VC 行业的结构重组
主题综述
主题页(活文档)· 最近更新 2026-06-12 · 取材 6 篇访谈(逐字稿全文核对;Mercury 篇逐字稿仅存中文译稿)
更新日志
- 2026-06-12 — 首次综述。基于 6 篇访谈,核心结论:「钱已商品化、中型综合基金正在消亡」是六位发言人的真共识,但对同一道数学题(Botha 的 return-free risk)他们给出互斥解法——Sequoia 收缩供给做永久合伙制、a16z 扩表做权力平台、新基金拿分发换股权,三者对「每年到底有多少个赢家」的底层假设不可能同时为真。(本主题由旧库挖掘 /excavate 发现)
主流共识
一、资本本身不再稀缺——「投 VC 是无回报的风险」
这是全部六篇的底色。Botha 把账算得最狠:行业年投 1500–2000 亿美元,按 12% 净年化倒推需要每年 7000–8000 亿美元回流、超万亿美元总退出价值:
"You need 40 Figmas a year for the industry to make the returns work, which means that they don't. So, in my opinion, investing in venture is a return-free risk."「这个行业每年需要 40 个 Figma 才能实现回报,这意味着它们无法实现。所以,在我看来,投资风险投资是一种没有回报的风险。」Roelof Botha (Sequoia) · Sequoia's Roelof Botha: Why Venture Capital is Broken
Horowitz 在两年前的访谈里说的是同一件事,措辞几乎一样:
"The amount of capital versus the amount of great ideas, we already have more capital than great ideas."「资本的数量与伟大的想法的数量相比,我们已经拥有比伟大的想法更多的资本。」Ben Horowitz (a16z) · Ben Horowitz on How a16z Was Built
Andreessen 则把「钱不稀缺」直接翻译成创始人视角——这句话同时解释了为什么钱本身不再是 VC 的产品:
"Raising money from venture capitalists is the easiest thing you will ever do as a startup founder. We are sitting here with checkbooks waiting to write checks."「从风险投资家那里融资是你作为创业公司创始人所能做的最简单的事情。我们坐在这里,拿着支票簿,等着开支票。」Marc Andreessen (a16z) · Marc Andreessen | The Future of Venture Capital
种子端的实操者从下往上看到同一景象。Immad Akhund 说 AI 种子轮「同一个想法听到第四遍、4000 万美元估值还有顶级投资人抢」(该篇逐字稿仅存中文译稿):
「在种子阶段,做 AI 太难了。……这就像我第四次听到同一个想法的推销。创始人正在以大约 4000 万美元的估值进行融资。……但伟大的风投公司正在资助这些项目,对吧?每个人都在做。」
— Immad Akhund (Mercury) · 20VC Exclusive: Mercury Founder Launches First $26M Fund
二、幂律统治一切,平均值与 TAM 都是伪命题
回报结构上没有任何分歧:只有极少数公司决定一切,错过赢家(error of omission)是唯一致命的错误。Andreessen 把它推广到公开市场:
"The S&P 500 is no longer the S&P 500. It's like the S&P 492 and the S&P 8."「标准普尔 500 指数已经不再是标准普尔 500 指数了。它就像标准普尔 492 和标准普尔 8。」Marc Andreessen (a16z) · Marc Andreessen | The Future of Venture Capital
Shardul Shah 把同一逻辑用在对自己行业的拒绝上——拒绝均值、拒绝 benchmark、拒绝 TAM:
"We are in the business of finding fund returners. The power law dominates our business. Not being in a $10, $20, $50, $100 billion company is actually painful."「我们的业务是寻找能够带来基金回报的公司。幂律支配着我们的业务。没有投资到市值 100 亿、200 亿、500 亿甚至 1000 亿美元的公司实际上是很痛苦的。」Shardul Shah (Index Ventures) · 20VC: Index's Shardul Shah on Why Market Size is a Trap
"I don't know why we're talking about averages. None of us are in the business of mean reversion. So, you know, if some strategy at Greylock is to buy the index of series B's, like, God bless, but that's not what we're going to do at index."「我不知道我们为什么在讨论平均值。我们谁也不是在做均值回归的生意。所以,如果 Greylock 的某些策略是购买 B 轮指数,那就随他们去吧,但那不是我们 Index 要做的事。」Shardul Shah (Index Ventures) · 20VC: Index's Shardul Shah on Why Market Size is a Trap
