Ping An’s 2025 operating profit attributable to shareholders grew 10.3%. Yet its bank contribution fell, its life business grew only modestly, and a large part of the improvement came from smaller asset-management losses. To assess the owner’s return, first separate these businesses, then follow the cash they actually send to the listed parent.平安 2025 年归母营运利润增长 10.3%，但银行贡献下降，寿险增长相对温和，资产管理业务减亏又贡献了相当一部分增量。判断股东获得了什么，须先拆开这些业务，再追踪它们实际向上市母公司汇回多少现金。
One idea: Consolidated profit is not cash available at the holding company. Case · Ping An, 2024–2025 annual reports and 9M 2026 disclosures本期知识点：合并利润不等于控股母公司可分配的现金。真实案例 · 中国平安 2024—2025 全年及 2026 年前三季度披露
First separate annual results from nine-month results先把全年与前三季度分开
The subject is the listed insurance group (601318 / 02318), including life insurance, P&C, banking and other businesses. Annual figures below use the 2025 report’s 2024 comparatives; the nine-month comparison uses the issuer’s 2026 release. Amounts are RMB millions. Information is checked to 31 October 2026: 2026 is incomplete and the nine-month results are unaudited.研究对象是上市的中国平安保险集团（601318 / 02318），包括寿险、产险、银行及其他业务。全年数据采用 2025 年报所列的 2024 年比较数；前三季度采用公司 2026 年公告的同口径比较数。金额均为人民币百万元。资料核对截至 2026 年 10 月 31 日：2026 年尚未结束，前三季度数据未经审计。
| Metric指标 | 2024 | 2025 | Change变化 |
|---|---|---|---|
| Attributable net profit归母净利润 | 126,607 | 134,778 | +6.5% |
| Attributable operating profit归母营运利润 | 121,862 | 134,415 | +10.3% |
| L&H new business value寿险及健康险新业务价值 | 28,534 | 36,897 | +29.3% |
| P&C combined ratio产险综合成本率 | 98.3% | 96.8% | −1.5 pp |
| Dividend per share · RMB每股股息 · 元 | 2.55 | 2.70 | +5.9% |
2025 annual report: printed pp. 36–40, 50–53 and 86; the dividend includes the proposed final payment at the report date.2025 年报：印刷页码第 36—40、50—53、86 页；年度股息包括报告日尚待批准的末期股息。
| Metric指标 | 2025 9M | 2026 9M | Change变化 |
|---|---|---|---|
| Attributable net profit归母净利润 | 132,856 | 96,645 | −27.3% |
| Attributable operating profit归母营运利润 | 116,264 | 125,905 | +8.3% |
| L&H new business value寿险及健康险新业务价值 | 35,724 | 30,987 | −13.3% |
| P&C combined ratio产险综合成本率 | 97.0% | 95.6% | −1.4 pp |
Ping An’s issuer-distributed 28 October 2026 release, §1.2 Key Figures (EQS); all comparisons are nine months.平安于 2026 年 10 月 28 日通过 EQS 发布的公司公告，第 1.2 节关键数据；比较期间均为前三季度。
The 2026 picture is mixed: the adjusted operating measure rises, reported profit falls, and life NBV declines despite better P&C underwriting. Neither the 2025 NBV growth rate nor the 2026 operating-profit growth rate describes the whole group. Do not multiply nine-month profit by four-thirds: investment markets and insurance sales are not evenly distributed through a year.2026 年的信号并不一致：调整后的营运利润增长，报表净利润下降，产险承保改善，寿险新业务价值却下降。2025 年的新业务价值增速或 2026 年的营运利润增速，都不能单独代表整个集团。也不宜将前三季度利润简单乘以 4/3 来预测全年：投资行情与保险销售并非均匀发生。
Locate the source of profit growth找出利润增量来自哪里
| Business业务 | 2024 | 2025 | Difference增减额 |
|---|---|---|---|
| Life and health寿险及健康险 | 96,975 | 99,752 | +2,777 |
| P&C财产保险 | 14,952 | 16,923 | +1,971 |
| Banking银行 | 25,796 | 24,711 | −1,085 |
| Asset management资产管理 | −11,899 | −3,785 | +8,114 |
| Group total, including other businesses and eliminations集团合计，含其他业务与抵销 | 121,862 | 134,415 | +12,553 |
2025 annual report, printed pp. 37–39: attributable segment operating profit, after non-controlling interests.2025 年报第 37—39 页：按归属于母公司股东口径列示的分部营运利润，已区分少数股东权益。
