Investment Holdings prints ROE 16.8% and PBR 0.94× through that volatility.

The most important gap is the form of capital return is still dividend-heavy. That distance from the Meritz model is itself the time-gap alpha. If Korea Investment Holdings shifts the form toward share-buyback-and-cancel, EPS accretion accelerates at the same ROE, and the PBR multiple follows. That is the meaningful frame inside the post-recognition market.

3.3 DB Insurance — The Cleanest ROE-Price Alignment in Insurance

In insurance, K-ICS solvency and capital sensitivity tend to dominate the multiple before ROE. So insurance PBRs at the same ROE are less consistent than in banking or securities.

DB Insurance shows the cleanest alignment in the sub-sector:

The company’s stated intent to raise payout above 35% once K-ICS settles in the 200–220% range matters. It is the fastest-following case of the Meritz standard inside insurance. A 30% → 35% move sounds small, but at 16.6% ROE it accelerates EPS accretion meaningfully.

3.4 Hana Financial — Banking’s Price-Efficiency Leader

Bank holdcos are structurally pinned in 9–11% ROE territory. CET1 ratio sets the absolute ceiling on payout, and there is always a tradeoff between asset growth and capital return. Banking will never look “like Meritz.”

Hana’s position inside the cohort:

Arithmetic check: ROE/PBR = 10.5 / 0.7 = 15.0% — clearly above the banking-cohort average (~12–13%).

Read: Hana is the bank most aggressively following the Meritz standard inside banking. Payout ratios are now near KB and Shinhan; quarterly buyback-and-cancel is running. The ROE ceiling sits at the banking-cohort limit (~10%), so the price will not reach Meritz’s 1.6×. But the path from 0.7× toward ~1.0× normalization still exists as a time gap inside the post-recognition market.

3.5 Summary — Same Matrix, Different Clocks

This table compresses the post’s argument. The standard Meritz built is already proliferating across the cohort — at different speeds in different sub-sectors. In the post-recognition market, the meaningful difference is no longer “discovery alpha” but “speed-of-adoption alpha.”

4. Honest Limits of the Meritz Model

Maintaining a constructive tone shouldn’t mean overstating model durability. Two real limits.

4.1 Capital Sensitivity Is Not Risk-Free

Meritz is an insurance + securities composite. Meritz F&M is a non-life insurer; Meritz Securities is exposed to the full breadth of securities-industry volatility. Per Samsung Securities’ framework, a 100bp rate up move stresses Meritz F&M’s capital -10% — the largest among non-life insurers. K-ICS and rate environment can move capital strength, and capital strength is what feeds the buyback-and-cancel algorithm.

Meritz Securities also carries real-estate PF, alternative investments, and overseas asset valuation P&L. The model is powerful as a “capital-allocation algorithm,” but the capital itself is cyclical.

4.2 At Higher PER, the Algorithm’s Efficiency Falls

Buyback earnings yield = 1 / PER. PER 7.2× → 13.9%. PER 10× → 10.0%. PER 12× → 8.3%. The same capital-return budget generates less EPS accretion as the multiple rises. The model is designed to work best in the low-PER region.

The implication: as the price moves higher, both the company-level capital efficiency of buybacks and the marginal investor’s incremental upside fall. This is not a weakness as much as a self-stabilizing feature of the model — it limits how quickly the price can extend even in a constructive case.

5. Signals Worth Tracking — Post-Recognition Observation Points

Not trading triggers — observation points that show how the model evolves.

5.1 Meritz — Model-Durability Verification

• ROE stability. 22% → low-20s through the trajectory? 2026–2027E estimates show 22.4% → 21.1%. Settling above ~21% is the first-line durability check.
• Annual payout ≥ 50% defense. 2025 at 61.7%; 2026E at 62.5%. Below 50% would weaken the model claim.
• Buyback-and-cancel cadence continuity. Time gap between buyback and cancellation.

5.2 Korea Investment Holdings — Form-Transition Signals

• Buyback-and-cancel disclosures. Frequency and size — is the dividend-led form starting to shift?
• 2026–2027 ROE stability post-2025 high-base normalization, holding 16–17% range.

5.3 DB Insurance — Payout-Lift Signals

• K-ICS 200–220% stabilization — the company’s stated precondition for payout uplift.
• Stepped path toward 35%+ payout (30% → 32% → 35%) over 2026–2027.

5.4 Hana Financial — Price-Normalization Signals

• Quarterly buyback-and-cancel routinization. Execution speed of the 1H 2026 ₩400B plan, plus whether 2H plans add on.
• CET1 ratio capital-return capacity — sustaining payout near 50% while preserving capital ratios.

