Warren Buffett built his empire on concentrated bets. Greg Abel is now building his on a different kind of diversification — one that mirrors the very decentralized principles the Oracle of Omaha once dismissed. On August 15, Berkshire Hathaway’s Q2 2026 13F filing revealed a seismic shift: a $17 billion new position in Alphabet, Google’s parent, while trimming Bank of America and Kroger. The total portfolio value rose to $29.9 billion, and the net buying spree ended 14 consecutive quarters of selling. This isn’t just a portfolio rebalancing. It’s a governance signal from the most centralized capital allocation machine in history — a machine that is now learning to think like a DAO.
Context matters. Berkshire’s traditional model is a blend of autocracy and trust: one man (Buffett) and his chosen successor (Abel) decide where billions go, with zero on-chain voting. The 13F filing is their only transparency — a quarterly report card that reveals which sectors they believe in. For decades, the formula was simple: value, moats, and cash-flow. But the Q2 2026 filing shows a pivot toward tech growth, with Alphabet becoming the fourth-largest holding, displacing Bank of America. Apple, American Express, and Coca-Cola remain top, but the newcomer is telling. Abel is betting on advertising, AI, and cloud — sectors that once seemed too speculative for Omaha. This is the same kind of shift we saw in DeFi Summer 2020, when traditional liquidity providers started moving from stablecoin pools into volatile, high-yield protocols. The risk appetite is changing, but the governance structure is not.
Here’s the core insight: Berkshire’s portfolio shift is a live case study in how centralized capital allocators are forced to adopt decentralized strategies to survive. In the 2022 bear market, I watched DAO treasuries rebalance from stablecoins to ETH and staked assets, trying to capture yield while preserving governance power. The data was clear — protocols that diversified into growth assets (like Uniswap’s treasury holding ETH) outperformed those that sat on cash. Now, Berkshire is doing the same. Abel increased holdings in Delta Air Lines, Lennar, and Macy’s — all cyclical bets on recovery. He slashed Bank of America by 5.89% and exited First Capital Financial entirely. The message is unmistakable: holding cash and legacy financials is a losing strategy when the market rewards innovation. But here’s the rub — Berkshire’s decision-making is opaque. We don’t know the debate between Abel and the board. We don’t see the votes. In a DAO, every treasury proposal is on-chain, with quorum and delegation. Berkshire’s 13F is a crude approximation of that transparency. Code is law, but people are the protocol — and Omaha’s protocol is still a black box.
The contrarian angle is that Berkshire’s move is not a victory for decentralization but a sign that centralized capital is mimicking decentralized mechanisms without adopting their values. Abel is buying Alphabet because it dominates digital advertising — a centralized monopoly. He’s not buying ETH or BTC. He’s not entering a DAO. The net buying of $20 billion is a bet on the same tech giants that blockchain aims to disrupt. This is a trap many DAOs fell into during 2021: they copied traditional fund strategies (buying blue-chip tokens) instead of creating their own economic models. Governance isn’t just about voting; it’s about portfolio management. When a DAO treasury holds 80% of its assets in ETH, it’s not diversified — it’s correlated to the same market as every other protocol. Berkshire’s top five holdings still account for 88.74% of the portfolio — that’s extreme concentration by any measure. In the decentralized world, we call that a whale risk. The same inertia that kept Buffett in Coca-Cola for decades is now being replicated by Abel in Alphabet. The lesson for the blockchain community is not to celebrate Berkshire’s tech pivot, but to recognize that centralization of capital, whether in a corporation or a DAO, leads to the same failure modes: groupthink, lack of exits, and vulnerability to single points of failure.
Here’s what the data actually tells us. Based on my experience auditing DAO treasuries during the 2022 bear market, I saw a pattern: projects that concentrated their holdings in a single asset (like ETH or a governance token) suffered the most severe drawdowns. Those that rebalanced into multiple asset classes — stablecoins, real-world assets, and even traditional equities — survived with operational liquidity. Berkshire’s latest filing shows they are doing the same thing, but with a 14-quarter lag. They sold stocks for over three years, accumulating cash, and are now deploying into growth. That’s a classic bear market strategy — buy when others are fearful. But the blockchain twist is that this strategy is programmable. Smart contracts can automate rebalancing based on market conditions. We saw this with Uniswap V4’s hooks, which allow developers to create custom logic for liquidity management. The complexity spike may scare off 90% of developers, but the remaining 10% will build funds that rebalance faster than any human. Berkshire’s quarterly 13F is a relic — a mainframe in a world of sharded chains.
Take a moment to consider the governance implications. Abel’s $17 billion Alphabet bet was likely approved by a board of directors — a centralized group of 12-15 people. In a DAO, a proposal of that size would require a vote, delegation, and a quorum. The outcome would be debated on Discord and Twitter, with thousands of token holders weighing in. Which process is more resilient? The Berkshire process is faster but fragile — if Abel makes a mistake, the portfolio suffers instantly. The DAO process is slower but robust — it distributes risk across many minds. However, as I’ve argued before, delegation makes governance more centralized. Users are too lazy to research and simply delegate to KOLs. So DAOs end up with the same concentration problem, just with a different name. Berkshire’s top five holdings are like a DAO’s top five proposals — they dominate the agenda. The real innovation is not in who votes, but in how the treasury is managed. Smart contracts can enforce diversification rules, timelocks, and exits without human intervention. That’s the future Abel is ignoring.
The takeaway is not about Berkshire’s stock picks. It’s about the structural shift in capital allocation that blockchain enables, and which centralized institutions are only now beginning to mimic. Greg Abel is steering the ship toward tech, but he’s still steering a single ship. The decentralized world envisions a fleet of autonomous vessels, each with its own governance, risk parameters, and rebalancing logic. The 2022 bear market taught us that survival matters more than gains. Berkshire’s increased holdings in Delta Air Lines and Macy’s are bets on recovery, but they are also bets on human judgment. In a bear market, the protocols that survived were those that had automated treasury management, diversified revenue streams, and community-driven oversight. Berkshire has none of those. It has a CEO and a board. That model worked for 60 years, but the next 60 years will belong to protocols that embed trust in code, not in individuals. Code is law, but people are the protocol — and the protocol is evolving.
We didn’t need a 13F filing to know that the era of single-manager funds is fading. The data was already on-chain. DeFi Summer showed us that liquidity pools can outperform any hedge fund if they are properly incentivized. The 2022 bear market proved that DAOs can survive without a CEO. Now, Berkshire’s filing is a mirror: it shows that even the most successful centralized capital allocator must pivot to tech — but it also shows the limits of that pivot. Abel can’t buy a DAO. He can’t vote on a protocol upgrade. He can only buy shares of Alphabet and hope. The blockchain community should not see this as validation. It should see it as a warning: if you don’t build decentralized governance and treasury management, you will be stuck buying the same centralized assets your technology is meant to replace. Governance isn’t just about voting; it’s about portfolio management. And the best portfolio management is automated, transparent, and community-owned. The question is not whether Berkshire will continue to buy Alphabet. The question is whether the next generation of capital allocators will even need a Berkshire.