Practitioner reads on Berkshire Hathaway’s positions, deals, and annual-letter signals.
Authored by Philip A. Baratelli, CPA, MBA. Former Family Office CFO and public-company CFO; corporate controller and treasurer in earlier roles. Currently a Principal, investing on his own behalf. Owns BRK.B personally; does not advise on Berkshire, TMHC, or any other security.
When Berkshire deploys capital, the Institute publishes the practitioner read when the analysis is ready — deal mechanics, price discipline, library tie-back. Cases ship when they stand up to the depth every other Berkshire Read holds to. We’d rather be late and right than fast and shallow. Hypotheticals test the pattern against named, ranked candidates at the same flagship depth.
Phil Baratelli
CPA · MBA
NYSE Closing Bell · Nov 4, 2003 Armor Holdings (NYSE: AH)
Where journalists land first
The quarterly cash data, pulled together in one place — we’re trying to make it easier to follow than the filings alone.
Berkshire Read is the Baratelli Institute’s recurring practitioner read on Berkshire Hathaway — the deals, the equity-stake changes, the annual-letter signals. Berkshire’s behavior is the closest thing to a public-market apprenticeship the world has, and the Institute’s library is built on the same first principles Buffett has practiced for sixty years. Cases walk an actual Berkshire move with a 30+ page case study, a 10-tab Excel model, a Library Crosswalk, a presentation deck, and a single combined PDF. Strategic Briefs pair to each case as hypothetical analytical exercises — named, ranked Berkshire candidates tested against the same five-fingerprint pattern, with the discipline banner that distinguishes documented fact from Institute-hypothesized structure from Buffett-pattern inference.
Companion to Berkshire Read · The Operator’s Chair
If We Ran Berkshire
Where Berkshire Read walks the deals Berkshire actually did, the Operator’s Chair runs the hypotheticals — the acquisition lenses we’d hunt through if the capital were ours, each idea graded against the same five Buffett filters. The discipline is the product; the candidates graduate into full case studies.
Featured Hypothetical · Operator’s Chair Candidate
If We Ran Berkshire — Acquire NBCU
Comcast's June 29, 2026 announcement to spin NBCUniversal as a tax-free §355 distribution puts a $60-75B integrated entertainment-and-experiences platform inside our addressable set. Under Greg Abel — with BNSF and BHE as the operating template and roughly $300-400 billion of un-deployed cash compounding at T-bill yields — NBCU is the rare opportunity that absorbs meaningful Berkshire capital while opening an entire new investment category. Studios + Streaming + Theme Parks + Sky international + Sunday Night Football reconstructs Disney's architecture in one transaction. The next decade of bolt-ons (Six Flags at cyclical troughs, cruise lines, hotels at REIT-cycle discounts, sports rights at renewal cycles) is what the platform enables. The Institute's editorial position: we would buy NBCU outright if we ran Berkshire.
Featured Hypothetical · Operator’s Chair Candidate · NEW
If We Ran Berkshire — Acquire Copart
Copart is one of only two players in a genuine two-sided-network-effect duopoly — the specific rare event Berkshire waits decades for. Moody’s / S&P is the analog. Coca-Cola / Pepsi is the analog. IAA sits inside RB Global; Copart is the only realistic path to own half the salvage-auction duopoly. Founder Willis Johnson is 79. Returning CEO Jay Adair is his son-in-law. Owner earnings capitalized at 20× produce an intrinsic anchor of ~$30 billion; the sum-of-the-parts (owner earnings + real estate at market + cash and Treasuries) produces a central of ~$38 billion against a $27.8 billion market cap. Zero long-term debt; 18.6× TTM P/E on $1.61 EPS. BNSF (2010, ~$44B) and Precision Castparts (2016, ~$37B) are the acquirer precedents at this scale. Two Berkshire-specific drivers — GEICO integration margin capture and international consolidation at Berkshire capital cost — add ~$245M/year of value creation vs. baseline. Recommended: Scenario B at 1.15× (~$32B EV).
Featured Hypothetical · Operator’s Chair Candidate · NEW
If We Ran Berkshire — Acquire Penske Automotive
Why Berkshire is the natural buyer, if the family ever sells.