三、「等客上门」的旧模式死了,行业被迫重组
Andreessen 给这个主题下了定义性的判断,也贡献了最生动的旧时代标本——2009 年某顶级基金 GP 对他说风投就像回转寿司店:
"He said, you know, if you miss one, it doesn't matter. Cause there's another sushi boat coming up right behind it. And he's just like, you just sit and watch the sushi go by."「他说,你知道,如果你错过了一个,没关系。因为后面还有另一艘寿司船会过来。他只是说,你只要坐在那里,看着寿司经过。」Marc Andreessen (a16z) · Marc Andreessen | The Future of Venture Capital
"I think what's happened is like the industry, the venture industry, it had to restructure in order to basically accommodate the change in the market."「我认为现在发生的情况是,风险投资行业必须进行重组,才能适应市场的变化。」Marc Andreessen (a16z) · Marc Andreessen | The Future of Venture Capital
四、杠铃化:两端活、中间死
对行业最终形态的描述,a16z 两人与 Mercury 的 Immad 高度一致——规模端与专精端成立,中间的传统综合基金被挤出。Andreessen 说得最绝对:
"One is the death of the middle, which is it just doesn't make sense to have the old fashioned, you know, series A, series B, six GPs, $300 million fund sitting outside the road waiting for people to walk in the door. Like those days are over and those funds are, you know, those funds are shutting down. Like that model's going away."「一是中间环节的消亡,即拥有老式的 A 轮、B 轮、六位普通合伙人,3 亿美元的基金坐在路边,等待人们走进来,这已经没有意义了。那些日子已经结束了,那些基金正在倒闭。这种模式正在消失。」Marc Andreessen (a16z) · Marc Andreessen | The Future of Venture Capital
Horowitz 的版本把两端点名为「品牌」和「专精」,并把 Sequoia 划进自己一端:
"The kind of classical venture firm that is just like a collection of smart, you know, investors. Like, I think that's probably run its course. So I think you have to be like a top end, like serious brand that can marshal resources and money and considered smart money and people want to follow. I put us in that category, Sequoia. There's that class of thing and then there's people who are very specialized in a very kind of specific part of the market."「就像那种传统的风险投资公司,就像一群聪明的投资者。我觉得那可能已经过时了。所以我认为你必须成为顶级的,像一个可以调动资源和资金的知名品牌,被认为是聪明的资金,人们愿意追随。我把我们和 Sequoia 归为这一类。有这类事物,还有一些人在市场的某个非常特定的部分非常专业。」Ben Horowitz (a16z) · Ben Horowitz on How a16z Was Built
Immad 在小端独立得出同样结论,但留了一分不确定(该篇逐字稿仅存中文译稿):
「存在一种哑铃式的概念,对吧?像我这样投资较小金额的人,我们会做得很好,多阶段基金也会做得很好。我不知道中间会发生什么。我认为中间环节会遇到更多问题。」
— Immad Akhund (Mercury) · 20VC Exclusive: Mercury Founder Launches First $26M Fund
注意一个温差:Shardul 不接这套「赢家/输家」叙事。他承认要适应,但认为行业淘汰速度被高估了——
"But given the industry is so driven by the power law, I think people can withstand cycles longer than we probably give them credit for."「但是,鉴于这个行业是由幂律驱动的,我认为人们可以承受比我们可能认为的更长的周期。」Shardul Shah (Index Ventures) · 20VC: Index's Shardul Shah on Why Market Size is a Trap
分歧在哪
共识止于「钱太多、中间死」。往下走,每一道关键题——基金该多大、VC 到底卖什么、要不要上市、种子怎么拼、决策权放在哪——桌上的人给出的是互斥答案。