Asset management’s smaller loss explains approximately 8,114 ÷ 12,553 = 64.6% of the group’s operating-profit increase. That is helpful, but it cannot recur indefinitely: once a loss reaches zero, growth needs another source. Life and health operating profit rises only 2.9%, and banking declines 4.2%. Judge turnaround progress separately from the growth of already profitable businesses.资产管理减亏对集团营运利润增量的贡献约为 8,114 ÷ 12,553 = 64.6%。减亏当然有价值，但不能无限重复：亏损归零以后，进一步增长须有新的来源。寿险及健康险营运利润仅增长 2.9%，银行下降 4.2%。应把扭亏进展与原有盈利业务的增长分开评价。
Operating profit is a management-defined measure, not another name for cash earnings. For the relevant non-VFA life business, the 2025 methodology replaces short-term investment variation with a 4.0% long-run return assumption and removes specified non-operating items. Read the reconciliation and challenge the assumed return; a persistent reinvestment shortfall does not become harmless because an adjusted measure excludes it.营运利润是公司定义的经营评价指标，不是现金利润的别称。2025 年对相关的非浮动收费法寿险业务，公司按照 4.0% 的长期投资回报假设剔除短期投资波动，并调整特定非营运事项。读者需要核对调节表，也需要检验这个回报假设；如果再投资收益长期不足，不能因为调整口径暂时将其剔除，就认为经济损失消失了。
2025 annual report, printed pp. 37–38 and 79–80: definition and reconciliation of operating profit.2025 年报第 37—38、79—80 页：营运利润的定义与调节表。
Follow the cash to the listed parent追踪现金是否到了上市母公司
A controlled bank’s assets and earnings enter the consolidated accounts, but its cash remains inside a regulated legal entity. The bank must meet depositor obligations and capital requirements before paying dividends. The same distinction applies to insurance subsidiaries and policyholder liabilities. The parent’s free-cash reconciliation is therefore more informative for distributions than group operating cash flow less equipment purchases.控股银行的资产和利润进入合并报表，但现金仍属于受监管的独立法人。银行先要满足存款支付与资本要求，才能分红；保险子公司也要先履行保单责任。分析股东分配时，母公司自由现金变动表比“集团经营现金流减购建设备支出”更有解释力。
What happened to the parent’s available funds in 2025?2025 年，母公司可动用资金去了哪里？
- 74,565Opening free cash期初自由现金
- +53,022Dividends received from subsidiaries收到子公司分红
- −46,537Cash dividends paid to shareholders during the year年内实际支付股东分红
- −19,983Investment in subsidiaries对子公司投资
- +10,686Net convertible-bond proceeds发行可转债净收入
- −3,594Other movements其他变动
- =68,159Closing free cash期末自由现金
The subsidiaries paid the parent 53,022. After the year’s 46,537 shareholder payment, 6,485 remained before subsidiary investment and other movements. The bank’s cash dividend was 5,752, far below its 24,711 attributable operating-profit contribution: recognition of earnings and receipt of cash are different events. This is a timing and capital-allocation distinction, not proof that the remainder was lost.子公司向母公司分红 53,022，减去年内实际付给股东的 46,537，尚余 6,485，之后还要考虑对子公司投资等用途。其中银行实际汇入的股息为 5,752，明显少于其 24,711 的归母营运利润贡献：确认利润与收到现金是两件事。这反映了分配时点和资本配置的差别，不能据此认定其余利润已经损失。
The 46,537 paid during calendar 2025 is also different from the 48,891 dividend attributed to FY2025, which includes a final payment proposed after year-end. Use the former to reconcile cash movements and the latter to discuss the fiscal-year payout policy; mixing them creates a false cash gap.2025 日历年实际支付的 46,537，与归属于 2025 年度、包含年后拟派末期股息的 48,891，也不是同一个期间口径。核对现金变动使用前者，讨论年度分配政策使用后者；混用会算出并不存在的现金差额。
2025 annual report, printed pp. 85–86: upstream dividends, parent cash and fiscal-year dividends.2025 年报第 85—86 页：子公司分红、母公司自由现金与年度股息。
Test whether distributions leave enough capital分红之后，还剩多少风险承受能力
At 2025 year-end, the group’s comprehensive solvency ratio was 193.3%, down from 204.1%; Ping An Life’s was 175.7%, down from 189.2%. These are different consolidation scopes. The bank uses a separate capital regime: its core tier-one capital ratio was 9.36%. A group ratio cannot authorize the transfer of cash from any particular subsidiary.2025 年末，集团综合偿付能力充足率由 204.1% 降至 193.3%，平安寿险则由 189.2% 降至 175.7%，两者合并范围不同。银行另用资本监管体系，核心一级资本充足率为 9.36%。集团总体达标，不意味着任意一家子公司的现金都可以自由上划。
