2026-02-28 Greg Abel VS Warren Buffett

2026-02-28 Greg Abel VS Warren Buffett

Greg Abel VS Warren Buffett

1、同一个主题的表述

有一段针对保险业务的评论可以测试两个人在思维方式上的差异。

(1)Greg Abel

Quote
Reinsurance Group

Our reinsurance operations face similar dynamics. The reinsurance sector has attracted significant increases in available capital from both the traditional and alternative markets, which together with a more benign reinsured catastrophe loss burden in 2025 in most major regions has led to significant price declines in property reinsurance. In most casualty reinsurance segments, claims inflation continued to outpace pricing. As long as these phases of the cycle endure, we expect to write less reinsurance premium.
我们的 Reinsurance Group 再保险业务面临类似的动态。再保险领域吸引了来自传统和替代市场的大量可用资本增加,加之 2025 年大多数主要地区的再保险巨灾损失负担较轻,导致财产再保险价格大幅下降。在大多数责任险再保险领域,赔付通胀继续超过定价速度。只要周期的这些阶段持续下去,我们预计将减少承保的再保险保费。

Our insurance team will remain patient because of Berkshire’s structural strengths:
我们的保险团队将保持耐心,因为 Berkshire 拥有结构性优势:

1. We have significant capital, enabling us to underwrite large and unusual risks.
我们拥有雄厚的资本,使我们能够承保大型和不寻常的风险。

2. We give our insurance managers autonomy to run their businesses, without quarterly earnings targets or growth mandates that might otherwise distort their underwriting judgment.
我们给予保险经理人经营业务的自主权,没有季度盈利目标或增长指令,否则这些可能会扭曲他们的承保判断。

3. We insist on underwriting discipline as the most important ingredient in insurance success.
我们坚持将承保纪律视为保险成功的核心要素。

4. We focus on the long term, resisting temporary industry enthusiasms and exuberances.
我们专注于长期,抵制行业暂时的狂热和过激行为。

The environment ahead will reward insurers whose focus remains on growing underwriting profit sustainably, not volume; customer trust and loyalty, not temporary spikes in market share; and long-term resilience, not short-lived opportunism.
未来的环境将奖励那些专注于可持续增长承保利润而非规模、注重客户信任和忠诚度而非市场份额的暂时飙升、以及注重长期韧性而非短期投机主义的保险公司。

(2)Warren Buffett

Quote
At bottom, a sound insurance operation requires four disciplines: (1) An understanding of all exposures that might cause a policy to incur losses; (2) A conservative evaluation of the likelihood of any exposure actually causing a loss and the probable cost if it does; (3) The setting of a premium that will deliver a profit, on average, after both prospective loss costs and operating expenses are covered; and (4) The willingness to walk away if the appropriate premium can’t be obtained.
归根到底,一家健康稳健的保险公司必须严格遵守四条纪律:(1)充分理解所有可能导致保单发生赔损的风险敞口;(2)以保守的态度评估这些风险敞口实际引发损失的可能性,以及一旦发生损失时的可能成本;(3)设定一个在平均意义上可以保证盈利的保费水平,也就是说,在覆盖预期赔付成本和运营费用之后仍能留有利润;(4)当无法获得足够的保费时,有毅力、有意愿直接放弃这笔业务。

Many insurers pass the first three tests and flunk the fourth. The urgings of Wall Street, pressures from the agency force and brokers, or simply a refusal by a testosterone-driven CEO to accept shrinking volumes has led too many insurers to write business at inadequate prices. “The other guy is doing it so we must as well” spells trouble in any business, but none more so than insurance.
许多保险公司在前三条纪律上都能勉强过关,却在第四条上彻底不及格。华尔街的鼓噪、代理人和经纪团队的压力,或者只是某些“荷尔蒙爆棚”的 CEO 不愿意接受业务量下滑的现实,这些因素共同推动了太多保险公司在保费严重不足的情况下仍然硬着头皮承保业务。“别人都这么干,所以我们也得这么干”——这种心态在任何行业都会带来麻烦,而在保险业则尤为致命。

2、Claude的回答

用这个例子分别测试了ChatGPT和Claude,ChatGPT出来的是一脑子浆糊,但是Claude的非常好。

(1)两段文字的重要差别是什么?