1. 同一道「钱太多」的数学题,Sequoia 选择缩表,a16z 选择扩表
Botha 的解法是把自己从过剩资本里摘出来——基金规模冻结在五七年前的水平,明确放弃费用收入和行业份额:
"Well, we stick to our knitting. The funds we operate today, our seed venture and growth funds today are no bigger than they were five, six, seven years ago. … We literally want to be the best net IRR and net multiple for our LPs. And we're not interested in maximizing fees or maximizing share of industry value creation. That's the game we've chosen to play."「嗯,我们坚守本业。我们今天运营的基金,我们的种子基金和成长基金,规模并不比五、六、七年前更大。……我们真的想为我们的 LP 提供最佳的净 IRR 和净倍数。我们对最大化费用或最大化行业价值创造份额不感兴趣。这是我们选择玩的游戏。」Roelof Botha (Sequoia) · Sequoia's Roelof Botha: Why Venture Capital is Broken
Horowitz 的解法正相反:赢家数量扩大了一个量级,所以基金必须等比例扩大——
"If you believe the market was fixed at 15 companies, then that's the exact right strategy. We don't believe that. … If there were 15 companies and now there's 150, then if you had a $400 million fund, then maybe you need a $4 billion fund to do the same deals, if you win the same percentage of them. That's just a simple math."「如果你认为市场固定在 15 家公司,那么这才是正确的策略。我们不这么认为。……如果之前有 15 家公司,现在有 150 家,那么如果你有一个 4 亿美元的基金,那么你可能需要一个 40 亿美元的基金才能做同样的交易,如果你赢得的百分比相同。这只是简单的数学。」Ben Horowitz (a16z) · Ben Horowitz on How a16z Was Built
"I'm not allowed to talk about our fund returns because we're an RIA, but if you look at our funds, I think our larger funds have at times outperformed our smaller funds."「我不能谈论我们的基金回报,因为我们是 RIA,但如果你看看我们的基金,我认为我们较大的基金有时表现优于较小的基金。」Ben Horowitz (a16z) · Ben Horowitz on How a16z Was Built
Andreessen 给出配套的赢家计数:
"150 companies a year that like really matter. And the reason is because there's so many more sectors now."「每年有 150 家真正重要的公司。原因是现在有更多的领域。」Marc Andreessen (a16z) · Marc Andreessen | The Future of Venture Capital
而 Botha 的计数直接否定这个前提——他数的是真金白银的退出,不是纸面估值:
"If you look at every single decade, there are only about 20 companies that end up getting exit values north of a billion dollars. Actual IPOs or M&As north of a billion. Not the paper write-ups. Only 20 companies. More money doesn't create more great ideas or more great founders."「如果你回顾过去几十年,最终退出价值超过 10 亿美元的公司只有大约 20 家。实际的首次公开募股或并购超过 10 亿美元。不是纸面上的账面价值。只有 20 家公司。更多的资金并不能创造出更多伟大的想法或更伟大的创始人。」Roelof Botha (Sequoia) · Sequoia's Roelof Botha: Why Venture Capital is Broken
「每年 150 家重要公司」与「每十年 20 个十亿美元退出」不可能同时支撑各自的基金规模策略。这是全题最硬的一组对撞——而且两边都拿不出对方接受的计数口径。Horowitz 还为扩表加了一层使命论护甲,等于宣布回报不是唯一记分牌:
"Our mission isn't to maximize the number of money per partner. Our mission is to be the resource for building great technology companies."「我们的使命不是最大化每个合伙人的金钱数量。我们的使命是成为构建伟大科技公司的资源。」Ben Horowitz (a16z) · Ben Horowitz on How a16z Was Built