2025 annual report, five-year highlights and printed pp. 46, 87: subsidiary capital and group solvency.2025 年报五年摘要及第 46、87 页：子公司资本与集团偿付能力。
The owner question is whether additional capital retained in each business earns an acceptable risk-adjusted return. Munger’s counter-test is to ask what would break the payout story: weaker upstream dividends, repeated capital injections, persistent investment shortfalls or cash distributions increasingly financed by borrowing. A convertible-bond inflow is financing, not evidence of improved underwriting.从所有者角度，要问留在各业务中的新增资本，能否取得与风险相称的回报。反过来检验，则要观察什么会破坏稳定分红的判断：子公司上划减少、持续注资、投资收益长期不足，或者分配越来越依赖融资。可转债流入是融资来源，不能证明承保能力改善。
Uneven business recovery; dividends depend on upstream cash集团改善不均，分红取决于现金上划
Ping An’s improvement is real but uneven. The evidence supports a recovering financial group, not uniform double-digit growth across its businesses. P&C underwriting improved and asset-management losses narrowed, while banking and the 2026 life NBV figures require separate scrutiny.平安的改善真实存在，但分布并不均匀。现有证据支持“多项业务修复中的金融集团”，尚不足以认定各业务都已进入稳定的两位数增长。产险承保改善、资产管理减亏值得肯定，银行表现与 2026 年寿险新业务价值则需要单独检验。
Owner earnings start with cash that can leave the subsidiaries. In RMB millions, the parent received 53,022 in subsidiary dividends and paid shareholders 46,537 during 2025. That leaves 6,485 before subsidiary investment and other movements; it is not all available for an additional payout. A defensible distribution estimate starts with sustainable upstream dividends, then allows for necessary parent expenses, financing obligations and capital injections. The group’s 134,415 operating profit is useful evidence about business performance, but it is not the starting cash balance for that calculation.所有者收益，要从子公司能够汇出的现金算起。以下金额均为人民币百万元。2025 年母公司收到子公司分红 53,022，向股东实际支付 46,537，两者相减剩下 6,485，尚未考虑对子公司投资与其他用途，不能全部视为额外可分配资金。估计未来分配能力，应从可持续的子公司分红出发，再考虑必要的母公司费用、融资义务和注资需求。集团营运利润 134,415 有助于判断经营表现，却不能直接充当这项计算的现金余额。
Capital allocation determines whether retained earnings help the owner. The bank’s smaller cash transfer than its attributable operating-profit contribution does not by itself indicate a loss: retained capital may support future earnings. It earns that justification only if incremental returns compensate for the risk and later improve distributable capacity. Likewise, asset-management loss reduction explains 64.6% of the group’s 2025 operating-profit increase. Treat the reduction as progress, but do not compound that one-year contribution indefinitely or assume further injections will earn adequate returns without evidence.留下来的利润是否有价值，要看资本配置的结果。银行汇入的现金少于其归母营运利润贡献，本身不说明资金损失；留存资本可能支撑未来盈利。但只有新增回报足以补偿风险，并最终改善可分配能力，留存才得到经济上的支持。同样，资产管理减亏贡献了 2025 年集团营运利润增量的 64.6%。可以肯定修复进展，却不能把这一年的贡献率一直外推，也不能在缺少证据时认定后续注资一定划算。
2025 annual report, printed pp. 37–39 and 85–87: segment profit, parent cash and capital constraints; the differences and 64.6% contribution are calculations in this article.2025 年报第 37—39、85—87 页：分部利润、母公司现金与资本约束；差额及 64.6% 的贡献比例由本文计算。
Inversion makes the distribution thesis falsifiable. A claim of improving owner returns weakens if upstream cash falls while recurring capital injections rise, or if distributions increasingly depend on new borrowing. Those conditions would matter even with higher adjusted profit. Incentives belong in this test: if remuneration rewards premium volume or adjusted earnings without charging for the capital and risk they consume, growth can benefit the reported metric before it benefits the owner. That is a risk to check against disclosed remuneration and capital-allocation outcomes, not a finding about Ping An’s actual incentive arrangements.逆向检验，要写出什么情况会推翻分配能力改善的判断。如果子公司上划现金下降、反复注资增加，或者分红越来越依赖新增借款，即使调整后利润仍增长，也应下调对股东回报质量的评价。激励机制也要纳入检验：如果考核奖励保费规模或调整后利润，却不充分计入资本消耗和风险，增长可能先改善考核指标，未必改善股东收益。这是需要对照薪酬披露和资本配置结果核查的风险，不是本文对平安实际考核机制已经作出的认定。