Penske Automotive Group (NYSE: PAG) is the third case we make for Berkshire to acquire — and the one where the architecture is already substantially in place. Roger Penske is 89. Berkshire already owns approximately 19% of PAG, a position built patiently over more than a decade that has returned roughly $400 million in cumulative dividends since 2015. The three-entity architecture — PAG (the public consolidator), Penske Corporation (the private family control block), and PTS (the Mitsui joint venture housing truck leasing and logistics) — produces four plausible succession outcomes (status quo continues, family retains, IPO of Penske Corporation, strategic sale), and Berkshire is the natural counterparty in one of them. Post-Mitsui SOTP builds to ~$12.6B in equity value; a strategic sale to Berkshire would require roughly ~$10.2B of fresh capital, net of the 19% already held at cost. The memo does not forecast a sale — it maps the architecture the practitioner shelf should hold in view, and walks the Berkshire-declines counter-case in the same voice as the primary thesis.
Twelve free Baratelli tools curated for the reader who wants to read a business the way Buffett reads one — three-statement coherence, intrinsic value, capital allocation discipline, deferred-tax math, and owner-operator alignment. No login. No paywall.
Berkshire Acquires Taylor Morrison Home Corporation
An $8.5B all-cash acquisition of a Sun Belt production homebuilder — the largest residential-construction deal in Berkshire history, and the first time Berkshire owns a publicly-listed national homebuilder. A practitioner read of the deal mechanics, the Buffett-discipline lens, and what it signals about Berkshire’s residential thesis going forward.
Announced May 31, 2026 · Filed 8-K, May 31 · Published June 1, 2026
Berkshire paid a modest 1.08x book value and 6.6x trailing Adjusted EBITDA, all-cash from the ~$397B war chest, for a Sun Belt homebuilder with a 13-year public track record, an aligned owner-CEO (Sheryl Palmer) who stays in the seat, and an asset-light land book (~54% lots controlled off-balance-sheet). The break from Berkshire pattern is the industry; the discipline is textbook. The clearest read is that Greg Abel is signaling a Clayton-Homes-plus-TMHC-plus-HomeServices site-built housing platform, not a one-off opportunistic buy.
$72.50Per-share cash
~$6.8BEquity value
~$8.5BEnterprise value
24%Premium to 5/29 close
1.08xP / Book
6.6xEV / Adj. EBITDA
8.7xP / FY25 EPS
H2 2026Expected close
Deal terms and multiples as of June 1, 2026 (announcement date May 31, 2026; figures reconciled to the 8-K and the Q1 2026 10-Q). Re-verify against the latest filing before quoting at deadline.
Drawn from the substantive analysis in the case. Each angle is the headline; the case + model + crosswalk supply the reporting backbone.
ANGLE 1
“The cash-and-forever advantage: why Berkshire just beat every PE sponsor at homebuilder M&A”
A financial sponsor (KKR, Apollo, Blackstone, Brookfield) buying TMHC at 8.5x EV would face two constraints Berkshire does not: ~5x leverage on a cyclical homebuilder credit (which the rating agencies and banks will resist) and a five-to-seven-year hold period (which the homebuilder cycle does not cooperate with). Berkshire pays cash, holds forever. That combination is Berkshire’s structural advantage over the sponsor universe in homebuilder M&A — and the TMHC deal is the proof of concept. Case study pp. 9–10 walks the math.
ANGLE 2
“Greg Abel’s site-built platform: TMHC plus Clayton Homes equals a top-10 U.S. builder”
Berkshire has owned Clayton Homes since 2003. Clayton Properties Group (site-built division) shipped ~10,000 site-built homes in 2025 — combined with TMHC’s 12,997 closings, Berkshire vaults from a modest site-built footprint to a top-10 U.S. production builder overnight. This is not an opportunistic single buy; it is the architecture of a platform. Case study pp. 9, 11.