2. VC 的真实产品:权力、声誉、判断,还是分发?
四种答案,彼此并不兼容。Andreessen 的答案是权力——为还没长出肌肉的公司投射权力,品牌只是权力的一种信用形式:
"This has been one of our theories, how we built our firm is you optimize for maximum amount of power in order to be able to give the startups access to it."「这一直是我们的理论之一,我们建立公司的方式是优化最大程度的权力,以便让初创公司能够获得它。」Marc Andreessen (a16z) · Marc Andreessen | The Future of Venture Capital
"I sometimes describe it as venture firm is providing a bridge loan of a brand."「我有时把它描述为风险投资公司提供品牌过桥贷款。」Marc Andreessen (a16z) · Marc Andreessen | The Future of Venture Capital
Horowitz 的答案是给创始人的产品 + 声誉——firm 先于投资人,靠「陪到最后」卖信用:
"We're kind of a product first and then a team of investors second, whereas every other firm I think is the opposite."「我们是产品第一,投资者团队第二,而我认为其他公司则恰恰相反。」Ben Horowitz (a16z) · Ben Horowitz on How a16z Was Built
"Take us as your partner. We're going to be there till the bitter end. … our whole idea is that we sell on reputation."「把我们当作你的合作伙伴。我们会陪你走到最后。……我们的想法是靠声誉来销售。」Ben Horowitz (a16z) · Ben Horowitz on How a16z Was Built
Botha 承认平台化让创始人受益("really the founders are the ones who benefited"),但 Sequoia 的运营团队不为创始人服务、为投资人自己服务——产品是判断本身:
"We've decided to not build as big an organization. Most of the operating teams we have at Sequoia help us. So we have about as many developers at Sequoia as we have investors and they're building products for us so that we are much more effective and productive than we might have been 20 years ago."「我们已经决定不建立那么大的组织。我们在 Sequoia 的大部分运营团队都在帮助我们。所以我们在 Sequoia 的开发人员和投资者数量差不多,他们为我们开发产品,以便我们比 20 年前更有效率和生产力。」Roelof Botha (Sequoia) · Sequoia's Roelof Botha: Why Venture Capital is Broken
买方证词站在 Botha 这边、打在 a16z 身上——Immad 是 a16z 的被投创始人,他对「平台增值」的评价不客气(该篇逐字稿仅存中文译稿):
「最近 Horowitz 是我们的种子轮投资者……他们提供的很多附加值并不是非常有用。我认为,归根结底,任何风险投资公司最有价值的两件事是,你知道,合伙人是谁,对吧?……还有第二件事,我认为人们不太理解的是,你可以获得创始人网络。」
— Immad Akhund (Mercury) · 20VC Exclusive: Mercury Founder Launches First $26M Fund
而 Immad 自己的新基金给出第四种答案:分发。他不靠平台也不靠品牌,靠 Mercury 在创业公司里 30–40% 的渗透率自动产生交易流(该篇逐字稿仅存中文译稿):
「其次,当 30% 到 40% 的初创公司使用 Mercury 时,我的交易流量又上了一个台阶。每个人都知道我,这意味着很多人都希望我投资,我到了一个无法处理所有这些事情的地步。」
— Immad Akhund (Mercury) · 20VC Exclusive: Mercury Founder Launches First $26M Fund
媒体是分发的另一形态。Shardul 公开承认自己在这件事上转了 180 度——而把他说服的理由,恰恰是「个人品牌是基金的获客资产」:
"I was totally opposed to participating with media. I thought it was self-serving, brand building. And here I am!"「我完全反对与媒体合作。我认为那是自我服务,品牌建设。而我今天就在这里!」Shardul Shah (Index Ventures) · 20VC: Index's Shardul Shah on Why Market Size is a Trap