5.5 Cohort-Level Meta Signals

• Disappearance pace of “low-PBR discount” framing in Korean sell-side material. The further it fades, the deeper the recognition shift has anchored.
• Multiple-spread convergence or persistence across financials / securities / insurance cohorts. Convergence = standard fully diffused. Persistent spread = remaining time-gap alpha.

6. The Single-Frame Summary

The Korean financials landscape has changed. The “low-PBR discount asset” era is over; the market now reprices Korean financials through ROE × payout × EPS-growth. Meritz Financial Holdings is the peak of the new standard. Korea Investment Holdings, DB Insurance, and Hana Financial are the same standard following at different speeds across securities, insurance, and banking.

In the post-recognition market, the meaningful alpha is no longer “discovery” — it is “speed and durability.” Meritz’s frontier check is whether the model is preserved over time. Korea Investment Holdings’s check is whether form transitions from dividend toward buyback-and-cancel. DB Insurance’s check is whether payout migrates from 30% toward 35%+. Hana’s check is whether PBR normalizes from 0.7× toward ~1.0×.

And the most important single thing across the entire landscape: how capital is allocated has become the identity of Korean financials. A financial-holdings company in Korea where the top line falls but per-share value still compounds — the existence of that pattern alone is enough reason for this series to track the cohort.

The next post in the series returns when (1) Meritz quarterly payout ratio prints, (2) followers’ capital-return policy disclosures arrive, and (3) the ROE-PBR matrix updates with the next 1–2 quarters of data.

Appendix — Evidence Tier

[Fact]

• Meritz 2025A: revenue ₩35.26T (-24.3% YoY), operating profit ₩2.87T (-9.9% YoY), net income ₩2.35T (+0.7% YoY), ROE 22.7%.
• Meritz 2025 payout ratio 61.7%; 2025 buyback ₩1.45T; 2026E capital return ~₩1.55T; 2026E payout ratio ~62.5%.
• Meritz 2026E EPS ₩15,330 (+13.6% YoY); 2026E BPS ₩72,803 (+20.2% YoY); 2027E EPS ₩17,209; 2027E BPS ₩85,960.
• KB Financial 2025 payout 52.4%; Hana 46.8%; Meritz 61.7% (cohort-level standard now in 40–60% range).
• Hana Financial 2026 1H disclosed ₩400B share-buyback-and-cancel plan.
• DB Insurance has stated intent to raise payout to 35%+ once K-ICS stabilizes in the 200–220% range.

[Inference]

• Korean financial repricing from “low-PBR discount” to “ROE × payout × EPS growth” is materially complete; remaining alpha is in the speed at which the rest of the cohort adopts the standard.
• The 5.8 percentage-point gap between Meritz net-income growth (+7.8%) and EPS growth (+13.6%) is the accounting-level definition of “capital-buyback compounding.”
• Forward PBR 1.5–1.6× for Meritz sits inside justified ranges under conservative cost-of-equity / sustainable-ROE assumptions; further multiple expansion is bounded.
• ROE/PBR earnings-yield proxy ranks Korea Investment Holdings (17.9%) > DB Insurance (16.6%) > Hana Financial (15.0%) > Meritz (14.0%) — the time-gap distribution.

[Speculation]

• Korea Investment Holdings shifting capital-return form from dividend toward buyback-and-cancel would meaningfully accelerate EPS accretion at unchanged ROE.
• DB Insurance moving payout from 30% to 35%+ over 2026–2027 would tighten its discount to Meritz.
• Hana Financial PBR can normalize from 0.7× toward ~1.0× while preserving CET1 capacity.

[Blocked]

• Per-quarter payout ratios across the cohort beyond what has been disclosed.
• Specific timing of Korea Investment Holdings’s potential capital-return-form transition.
• Forward K-ICS sensitivity tables across non-life insurers needed to verify capital headroom for further payout uplift.
• Per-firm CET1 trajectory among bank holdcos under different macro scenarios.

Disclaimer: This post is research commentary, not investment advice. ROE / payout ratio / total yield / PBR scenarios are based on publicly available sell-side estimates (Samsung Securities, Kiwoom Securities, Yuanta Securities, others) and company IR materials; actual results may differ. Tickers cited are illustrative for the framework, not recommendations. Do your own due diligence and consult licensed advisors before any investment decision.

Disclaimer: For research and information purposes only. Not investment advice. Names cited are for analytical illustration; readers should perform their own due diligence and consult licensed advisors before any investment decision.