ANGLE 3
“Mid-cycle, not mid-peak: the 1.08x book multiple is the entire signal”
Berkshire paid 1.08x book, 6.6x EBITDA, 8.7x earnings on FY2025 trough-cycle numbers (gross margin compressed 190 bps YoY, backlog down 41.8%, net orders down 9.6%). These are not 2021 multiples. The deal is the textbook Buffett move — a durable, simple business at a price that requires the seller to deliver merely normal operating performance for the buyer to do well. A practitioner reads the multiple as the cycle-call signal, not the price-tag. Case study pp. 6, 9, 12.
PHILIP A. BARATELLI, CPA, MBA · FOUNDER, BARATELLI INSTITUTE
Quotable lines a journalist can paste with attribution
Drawn directly from the case-study analysis. Preferred attribution: “Phil Baratelli, Baratelli Institute.”
The TMHC acquisition is a Berkshire-discipline deal in unfamiliar clothing. The price is textbook; the industry is the surprise.
— Phil Baratelli, Baratelli Institute · on the May 31, 2026 deal announcement
Berkshire’s hold period is forever — or until a different opportunity appears. Forever is the correct hold period for a cyclical land-and-build business.
— Phil Baratelli, Baratelli Institute · on why a sponsor could not have bid this
1.08x book, 8.7x earnings, 6.6x EBITDA — these are not 2021 multiples. These are mid-cycle-trough multiples. Berkshire’s habit is to acquire at exactly this point in the cycle.
— Phil Baratelli, Baratelli Institute · on Berkshire’s price discipline
The clearest read is that Greg Abel is signaling a Clayton-Homes-plus-TMHC-plus-HomeServices site-built housing platform, not a one-off opportunistic buy.
— Phil Baratelli, Baratelli Institute · on the strategic frame
Sheryl Palmer stays. That single fact tells you this is the Buffett pattern — Tony Nicely at GEICO, Pete Liegl at Forest River, Greg Abel at BHE, Kevin Clayton at Clayton Homes. The textbook owner-operator alignment chapter.
— Phil Baratelli, Baratelli Institute · on the management-continuity signal
The cash absorbs ~2% of the $397 billion war chest. The cash-deployment problem at Berkshire argues for more of these deals, not fewer.
— Phil Baratelli, Baratelli Institute · on Berkshire’s next move
Pay a price that requires the seller to deliver merely normal operating performance for the buyer to do well. That is the third great Buffett principle, and TMHC is what it looks like in 2026.
— Phil Baratelli, Baratelli Institute · on margin-of-safety pricing
CONTEXT · SCALE OF THE PARENT
BRK’s River of Cash — The Places Berkshire Funds From, Plus the Two Low-Cost Financings on the Liability Side
Every Berkshire Read case is a transaction. Every transaction gets funded from one of four asset-side places — one big balance that sits on the asset side of the balance sheet (cash), and three flows or trims the quarter delivers (OCF, capex, public-securities sales). Sitting alongside cash on the asset side is the ~$285B public-equity portfolio. And sitting behind both of those, on the liability side, are Berkshire’s two low-cost financings — insurance float and deferred tax liability — that combined to ~$266B at Q1 2026 and quietly fund the broader investment portfolio. The plain-English read: a Berkshire deal almost never has to choose, and the long-run compounding engine is the two low-cost financings sitting underneath the asset stack. The cards below explain each source the way Phil would explain it to a smart friend who isn’t in finance.
THE PROMISE THAT EARNS THIS PAGE
Our goal: make Berkshire’s cash and capital flows easier to follow than the source filings make them on their own — showing the river of cash-flow rollforward, and the running totals of cash, equity portfolio, float, and deferred tax liability — the full balance-sheet picture, with the math and the why. Work in progress — and we will try.
Mentoring at Scale.
THE FOUR SOURCES · WHAT THEY ARE, WHERE THEY COME FROM
Two balances and two flows — the four sources, and what’s actually deployable
Plain-English explanations first. Quarterly numbers below. Reader doesn’t need to be a CFA — just curious.
Balance · Stock
1. Total Cash Balance
~$397.4Bend of Q1 2026 · record high · $58B cash + $339B short-term Treasuries
What it is. The actual money sitting in Berkshire’s account — mostly short-term U.S. Treasury bills. The famous “war chest” — and as of Q1 2026 (Greg Abel’s debut quarter), an all-time high.