"Anna who runs marketing at index took me aside and said, you're an idiot. This is not for you. This is for index and your founders."「在 Index 负责市场营销的 Anna 把我拉到一边说,你真是个白痴。这不是为了你。这是为了 Index 和你的创始人。」Shardul Shah (Index Ventures) · 20VC: Index's Shardul Shah on Why Market Size is a Trap
主持人 Harry Stebbings——本人就是「媒体先于基金」模式的实证——把这条路线说穿了:
"Bill Gurley built Benchmark's brand by selling the brand of Bill Gurley. You sell index by selling the Shardul brand."「Bill Gurley 通过推销他自己的品牌来打造 Benchmark 的品牌。你可以通过推销 Shardul 这个品牌来推销 Index。」Harry Stebbings (20VC,主持人) · 20VC: Index's Shardul Shah on Why Market Size is a Trap
3. 上市还是永久合伙制:三种互斥的终局想象
Immad 认为多阶段基金上市不可避免、且资本还会继续涌入(该篇逐字稿仅存中文译稿):
「似乎不可避免地,这些多阶段基金中的一些会进行首次公开募股,成为上市公司,对吧?我们听到了一些关于 GC 这样做的消息。我认为这肯定会发生,而且实际上会有越来越多的资金进入这个领域。」
— Immad Akhund (Mercury) · 20VC Exclusive: Mercury Founder Launches First $26M Fund
Horowitz 认为这是结构性自毁——公开市场给费用流估值,会激励基金把行业搞乱:
"The public markets value them on their fee stream much more than on their investment returns. … And as we saw, I think with both SoftBank and Tiger Global, if you try to change that demand supply imbalance, you just end up creating a mess. And so if you were public, you'd have a strong incentive to create a mess."「公共市场对它们的估值更多地基于它们的费用流,而不是它们的投资回报。……正如我们所看到的,我认为 SoftBank 和 Tiger Global 都试图改变这种供需失衡,你最终只会制造混乱。如果你是一家上市公司,你会有强烈的动机去制造混乱。」Ben Horowitz (a16z) · Ben Horowitz on How a16z Was Built
Botha 则用产权结构直接关死这扇门——不仅不上市,连代际之间都不交易:
"We've structured ourselves to be a private partnership in perpetuity to the extent possible under California law. … We didn't have to pay to get the partnership from the previous generation and nor will we charge the next generation. That's our motto."「我们已经将自己构建为一个永久性的私人合伙企业,在加州法律允许的范围内。……我们不必为了从上一代手中获得合伙企业而付费,我们也不会向下一代收费。这是我们的座右铭。」Roelof Botha (Sequoia) · Sequoia's Roelof Botha: Why Venture Capital is Broken
有意思的是 Horowitz 与 Immad 其实共享同一判断(上市激励是坏的),分歧在「坏的事会不会照样发生」;而 Botha 的永久合伙制和 Sequoia Capital Fund(IPO 后继续持有,三年半多赚 67 亿美元)把「不退出、不变现」做成了制度——这与 Immad 预言的「上市的、越做越大的多阶段基金」是两个物种。
4. 杠铃两端怎么拼:种子市场的三套互斥打法
都同意杠铃,但「大端怎么对待小端」答案完全不同。Andreessen 承认大基金在结构上做不了种子——这是给小端留生态位的版本:
"We cannot do all of the seed investments that we would like to do. In fact, we can't even do a tiny fraction of them. It's just like strategically, structurally, we just can't do it. … It's the same reason why Amazon can't give you the champagne experience, right?"「我们无法进行所有我们想做的种子轮投资。事实上,我们甚至连其中很小的一部分都做不到。这只是在战略上、结构上,我们就是做不到。……这也是亚马逊无法给你提供香槟体验的原因,对吧?」Marc Andreessen (a16z) · Marc Andreessen | The Future of Venture Capital
但他同时给小端划了红线——别长大、别走到中间来:
