My Berkshire Hathaway Thesis

After over two months of research, here's my final thesis on Berkshire.

We often say Flank is the gym for intelligent investors - I really felt that here. This was like training, and completing, a 500 lb deadlift. Feels really great.

My Berkshire Hathaway Thesis

I value Berkshire Hathaway at roughly $1.1 trillion with no margin of safety, and about $935 billion with a 15% margin of safety.

I believe this is conservative, and am considering buying a current prices (currently $1.04T).

Disclosure: I've held a small position in Berkshire for over a decade. I'm considering adding to my position.

Why Berkshire Is Hard to Value

Most investors look at Berkshire's $89 billion in 2024 GAAP net income, slap a P/E multiple on it on it, and call it a day.

That's not how Buffett told us to value this company. And honestly, after spending weeks running Berkshire through the Flank methodology, I understand why.

The usual metrics don't work here. Revenue jumps when Berkshire acquires a business, even if no value was created (which is why when a company utilizes a M&A growth strategy, we need to be aware of the fuckery that can be).

Net income swings wildly because accounting rules force unrealized market gains and losses through the income statement. In 2024, investment gains alone were $52.8 billion.

In 2022, investment losses were $67.9 billion. That volatility tells you about markets, not about operating performance.

Free cash flow gets distorted by insurance float dynamics, and total equity in absolute terms says nothing about whether shareholders are getting richer per share.

Buffett addressed this directly in his 2018 chairman's letter. He said to think of Berkshire as a forest made up of five "groves." Four are asset-laden groves you can value directly, and the fifth (insurance) finances the others through float. The whole is worth more than the sum because capital moves freely across the groves.

So that's how I valued it. Grove by grove.

The Five Groves

Grove 1: The Non-Insurance Operating Empire

This is BNSF railroad, Berkshire Hathaway Energy, and a massive collection of manufacturing, service, and retail businesses. Precision Castparts. Lubrizol. Clayton Homes. Duracell. Pilot Travel Centers. Forest River. McLane. NetJets.

Over 50 non-insurance operating businesses employing hundreds of thousands of people.

These businesses don't show up at their real value on the balance sheet. PP&E for railroad, utilities, and energy alone is $175 billion, carried at cost minus depreciation, not at what these franchises are actually worth as cash generators.

You can't wake up and decide to build a second BNSF or a parallel utility grid. Even with the money, you don't have the permits, the right-of-way, the regulatory approvals, or the decade it would take.

In 2024, stripping out investment gains, Berkshire earned about $55.7 billion pre-tax from operating businesses, up 29% from $43.3 billion in 2023. That improvement happened even though GAAP net income went down year over year (hence the need for the grove valuation)

I assumed conservative growth rates for my valuation: 2% for BNSF (below their 3-year CAGR of 2.49%), 6% for BHE (well below the 19.5% CAGR that includes wildfire accrual distortions), and 1% for the manufacturing, service, and retail bucket (its 3-year CAGR is only 0.71%, though I expect improvement). That puts Grove 1 at roughly $481 billion.

Grove 2: The Public Stock Portfolio

Berkshire's equity portfolio ended 2024 at $271.6 billion. The top five holdings (Apple, American Express, Coca-Cola, Bank of America, Chevron) account for 65% of the portfolio. Buffett is not afraid of concentration.

The big story in 2024 was that Berkshire was a massive net seller. They purchased about $9 billion in equities and sold roughly $143 billion. That's $134 billion of net selling, mostly Apple. This tells us Berkshire wasn't finding opportunities at prices it liked, and was deliberately moving capital from Grove 2 into Grove 4 (cash and T-bills).

I didn't assume any growth for these holdings in my valuation, which is ultra-conservative. I just took fair value and subtracted deferred taxes on unrealized gains. That puts Grove 2 at roughly $253 billion.

Grove 3: Shared-Control Businesses

This is the smallest grove. Kraft Heinz, Occidental Petroleum, and Berkadia. These are equity-method investments where Berkshire owns a significant stake but doesn't fully control the business. Combined book value is about $31.1 billion.

These businesses are not performing particularly well. I assumed 0% growth, which is conservative but reflects current reality. That puts Grove 3 at roughly $15 billion.

Grove 4: The Fortress

Berkshire ended 2024 with roughly $321 billion in cash, cash equivalents, and U.S. Treasury Bills (net of a $12.8 billion T-bill purchase payable).

This is not laziness. This is strategic. Buffett wants a balance sheet that can survive a financial crisis, a mega-catastrophe insurance event, and a broken credit market simultaneously, without selling anything important at the wrong time.

And in today's rate environment, this cash is no longer dead weight. "Interest, dividend and other investment income" was $21.8 billion in 2024, up from $15.8 billion. The fortress is producing meaningful income.

More importantly, this is Berkshire's optionality. When valuations get extreme in either direction, cash lets Berkshire do deals when others can't. I assumed no growth on this grove, not even accounting for the T-bill coupon yield - which will, in my view, keep this more aligned to the real purchasing power of that capital. That puts Grove 4 at roughly $390 billion.

Grove 5: The Engine (Insurance Float)

Grove 5 is not one of the four "asset-laden" groves. You don't value it as a standalone pile.

Its economic value is the structural funding advantage that lets Berkshire own more of Groves 1 through 4 without normal financing costs.

By year-end 2024, Berkshire's insurance float was approximately $171 billion. Float has grown from $46 billion to $171 billion over two decades, and when underwriting is disciplined, the cost of that float is zero or even negative.