Where it comes from. Sixty years of compounding operating profits the company chose not to spend.
How it deploys: totally unconstrained. Buffett can write the check tomorrow morning.
Liability · Low-Cost Financing
2. Insurance Float
~$176.9Bend of Q1 2026 · GEICO + General Re + BHRG
What it is. Insurance premiums Berkshire has collected but hasn’t yet paid out as claims. Berkshire invests it in the meantime.
Where it comes from. GEICO, Gen Re, and the reinsurance group — collected daily, paid out years later. Customers’ money, held legally.
How it works: free money to invest while it sits — regulator-constrained, not available for M&A.
Liability · Low-Cost Financing
3. Deferred Tax Liability
~$88.7Bend of Q1 2026 · tax on unrealized equity gains
What it is. Income tax Berkshire owes the IRS on unrealized gains in the equity portfolio (Apple, BAC, Coca-Cola, AmEx, Chevron). Not yet due, because the gains aren’t sold.
Where it comes from. The “Income taxes, principally deferred” line on the BRK balance sheet. An interest-free loan from the IRS, growing with the portfolio.
How it works: free leverage on the investment portfolio — becomes payable only when Berkshire sells.
Flow · This Quarter
4. Operating Cash Flow
~$10.4BQ1 2026 · recurring every 90 days
What it is. The cash the businesses generated this quarter from operations — BNSF, BHE, GEICO, the manufacturing/service/retail group, the whole shop.
Where it comes from. Customers paying invoices, premiums coming in, railroads moving freight.
How it deploys: recurring. Shows up again next quarter, whether Berkshire spent any of it or not.
Flow · This Quarter
5. Sale of Public Securities
VariableApple trims drove $30B+ in Q3 2024
What it is. Cash raised when Berkshire trims or exits a publicly-traded stock position — the Apple trims of 2024 are the textbook recent example.
Where it comes from. Selling shares on the open market through Berkshire’s portfolio managers.
How it deploys: fully discretionary. Buffett chooses when, what, and how much.
The Berkshire balance sheet at a glance · Q1 2026
~$266B of low-cost financing on the liability side partially supports the ~$682B Cash + Equity Stack.
Assets → what BRK owns
$397B
Cash + ST Tsy
~$285B
Equity portfolio
~$682B
Cash + Equity Stack
↔
Liabilities → low-cost financing
$177B
Float
$89B
DTL
~$266B
Combined
Float (insurance premiums held before claims) + DTL (taxes owed on unrealized equity gains, not yet due) = ~$266B of near-zero-cost financing on the right side, partially supporting the Cash + Equity Stack on the left (the rest is funded by Berkshire’s ~$650B of book equity). This is the part of the Berkshire balance sheet most analyses skip.
INSIDE THE EQUITY PORTFOLIO · Q1 2026 13F
Berkshire’s top equity holdings — what’s actually in the ~$263B book
The balance-sheet pill above shows the equity portfolio as a single number. Here is what makes it up. Five positions — Apple, American Express, Coca-Cola, Bank of America, Chevron — carry roughly two-thirds of the book. The newest name on the leaderboard is Alphabet, the first new tech position since Apple, profiled in Berkshire Read Case 3. Market values and weights are approximate, drawn from the Q1 2026 13F (holdings as of March 31, 2026); verify against the filing before quoting.
The first Alphabet line is the ~57.8M-share open-market stake (~$16.6B), the sixth-largest position. The highlighted pro-forma line counts Berkshire’s total committed Alphabet capital (~$25.4B) once the separate $10B June 2026 placement closes — which lifts the position above Chevron and roughly level with Bank of America, squarely inside the top five. Source: BRK Q1 2026 13F (filed ~May 16, 2026).
COMPANION REFERENCE · COST OF CAPITAL
The Berkshire & holdings WACC pages
The Baratelli WACC Reference publishes company-by-company cost-of-capital walkthroughs. Every Berkshire-adjacent name has its own page, with the practitioner note on what makes the calculation quirky — float and DTL for Berkshire, ownership dynamics for the holdings, Munger’s board seat for Costco.