"What does a successful seed investor do, right? He raises more money and wants to become a venture investor, right? But then you're going from one side of the barbell back to the middle, and you're creating that same problem again."「一个成功的种子轮投资者会做什么,对吧?他会筹集更多资金,并想成为一名风险投资者,对吧?但这样你就从哑铃的一端回到了中间,你又在制造同样的问题。」Marc Andreessen (a16z) · Marc Andreessen | The Future of Venture Capital
"I would not accidentally end up competing with Sequoia for Series A. Like I was just saying, that's a bad way to live."「我不会意外地最终与 Sequoia 竞争 A 轮融资。就像我刚才说的,那样生活不好。」Marc Andreessen (a16z) · Marc Andreessen | The Future of Venture Capital
Shardul 的打法是大端主动把蛋糕切给小端——包销整轮、再拆三份,正面回应「signaling」指控,并顺手点名了同行的话术:
"Most multistage funds say that signaling doesn't matter and the seed fund doesn't provide any value, so you should work with us. Our strategy is a little bit different. When I meet a founder at a seed stage, I'll tell him or her, I'll underwrite the entire round, so there's no financing risk, but we should split it into three sleeves. One sleeve is for index, one sleeve is for a seed fund. And the third sleeve is for angels operators. Because it takes a village."「大多数多阶段基金表示,信号无关紧要,种子基金不提供任何价值,所以你应该和我们合作。我们的策略有点不同。当我在种子阶段遇到创始人时,我会告诉他或她,我可以承担整个融资额,所以不存在融资风险,但我们应该把它分成三部分。一部分给 Index,一部分给种子基金。第三部分给天使投资人。因为这需要集体的力量。」Shardul Shah (Index Ventures) · 20VC: Index's Shardul Shah on Why Market Size is a Trap
Immad 站在小端,根本不觉得多阶段基金的种子产品威胁到他——因为大基金只认得出一种创始人(该篇逐字稿仅存中文译稿):
「这对你来说可能很难,但对我来说不一定,因为我就像,好的,当然,(大基金)在领投,我加入。……种子轮的定义就是有很多未知因素,而且这些多阶段基金只会投资特定类型的创业者,对吧?比如,通常要么是来自大公司的高管在创业,要么是多次创业的连续创业者。」
— Immad Akhund (Mercury) · 20VC Exclusive: Mercury Founder Launches First $26M Fund
值得记一笔:Andreessen 本人的私人资产配置就是一座杠铃——「两端都好」不是外交辞令,是他自己的钱的去向:
"They're both good. … If I were for some reason not doing this, I would immediately do what you're doing. … my liquid assets are basically tied up in either A16Z funds on the one side or I run a very aggressive personal investment program in basically angel and early stage seed funds. It's because I believe in the bar bill."「两者都很好。……如果我因为某些原因不这样做,我会立即做你正在做的事情。……我的流动资产基本上要么投入到 A16Z 基金中,要么我运行一个非常激进的个人投资计划,基本上是天使投资和早期种子基金。因为我相信哑铃策略。」Marc Andreessen (a16z) · Marc Andreessen | The Future of Venture Capital
5. 判断怎么生产:一票否决 vs 把扳机下放
同样面对「幂律下如何聚合判断」,Sequoia 与 a16z 的决策机器是镜像。Sequoia 要全体一致,任何一个合伙人都能杀死一笔投资:
"When we make investment decisions at Sequoia, it's a consensus decision, which blew my mind when I first got there."「当我们在红杉做出投资决策时,这是一个共识决策,当我第一次到那里时,这让我大吃一惊。」Roelof Botha (Sequoia) · Sequoia's Roelof Botha: Why Venture Capital is Broken