Use peers to test the explanation at the same business level. Compare Ping An Life with China Life and New China Life on new-business economics, persistency and capital strain under compatible actuarial assumptions; compare Ping An P&C with CPIC P&C and PICC P&C on underwriting and reserve development; compare the bank with banks. A strong subsidiary cannot erase a weak one merely because their profits consolidate into the same total. This business-level comparison is the appropriate basis for judging the group’s diversification.横向检验必须落在同一业务层面。平安寿险应与中国人寿、新华保险在可比精算假设下比较新单经济性、持续率与资本消耗；平安产险应与太保产险、人保财险比较承保和准备金发展；银行则与银行比较。优势业务不能因为合并报表汇总，就自动抵销弱项业务的经济问题。按业务逐项比较，才能判断集团的多元化究竟创造了多少价值。
Actionable conclusion: A holder can test a distribution-based thesis by updating three reconciled records at each reporting date: subsidiary earnings and cash transfers, capital retained or injected, and parent cash after distributions and financing. Keep the thesis only while the evidence supports sustainable transfers without progressively weaker capital coverage. For a new valuation, use a range of sustainable parent distributions and explicit capital needs rather than extrapolating the 10.3% operating-profit growth rate. The disclosures justify continued research into cash generation; they do not yet establish that the shares offer an adequate return at any particular price.可执行的结论：若持有逻辑是持续分红，每次披露后就更新三组能勾稽的记录：子公司利润与现金上划、留存资本与新增注资、分红及融资后的母公司现金。只有现金上划具有持续性、资本覆盖又没有不断削弱，这个持有逻辑才继续成立。进行估值时，应使用可持续母公司分配的区间，并明确资本需求，不能直接外推 10.3% 的营运利润增速。现有披露足以支持继续研究现金创造能力，但尚未证明股票在某个价格上具有足够回报。
Exercise: distinguish payout coverage from capital needs练习：分红有覆盖，是否意味着资金宽裕
Calculate subsidiary-dividend coverage of the parent’s cash dividends in 2025. Then deduct investment in subsidiaries. Does the initial coverage prove that the parent’s cash balance should rise?计算 2025 年子公司分红对母公司实际支付的股东现金分红的覆盖倍数，再扣除对子公司的投资。最初的覆盖倍数，能否证明母公司现金余额应当增加？
Check the reasoning核对推理
53,022 ÷ 46,537 ≈ 1.14 times. But 53,022 − 46,537 − 19,983 = −13,498 before financing and other movements. Add convertible proceeds of 10,686 and other outflows of 3,594: the net reduction is 6,406, matching 74,565 − 68,159. This does not establish that debt directly funded a particular dividend; it shows why distributions, reinvestment and financing must be considered together.53,022 ÷ 46,537 ≈ 1.14 倍。但 53,022 − 46,537 − 19,983 = −13,498，尚未计入融资与其他变动。加上可转债净收入 10,686，再减其他流出 3,594，资金净减少 6,406，恰好等于 74,565 − 68,159。这不能证明某笔债务直接支付了某笔股息，却说明必须把分配、再投资与融资放在一起判断。
Sources and reading route资料与后续阅读
Ping An 2025 annual report: printed pp. 36–40, 79–87. Annual comparisons, segment reconciliation, cash transfers and regulatory capital.中国平安 2025 年报，第 36—40、79—87 页：年度比较、分部勾稽、现金上划与监管资本。
Ping An 9M 2026 issuer release via EQS: §1.2, reported and operating profit, NBV and P&C combined ratio.中国平安 2026 年前三季度公司公告（EQS 发布），第 1.2 节：净利润、营运利润、新业务价值与产险综合成本率。
Ping An corporate 9M 2026 results page: operating progress and bank capital indicators.中国平安官网 2026 年前三季度业绩：营运进展与银行资本指标。
Next, China Life shows why an insurer’s investment performance must be read alongside its liabilities.下一期借助中国人寿的资产负债案例，理解为什么保险投资收益必须与负债一起读。