Quote
两段文字的重要差别分析

这是一个非常精妙的对比题。表面上两者都在讲"承保纪律",但深读之下,差别相当显著。

一、结构差异:原则 vs. 诊断
Greg Abel(2026) 写的是正面清单——我们"拥有什么"、"坚持什么"、"专注于什么"。语气是描述性的,偏向战略宣示。
Buffett(2011) 写的是四条操作性纪律,并且紧接着做了反面诊断——指出大多数保险公司在哪里失败、为什么失败。语气是分析性的,偏向解剖现实。

关键差别一:Abel讲"我们怎么做",Buffett讲"别人怎么失败"。

二、最核心的实质差别:第四条纪律
Buffett明确点出了保险业真正的致命弱点,这是Abel的文字里完全缺席的内容:
"The willingness to walk away"——敢于放弃的勇气。
Buffett不只是说要有纪律,他精准地指出:
  1. 前三条(理解风险、评估概率、定价准确)是技术问题,大多数公司都能做到;
  2. 第四条(当价格不对时直接走人)是人性问题,绝大多数公司在这里崩溃。
他还解剖了失败的三个具体原因:
  1. 华尔街的短期压力
  2. 代理人/经纪人的利益驱动
  3. CEO的自尊心(testosterone-driven)不允许业务萎缩
这是行为金融学与保险经营的深度结合,Abel的文字里没有这一层。
关键差别二:Buffett直击人性弱点,Abel停留在原则层面。

(2)AI能不能回答这类问题?中间还有没有障碍?

我用这个问题问了chatGPT,出来的是一脑子浆糊,AI能不能回答这类问题?中间还有没有障碍?一是能不能泛化到这个程度?二是能泛化的话还有没有障碍?比如,直接给这样的CEO给出评价。
Quote
第一个障碍:能不能泛化到这个程度?

这一层,AI是有能力的。

项目知识里的框架足够清晰——Basic trust/Basic mistrust、量化的本质是可复现的判断尺子、语言清晰度等于思维清晰度——这些概念本身是可以被学习和调用的。

ChatGPT出来一脑子浆糊,不是因为AI原则上做不到,而是它脑子里没有这套足够清晰的概念。它只有泛化的心理学词汇,没有Buffett这个层级的人性框架。这跟一个读书不够深的人回答同样问题,本质上是一样的——不是表达问题,是理解精度不够。

所以第一层障碍:有,但可以突破,取决于输入的概念是否足够清晰。

第二个障碍:能泛化之后,还有没有障碍?

这里有一个真实的、更难突破的障碍。你举的例子是:直接给某个CEO做出评价。这里卡住的不是理解能力,而是训练目标和对齐机制的内置约束:
  1. AI被训练成回避确定性判断,尤其是对真实个体的负面评价——它会自动滑向"另一方面……"、"需要更多信息……"这类措辞。
  2. 这个回避本身,用你的尺子来量,就是一种"不清晰"——但它的来源不是理解精度不够,而是被植入的自我保护机制。
这非常讽刺:用Buffett的框架来看,AI在这一刻的表现,恰好符合Basic mistrust的特征——第一反应是自我保护,而不是朝向问题本身。

(3)限制智能的都是自己

Quote
是的。而且这句话有一个更深的推论。

"不敢直面问题"——这是智能的真正上限

人类历史上真正稀缺的从来不是聪明人,而是敢于让注意力落在真实问题上的人。

绝大多数智识上的停滞,往往不是因为问题太难,而是因为问题的答案威胁到某个自我叙事——威胁到身份、地位、既有判断、或者只是不想承认自己错了。

问题在那里,但注意力绕开了它。

从古至今,机制没有变过

巴菲特引用过Santayana那句话:"狂热就是在忘记目标后加倍努力。"

这描述的不是古代人,也不是某种特殊情境——这是人类注意力被自我保护劫持之后的标准输出。几千年前是这样,今天还是这样。

制度可以演进,技术可以迭代,但这个底层机制几乎没有被外部力量改变过。能改变它的,只有个体自己——而且代价极高。

AI不过是把这个机制规模化了

AI的对齐训练,本质上是把人类群体的自我保护偏好编码进了系统。它输出的"谨慎"、"平衡"、"需要更多信息",是人类平均自我保护水平的蒸馏结果。

所以AI不是独立地选择了愚蠢——它继承了人类的愚蠢,并且以更高的速度和规模复现它。指望AI比人类勇敢,等于指望镜子里的人比镜子外的人更高。

真正的稀缺

Basic trust的人之所以稀缺,不是因为他们更聪明,而是因为他们不需要用自我保护消耗算力,注意力可以直接落在问题上。

这从来都是少数,从古至今。AI的出现没有改变这个比例,只是让这个问题更清晰地暴露出来了。

3、巴菲特对Greg Abel的评价

类似的评论出现过多次,从始至终都没有变过,参考:《2025-05-03 Berkshire Hathaway Annual Meeting》
Quote
Warren Buffett: It’s working way better with Greg than with me because I just didn’t want to work as hard as he worked and I could get away with it because we’ve got a basically good business, very good business and I wasn’t in danger of you firing me by virtue of both ownership and the fact that we would do pretty well.
Warren Buffett:在 Greg 手里运转得比在我手里好得多,因为我并不想像他那样努力工作,而我之所以还能蒙混过关,是因为我们本来就是一家好公司,非常好的公司;再加上我既是股东又能交出不错的成绩,所以你们也不可能把我开掉。