a16z 把扳机交给离子弹最近的人,两位创始人明确不审批:
"We have investment verticals. They're discrete teams. … they have like trigger pull … authority. They can make investment decisions. Like we don't run the firm or Ben and I sit and decide, is this a good investment or bad investment? Like our specialists make those decisions."「我们有投资垂直领域。它们是独立的团队。……他们拥有启动决策的权限。他们可以做出投资决策。就像我们没有运营公司,也不是我和 Ben 坐下来决定,这是否是一项好的投资或糟糕的投资?我们的专家会做出这些决定。」Marc Andreessen (a16z) · Marc Andreessen | The Future of Venture Capital
但 a16z 在治理层又比谁都集权——经济利益共享、控制权不共享,Horowitz 认为这恰恰是能持续重组的前提:
"We have a different structure where we're shared economics, but we've kind of centralized control. And that enables us by not having shared control, we can change the structure of the firm very easily."「我们的结构不同,我们有共同的经济利益,但我们将控制权集中起来。通过不拥有共同的控制权,这使我们能够非常容易地改变公司的结构。」Ben Horowitz (a16z) · Ben Horowitz on How a16z Was Built
所以严格说这是两套完整哲学的对峙:Sequoia = 投资上全体一致 + 产权上永久合伙;a16z = 投资上单点拍板 + 治理上中央集权。两家都赢,说明「正确的判断生产方式」这道题在样本里没有收敛。
6. 桌子另一头:创始人证词里的 VC
这组分歧最微妙——VC 说自己是创始人的客户服务业(Andreessen:"There are two customers. They're the LPs and they're the founders."),但两位创始人的证词给这个说法上了压力。Ivan Zhao 回忆是谁把「招职业经理人、扩张团队」的剧本塞给他的——
"That's a good question. Who tell us? Some investors."「这是个好问题。谁告诉我们的?一些投资者。」Ivan Zhao (Notion) · Notion's Ivan Zhao: The Refounder
"It was conventional wisdom amongst the investor class."「在投资者圈子中,这是传统的智慧。」主持人(Sequoia 合伙人、HubSpot 联合创始人) · Notion's Ivan Zhao: The Refounder
但 Ivan 没有把账全算在投资人头上——他抵制传统打法的那部分(不建销售团队、想用第一性原理发明全新 PLG 动作)恰恰是他自认的错误:"In retrospect, I should have started Sales Team earlier… That's a mistake."(「回想起来,我应该早点开始销售团队……那是个错误。」)创始人证词在这里是双向的:投资人的剧本既害人也救人。
Immad 的版本更釜底抽薪——最好的投资恰恰是不需要他的:
「我以 2000 万的估值投资了他们(Rappi),他们就再也没有和我联系过。……我当时就想,如果我最好的投资对象都不和我交流,那我做投资人还有什么意义呢?」(该篇逐字稿仅存中文译稿)
— Immad Akhund (Mercury) · 20VC Exclusive: Mercury Founder Launches First $26M Fund
Shardul 被问到 Founders Fund 的「最好的创始人不需要 VC」时,回答是半让步半反击:
"Well, I'm glad they said that. Good luck."「很高兴他们这么说。祝他们好运。」Shardul Shah (Index Ventures) · 20VC: Index's Shardul Shah on Why Market Size is a Trap
而对 2026 年这个当口,创始人和(前创始人出身的)投资人之间出现了罕见的同频悲观——主持人转述「2026 年份的 venture 会很糟」,Ivan 不反驳:
"In times like 2026, you get a lot of tourist entrepreneurs. And so I've heard people say, you know, 2026 Ventures is going to be crap."「在 2026 年这样的时期,你有很多游客型企业家。所以我听到有人说,2026 年风险投资会很糟糕。」主持人(Sequoia 合伙人、HubSpot 联合创始人) · Notion's Ivan Zhao: The Refounder
"A little bit. I don't disagree. Market will do their thing. Market is kind of beautiful in that way."「有一点。我不反对。市场会按照自己的方式发展。从这个角度来看,市场很美妙。」Ivan Zhao (Notion) · Notion's Ivan Zhao: The Refounder