A share of stock is a share of a business’s profit — so the number that matters is Berkshire’s slice of each company’s earnings, not its share price.
Buffett introduced “look-through earnings” in the 1990 shareholder letter to fix a blind spot in GAAP: Berkshire’s income statement only records the dividends it receives from its stock holdings, not the much larger profits those companies earn and retain on Berkshire’s behalf. Look-through earnings add the retained slice back in. The formula is simple:
Look-through earnings = Berkshire’s % ownership × the company’s net income
(summed across every holding — dividends Berkshire actually pockets plus the earnings retained inside each business)
Worked example from Case 3: a ~0.71% economic stake in Alphabet against ~$132B of net income is roughly ~$0.94B of look-through earnings — the real owner-economics of the position, most of which never appears on Berkshire’s own income statement. It is the lens the Institute uses to read every Berkshire equity position: own the earnings, not the ticker.
THREE WAYS THE TOP FIVE PAY BERKSHIRE · FY2025
Net income, dividends, buybacks — and the slice that belongs to Berkshire
Every dollar these five companies earn gets used three ways: paid out as dividends, spent on buybacks, or retained inside the business. Berkshire benefits from all three — but only one shows up on its income statement. The table below lays out each company’s full-year 2025 figures and Berkshire’s estimated economic ownership; the column on the right is the look-through earnings that ownership represents.
Holding
Net income
Dividends
Buybacks
Est. BRK ownership
BRK look-through earnings
BRK dividends received
BRK look-through buybacks
Apple (AAPL)
~$112.0B
~$15.4B
~$89.3B
~2.0%
~$2.2B
~$0.3B
~$1.8B
American Express (AXP)
~$10.8B
~$2.3B
~$5.3B
~21.8%
~$2.4B
~$0.5B
~$1.2B
Coca-Cola (KO)
~$12.9B
~$8.8B
~$1.0B
~9.3%
~$1.2B
~$0.8B
~$0.1B
Bank of America (BAC)
~$30.5B
~$8.5B
~$21.0B
~7.0%
~$2.1B
~$0.6B
~$1.5B
Chevron (CVX)
~$12.3B
~$12.3B
~$11.7B
~6.0%
~$0.7B
~$0.7B
~$0.7B
Top-five total
~$178.5B
~$47.3B
~$128.3B
—
~$8.7B
~$3.0B
~$5.2B
All figures full-year 2025, $ billions, approximate — verify against each company’s 10-K and Berkshire’s Q1 2026 13F before quoting. Ownership percentages are estimates from Berkshire’s reported share counts against each company’s shares outstanding (Apple ~300M shares; AXP 151.6M; KO 400M; BAC and CVX after 2025–26 trims). Look-through earnings = ownership % × net income; BRK dividends received = ownership % × dividends; BRK look-through buybacks = ownership % × buybacks. The three BRK columns are not additive — dividends and buybacks are funded out of the same net income that look-through earnings already captures in full, so summing them would double-count. Coca-Cola net income is GAAP; buyback figures are gross. Sources: Apple FY25 10-K; AXP, KO, BAC, CVX FY2025 results; BRK Q1 2026 13F.
What actually lands at Berkshire
Dividends are direct. Berkshire’s slice of the ~$47.3B these five paid out — roughly ~$3.0B — arrives as cash and is the only piece GAAP records on Berkshire’s income statement.
Net income is look-through. Berkshire’s economic share of the ~$178.5B these companies earned is about ~$8.7B — nearly three times the dividends. The difference is earnings retained and reinvested on Berkshire’s behalf, invisible to GAAP but real to an owner.
Buybacks compound silently. The ~$128.3B these five spent retiring their own shares lifts Berkshire’s ownership percentage every year — without Berkshire buying a single share or spending a dime. Coca-Cola at a fixed 400M shares has drifted from ~6% to ~9.3% of the company this way. It is the quietest of the three benefits and, over decades, often the largest.