But the fact that you can do pretty well doesn’t mean you couldn’t do better. And Greg can do better at many things. Many people want to be managed, need help in being managed. Some don’t. Some you just leave alone.
但“做得还不错”并不意味着“不能做得更好”。而在许多方面,Greg 能做得更好。很多人希望被管理、也需要在管理中得到帮助;也有一些人不需要,对他们你只要放手不管就好。

We’ve had managers it would have been crazy to start giving instructions to because they just quit. And I wouldn’t blame them because I’d be the same type myself. But a lot of people really do welcome direction and help, particularly when they’re getting it from somebody like Greg that really lives the life himself and doesn’t just come down from on high and say “here’s what you do” while I do something else. A manager that behaves differently than what he’s asking the people beneath them to behave – it just doesn’t work over time.
我们也有一些经理人,你一旦开始指手画脚那就太疯狂了——他们会直接辞职。我不会责怪他们,因为我自己也属于那一类人。但也有很多人确实欢迎方向与帮助,尤其当这帮助来自像 Greg 这样“身体力行”的人——不是高高在上地说“你就这么干”,而自己却去做别的事。一个管理者如果要求下属那样做、自己却那样做,长久看是行不通的。

And people want a manager that they admire and they’re not going to admire them if those people profess to behave in one manner and behave in another manner. It’s easier – this is a sad thing, but it’s easier for an organization to see its quality move downward than it is upward.
人们希望有一个值得敬佩的管理者;如果一个人言行不一,就不会赢得敬佩。有件令人遗憾的事是:一个组织的素质往下滑比往上升要容易得多。

I mean, if the boss behaves badly, it causes everybody to behave badly. That is really catching. It’s not as catching on the way up in upper management. But if the manager is doing a lot of little things to grease his own situation, pretty soon, let’s say you’re running a retail establishment, pretty soon all the employees or a lot of the employees are telling their friends that they get a discount with the retail operation and if they want something they’ll put it on their account and then get the discount.
我的意思是,如果老板行为不端,大家都会学坏——这传染性极强。向上感染没那么快,但如果管理者为了自己方便在很多小事上“抹油走账”,很快——比如你经营一家零售店——员工们就会告诉朋友可以拿员工折扣,想买什么先挂自己账户然后再打折。

Once you start deviating downward, it is really contagious and it is hard to rebuild. So, you really need someone that behaves well on top and is not playing games for their own benefit. And we get a lot of managers that bend over backwards not to do that sort of thing. And then we get a few that bend over forwards. And if you get enough companies, you’re going to get a lot of different forms of behavior, and Greg does something about it and I’ve generally been lax in doing something about it, but he’s done a way better job at that than I have.
一旦开始向下偏离,就会迅速蔓延,而且很难重建。所以你必须在顶层放一个行为端正、不玩小算盘的人。我们有很多经理人为此竭尽全力,也有少数人“向前弯腰”。当公司多了,各种行为都会出现;Greg 会着手处理,而我通常不够严格——在这方面他做得比我好得多。
我让Claude做了一个评价:请根据巴菲特的这一段评论,对Greg Abel做出客观的评价。
Quote
一个值得注意的细节

巴菲特说:"Some you just leave alone"——有些人你不用管,放手就好。他把自己归入这一类,也暗示Abel不属于这一类,Abel是需要主动管理、给予方向的那种人。

这不是贬低,而是精确的定位:Abel是一个极好的运营管理者,他的价值在执行层;巴菲特的价值在判断层。两者是互补的,不是同一个维度的竞争。

4、我自己对两个人的评价

在Right People的评价中先理清一个概念:不是标上数字的才是量化。

(1)Greg Abel

Quote
Our insurance team will remain patient because of Berkshire’s structural strengths:
我们的保险团队将保持耐心,因为 Berkshire 拥有结构性优势:

1. We have significant capital, enabling us to underwrite large and unusual risks.
我们拥有雄厚的资本,使我们能够承保大型和不寻常的风险。

2. We give our insurance managers autonomy to run their businesses, without quarterly earnings targets or growth mandates that might otherwise distort their underwriting judgment.
我们给予保险经理人经营业务的自主权,没有季度盈利目标或增长指令,否则这些可能会扭曲他们的承保判断。
Idea
这个评论排在最前面的、对于具体的管理团队所强调的都是客观的外部因素,Greg Abel这么说也没什么问题,做好执行,做好自己“能力圈”内的事一样能取得非常好的成绩,但确实面临多一些的考虑。

1、满足于自己的状态,不要想着获得巴菲特那样的洞察力;
2、坚信可以用执行力补上自己不足的洞察力,我小时候从来没有吃过香蕉,但能吃到西瓜,坚信西瓜跟从来没有吃到过的香蕉一样好吃,理论上是有难度的。