都没说透的
1. 「钱怎么变少」没有人回答。 Chamath 当面问 Botha 资本量如何回归合理,得到的是 "I've been wondering that for 20 years."(「我已经困惑了 20 年了。」)Botha 把 LP 的持续供给归因于 J 曲线下的「hope springs eternal」,但他自己的 scout 计划、a16z 的扩表、Immad 预言的基金 IPO,全都是往系统里加钱的机制。桌上每个人都说行业钱太多,没有一个人的策略是让自己变小——过剩永远是别人。
2. 冲突政策是规模端的隐形天花板,但没人给它定量。 Andreessen 说限制 a16z 变大的不是人才而是 "Conflicts. Conflict policy."(「冲突。冲突政策。这是迄今为止最大的问题。」),又同时押注 AI 让「所有 incumbents 被摧毁」、每个品类都会塌缩出全栈赢家。这两个论断放在一起意味着冲突网格会越收越密——大端的扩张极限到底在哪,访谈里被绕开了。
3. 退出端的管道坏了,修法无人谈。 Botha 的整套数学以退出价值为分母;Immad 单独指出公开市场的结构性破损(被动化、上市成本、低于 100 亿美元没人看),但他的结论是「我不知道该如何解决」。Sequoia Capital Fund 的「IPO 后继续持有」其实是绕开退出而非修复退出。「每年 150 家重要公司」(Andreessen)和「每十年 20 个十亿美元真退出」(Botha)之间巨大的剪刀差——纸面赢家如何变成现金——没有任何一篇正面处理。
4. 杠铃的「专精端」语义在漂移,没人对齐。 Horowitz 说的专精是品类基金(pure crypto / pure games / pure AI firm),Shardul 说 "I believe in concentrations, not majors" 且 Index 全员 stage agnostic(与全球按阶段分工的主流相反),Immad 的小端靠分发而非领域专长,Ivan Zhao 的 barbell 干脆是讲工程师招聘的(super junior + super senior)。同一个词在四张嘴里指四种东西——「中间死掉之后两端各自长什么样」其实远没有共识表面看起来那么清楚。
我的看法
以下是判断,不是转述。其一,「资本商品化」之后 VC 的竞争已经移到非资本资产上,而 Sequoia(永久制+小团队+判断)、a16z(平台+权力+子基金)、Mercury/20VC 式新基金(分发获客)这三套解法并不互斥——杠铃成立后它们是同一行业里的三个生态位,未来十年大概率并存而非互相淘汰;对此我把握较高。其二,真正可证伪、且十年内见分晓的分歧是赢家计数:如果 Botha 的「每十年 20 个真退出」继续成立,a16z 的 40 亿美元基金数学(15→150)撑不住,反之 Sequoia 的克制就是把份额白送人;我目前略偏 Botha 一侧,因为他数的是现金、对面数的是纸面,但把握中等,AI 这一轮可能真的改变退出量级。其三,「中型综合基金死亡」是全场最一致也最可信的结论——连立场最谨慎的 Shardul 也只是争辩死得没那么快,而不是不死。
还想知道什么
- 两套赢家计数的对账:Andreessen 的「每年 150 家重要公司」有没有 cohort 级数据(按什么口径、纸面还是退出)?与 Botha 的「每十年约 20 个 >$1B 真退出」用同一口径对齐后,剪刀差到底多大?
- a16z 大基金的净倍数:Horowitz 以 RIA 为由不披露("our larger funds have at times outperformed our smaller funds" 无法核验)。需要第三方 LP 数据(Cambridge 等)验证「扩表不稀释回报」是否成立。
- Sequoia Capital Fund 的对照基准:67 亿美元额外收益是相对「IPO 即分发、LP 自行卖出」的反事实算的,持有期的回撤风险和费用如何计提?LP 实际拿到手的 net 差异是多少?
- 分发型基金的兑现数据:Immad 的 60 家非主导组合(均额 15 万美元)与 20VC 式媒体获客基金,第一个完整周期的 DPI 是否真能证明「分发 > 品牌」?以及 Immad 预言的多阶段基金 IPO(GC 等)是否在 2026–2028 间发生——这将直接裁决 Horowitz 与 Immad 的分歧。
取材
- 20VC Exclusive: Mercury Founder Launches First $26M Fund(Immad Akhund)· 2025-06-18 ·(逐字稿仅存中文译稿)
- Marc Andreessen | The Future of Venture Capital(Jack Altman 主持)· 2025-08-19
- Ben Horowitz on How a16z Was Built(a16z Podcast,对话录制于 2023)· 2025-08-25
- 20VC: Index's Shardul Shah on Why Market Size is a Trap(Harry Stebbings 主持)· 2025-08-26
- Sequoia's Roelof Botha: Why Venture Capital is Broken & How Great Companies Are Built(All-In Summit)· 2025-10-13
- Notion's Ivan Zhao: The Refounder(主持人为 Sequoia 合伙人、HubSpot 联合创始人)· 2026-05-25