QUARTERLY VIEW · THE LAST 9 QUARTERS
The full balance-sheet view — assets on the left, liabilities on the right, flows in the middle
All figures in $ billions. Cash, Equity Portfolio, Float, and DTL are end-of-quarter balances (stocks) — the four running totals. OCF, Public-Sec. Sales, and Capex are quarter activity (flows). Sourced from BRK quarterly financial reports (10-K, 10-Qs) and 13-F filings; DTL is the “Income taxes, principally deferred” line; equity portfolio is the aggregate quarter-end fair value of public-equity holdings. Last updated 2026-06-04.
Plain-English orientation. The table is grouped to show the balance-sheet logic: ASSETS on the left (green — what BRK owns: cash and the public-equity portfolio) and LIABILITIES next to them (rose — float and DTL, what BRK owes at near-zero cost). The Cash + Equity Stack is partially financed by the liability stack (float + DTL); the remainder comes from Berkshire’s book equity (~$650B) and other liabilities. “OCF” = operating cash flow = the cash Berkshire’s businesses generated this quarter. “Capex” = capital expenditures at the operating subsidiaries. “DTL” = deferred tax liability = what Berkshire owes the IRS on unrealized capital gains in its equity portfolio (an interest-free IRS loan that grows with the portfolio). “Deployable subset” = the slice Abel is actually free to put to work after the $30B reserve and capex — excludes float and DTL (financing for the broader portfolio, not deal dollars) and excludes the equity portfolio itself (already invested).
Delta init. ~$2.6B Alphabet tripled to $15.6B TMHC announced 5/31 (Q2’26)
~$381
DEPLOYABLE-SUBSET TRAJECTORY
$167B → $381B over nine quarters
The capex-deducted firepower envelope (the right-most column above), quarter by quarter. Up ~128% across the period — the cash war chest growing faster than the operating businesses can deploy it.
◊Deals column shows named Berkshire transactions and equity-portfolio moves that Berkshire Read has analyzed (or will analyze). As new cases ship — Delta, Google, future M&A — they appear in the quarter they materially affected cash, float, or the equity portfolio. Deals are part of investment activity, not operating cash flow — they do not change the Deployable Subset formula (the formula already absorbs them via the Public-Sec. Sales and Cash columns). The column becomes part of the franchise mechanic: every case has a quarter and a dollar size in the rollforward. ¶Equity Portfolio = aggregate market value of Berkshire’s public-equity holdings at quarter-end (Apple, American Express, Bank of America, Coca-Cola, Chevron, Kraft Heinz, Occidental, and the long tail). Figures sourced from BRK 13-F filings, 10-Q balance-sheet equity-securities line, and the equity-securities cost/fair-value disclosures in quarterly notes; market value fluctuates with both stock prices and Berkshire’s buy/sell activity, so all figures are flagged approximate (“~”). The Q2’24 step-down reflects the well-publicized Apple trim; the Q4’25–Q1’26 figures reflect continued BAC and Apple trimming alongside new Alphabet and Delta positions. * Q2’24–Q1’25 public-securities sales reflect the well-publicized Berkshire equity-portfolio trimming (notably Apple, which drove the Q2’24 figure; subsequent BAC trims drove later quarters). †Quarterly capex is the recurring sustaining-and-growth investment at the operating subsidiaries — BHE (~$2.5–3B/qtr), BNSF (~$1–1.5B/qtr), and the manufacturing/service/retail group (~$0.5–1B/qtr) — combined ~$4–6B per quarter, ~$19–21B annualized (FY2024 $19.0B, FY2025 $20.9B). It’s a real claim on cash that has to be funded before any dollar is “deployable” for M&A or buybacks. § Insurance float disclosed by BRK at year-end and mid-year (Buffett’s annual letter + Q2 commentary); intermediate-quarter figures are linearly interpolated estimates — flagged as such, not represented as primary-source verified. §§ DTL figures for Q1’24, Q2’24, Q3’24, Q1’25, and Q2’25 are approximate, reflecting the larger unrealized-gain swings around the Apple