In 2024, underwriting profit was roughly $11.4 billion on $88.3 billion of premiums earned, an 87% combined ratio.

Buffett described it perfectly: "This collect-now, pay-later model leaves P/C companies holding large sums that will eventually go to others. Meanwhile, insurers get to invest this float for their own benefit." When underwriting is profitable, Berkshire is literally being paid to hold other people's money and invest it.

This is their hidden engine. Berkshire invests not just shareholder equity but a massive pool of float. That's why the whole is greater than the sum of the parts.

The Moat: Why This Can't Be Replicated

I spent a lot of time on this question because it felt a bit like moat inception. Buffett coined the term "economic moat" in the first place.

Berkshire's moat is a capital allocation flywheel built on four reinforcing advantages:

Structural access to cheap capital. Insurance float, tax float, and retained earnings give Berkshire a funding cost that no competitor can match. The cost of float was negative in both 2023 and 2024. Berkshire is being paid to hold investable capital by policyholders, not investors.
A reputation-driven deal moat. Berkshire is the buyer of choice. Sellers care about permanence, autonomy, and certainty of close. Private equity has finite fund lives. Strategic acquirers need synergies and integration. Berkshire offers something different: they buy you, they leave you alone, and they never sell. This advantage compounds over decades. More successful deals attract more inbound opportunities.
Scale and barriers to entry. BNSF's rail network can't be replicated. BHE's utility territories are regulated monopolies. Clayton Homes has 48% market share in manufactured housing. These are not businesses where a new entrant can show up with a good idea and compete. The capital requirements and regulatory moats are massive.
Resilience in bad markets. In 2022, Berkshire was up 4% while the S&P 500 dropped 18.1%. In 2008, Berkshire fell 31.8% versus the S&P's 37%. This isn't "never goes down." Rather, it shows structural resilience. From 1965 to 2024, Berkshire compounded at 19.9% versus 10.4% for the S&P 500. If Berkshire's stock price dropped 99%, it would still have beaten the index since 1965. That's the power of compound interest over 60 years.

There is only one Berkshire. This could not be replicated in the next 50 years by anyone. I'm pretty certain of that.

The Succession Question: Greg Abel

Greg Abel became CEO on January 1, 2026. This is the single most important question for any Berkshire investor right now: can Abel be Buffett 2.0?

The honest answer is that he doesn't need to be. Berkshire today is an empire of operating businesses that together employ over 392,000 workers. The job is less about stock picking and more about running businesses and deploying capital on big deals.

Abel's background fits this. He joined what became Berkshire Hathaway Energy in 1992, became its chief in 2008, and grew its annual earnings from $122 million to $3.4 billion through continuous reinvestment. He's an operator. Executives who work with him say he scrutinizes inventory levels, cash flows, capex, and EBITDA more closely than Buffett ever did.

Buffett himself said at the 2024 AGM: "I would leave the capital allocation to Greg. He understands businesses extremely well, and if you understand businesses you understand common stocks."

Abel is more hands-on than Buffett, who was famously hands-off with subsidiaries. That means more structured performance conversations with CEOs, more attention to operating metrics, and more willingness to intervene when things go wrong. That could be a net positive for a company this large and this diverse.

The risk is real though. Berkshire's deal moat is reputation-based, and some of that reputation is personal to Buffett. Will sellers still see Berkshire as the premium buyer? Will the culture of decentralized autonomy survive a more operational CEO? These are open questions. Abel seems like the right person, but execution still matters.

My sentiment on leadership: slightly bullish. I wish I knew more about him, but the information is limited, and the proof will be in the results.

What Would Make Me Wrong

The bear case on Berkshire is real and I take it seriously:

The law of big numbers. Berkshire's 1-year revenue CAGR is only 1.9%. At this scale, growth inevitably slows. The explosive compounding of the Buffett era may not repeat under any CEO.
Insurance tail risk. Mispricing or extreme catastrophe regimes could raise the cost of float. Berkshire itself stresses that danger always lurks in insurance. The California wildfires are a reminder that these risks are not theoretical.
GEICO's competitive position. Progressive has been executing better on pricing, segmentation, and telematics. GEICO has been rebuilding profitability and systems, often at the expense of growth. Berkshire is #3 in personal auto and the gap with Progressive has been widening.
Culture drift post-Buffett. The decentralized, hands-off management style is central to how Berkshire operates. If that changes too much under Abel, the "buyer of choice" advantage could erode.
The cash pile is also a question mark. $321 billion sitting in T-bills is safe and producing income, but if Berkshire can't find opportunities to deploy it, that capital earns below what a concentrated investment would return. Optionality has value, but it's not infinite.

My Bottom Line

For 2025, I value Berkshire at roughly $1.1 trillion before margin of safety. With a 15% margin of safety, that's about $970 billion. I believe this is too conservative given that I assumed no growth on the stock portfolio, no growth on cash, 0% on shared-control businesses, and completely ignored Grove 5's compounding power.

The operating businesses got 29% stronger in 2024 while GAAP net income declined.

The balance sheet is a fortress. The insurance engine continues to generate low-cost or negative-cost float.

The moats across rail, utilities, and insurance are wide and durable.

Five days before I finished this analysis, Greg Abel appeared on CNBC and told investors Berkshire would be reinitiating their buyback program, which had been suspended since 2024.

Berkshire only buys back its own stock when the CEO thinks shares are undervalued. My analysis lining up with management's view gives me confidence.

This is a company optimizing for flexibility and long-term compounding, not short-term cash flow or earnings beats. That's exactly the kind of business I want to continue to own.

Sentiment: Bullish.