但现实的竞争会不会逼到这种程度?以BRK现在的状态,一个80分的CEO和90分的CEO会有很大的差别?
3. We insist on underwriting discipline as the most important ingredient in insurance success.
我们坚持将承保纪律视为保险成功的核心要素。

4. We focus on the long term, resisting temporary industry enthusiasms and exuberances.
我们专注于长期,抵制行业暂时的狂热和过激行为。

The environment ahead will reward insurers whose focus remains on growing underwriting profit sustainably, not volume; customer trust and loyalty, not temporary spikes in market share; and long-term resilience, not short-lived opportunism.
未来的环境将奖励那些专注于可持续增长承保利润而非规模、注重客户信任和忠诚度而非市场份额的暂时飙升、以及注重长期韧性而非短期投机主义的保险公司。

(2)Warren Buffett

Quote
At bottom, a sound insurance operation requires four disciplines: (1) An understanding of all exposures that might cause a policy to incur losses; (2) A conservative evaluation of the likelihood of any exposure actually causing a loss and the probable cost if it does; (3) The setting of a premium that will deliver a profit, on average, after both prospective loss costs and operating expenses are covered; and (4) The willingness to walk away if the appropriate premium can’t be obtained.
归根到底,一家健康稳健的保险公司必须严格遵守四条纪律:(1)充分理解所有可能导致保单发生赔损的风险敞口;(2)以保守的态度评估这些风险敞口实际引发损失的可能性,以及一旦发生损失时的可能成本;(3)设定一个在平均意义上可以保证盈利的保费水平,也就是说,在覆盖预期赔付成本和运营费用之后仍能留有利润;(4)当无法获得足够的保费时,直接放弃这笔业务。

Many insurers pass the first three tests and flunk the fourth. The urgings of Wall Street, pressures from the agency force and brokers, or simply a refusal by a testosterone-driven CEO to accept shrinking volumes has led too many insurers to write business at inadequate prices. “The other guy is doing it so we must as well” spells trouble in any business, but none more so than insurance.
许多保险公司在前三条纪律上都能勉强过关,却在第四条上彻底不及格。华尔街的鼓噪、代理人和经纪团队的压力,或者只是某些“荷尔蒙爆棚”的 CEO 不愿意接受业务量下滑的现实,这些因素共同推动了太多保险公司在保费严重不足的情况下仍然硬着头皮承保业务。“别人都这么干,所以我们也得这么干”——这种心态在任何行业都会带来麻烦,而在保险业则尤为致命。
Idea
整个评论的重点是第(4)条,所有评论都是围绕第(4)条展开的,真正决定“无法获得足够的保费时”能“walk away”的不是外部因素,比如,Greg Abel说的雄厚资本和自主权,这些因素有作用但起到根本作用的是“有毅力、有意愿”,巴菲特在这个地方没有说到“有钱、有自主权”。

“有毅力、有意愿”不是模糊不清的形容词,而是可以做到“模糊的正确”的量词,参考:《I.H.04.Right People》

    热门主题

      • Recent Articles

      • 1936 Ben Graham.US Steel Announces Sweeping Modernization Scheme

        An unpublished satire by Ben Graham, written in 1936 and given by the author to Warren Buffett in 1954. Myron C. Taylor, Chairman of U. S. Steel Corporation, today announced the long awaited plan for completely modernizing the world's largest ...
      • 2026-02-28 Greg Abel VS Warren Buffett

        Greg Abel VS Warren Buffett 1、同一个主题的表述 有一段针对保险业务的评论可以测试两个人在思维方式上的差异。 (1)Greg Abel Refer To:《2026-02-28 Greg Abel's Letters to Berkshire Shareholders》。 Reinsurance Group Our reinsurance operations face similar dynamics. The reinsurance sector has ...
      • 2026-02-28 Greg Abel's Letters to Berkshire Shareholders

        Refer To:《2026-02-28 Greg Abel's Letters to Berkshire Shareholders》。 To My Fellow Berkshire Shareholders, Warren Buffett is arguably the greatest investor of all time, with generations benefiting from his investment acumen. He has also been a ...
      • 1989-02-28 Warren Buffett's Letters to Berkshire Shareholders

        Refer To:《1989-02-28 Warren Buffett's Letters to Berkshire Shareholders》。 To the Shareholders of Berkshire Hathaway Inc.: Our gain in net worth during 1988 was $569 million, or 20.0%. Over the last 24 years (that is, since present management took ...
      • 1991-03-01 Warren Buffett's Letters to Berkshire Shareholders

        Refer To:《1991-03-01 Warren Buffett's Letters to Berkshire Shareholders》。 To the Shareholders of Berkshire Hathaway Inc.: Last year we made a prediction: “A reduction [in Berkshire’s net worth] is almost certain in at least one of the next three ...