position peak and trims; verified year-end and last-three-quarter figures sourced directly from the BRK 10-K and 10-Q balance sheets (“Income taxes, principally deferred” line). ‡‡‡DTL = Deferred Tax Liability — income tax Berkshire owes on unrealized capital gains in its equity portfolio. The liability becomes payable only when (and if) Berkshire sells. Until then, it functions as an interest-free loan from the IRS that grows with the portfolio — the second of Berkshire’s two low-cost financing sources alongside float. ‡‡ Q3’24 OCF was depressed by a $15.3B income-tax payment in the quarter (per BRK Q3’24 10-Q MD&A); 9M’24 OCF was $26.0B, which net of Q1’s $10.6B and Q2’s $13.6B leaves $1.8B for Q3’24. ‡Q1 2026 = record-high $397.4B cash + short-term Treasuries ($58.1B cash & equivalents + $339.3B short-term U.S. Treasuries) — Greg Abel’s debut quarter as CEO (Jan 1, 2026 transition from Warren Buffett). Berkshire was a net seller of equities (~$24B sold, $16B purchased — $8B net) yet still initiated Delta (~$2.6B) and tripled Alphabet (to $15.6B); $235M in buybacks — Abel let the cash position grow. “Deployable Subset” equals Cash minus Buffett’s stated ~$30B minimum reserve, plus the quarter’s OCF, plus the quarter’s public-securities net sales, minus the quarter’s capex. Float and DTL are shown but deliberately excluded from this subtotal — they sit on the liability side of the balance sheet as low-cost financing for the broader investment portfolio, not as dollars available for the next acquisition. See the sidebar below for why. Sources: BRK Q1 2026 10-Q filed May 2, 2026 (1stqtr26.pdf; SEC EDGAR); BRK 2025 10-K filed Feb 2026 (EDGAR); BRK 2024 10-K filed Feb 22, 2025 (EDGAR); BRK 10-Qs for Q1’24 (EDGAR), Q2’24 (EDGAR), Q3’24 (EDGAR), Q1’25 (EDGAR), Q2’25 (EDGAR), Q3’25 (EDGAR). Cash + Short-term Treasuries reported as combined per BRK convention. Last verification pass: 2026-06-04.
THE FULL BALANCE-SHEET PICTURE · WHY BRK COMPOUNDS THE WAY IT DOES
Look at the table above. The left side (green) is the slice of what Berkshire owns that’s visible here: cash and the public-equity portfolio (the Cash + Equity Stack). The right side (rose) is the slice of what Berkshire owes that’s visible here: float and deferred taxes. This is not the full balance sheet — it excludes the wholly-owned operating businesses on the left, and book equity plus other liabilities on the right. It’s the slice that teaches the Buffett financing point.
The trick: those liabilities cost almost nothing. Most companies finance growth with debt at market interest rates. Berkshire’s two largest financings cost essentially zero — and most analyses of the company miss one of them entirely.
1. Float (~$177B). Customers’ insurance premiums Berkshire holds before claims are paid. As long as the insurance underwriting at least breaks even, the float is free money to invest. GEICO, Gen Re, and BHRG collect premiums daily; claims are paid out over years. Plain English: customers’ money, held legally, invested in the meantime.
2. DTL (~$89B). Income tax Berkshire owes the IRS on unrealized gains in its equity portfolio (Apple, BAC, Coca-Cola, AmEx, Chevron, Occidental). Not due until Berkshire sells. So Berkshire essentially borrows from the IRS at zero interest, and the loan grows with the portfolio.
Combined: ~$266B of near-zero-cost financing sitting on the liability side. That’s what partially funds the Cash + Equity Stack — the rest is funded by Berkshire’s ~$650B of book equity plus other liabilities (claim reserves beyond float, modest debt at BHE/BNSF, deferred revenue, accruals). Without float + DTL, Berkshire would need ~$266B more equity capital — or it would have to take on real debt at market rates. Instead: ~$397B cash + ~$285B equity portfolio = ~$682B Cash + Equity Stack, with ~$266B of the right-hand financing coming from policyholders and the IRS at essentially zero cost. Note: this is only the liquid-investment slice of Berkshire’s balance sheet — it excludes the wholly-owned operating businesses (BNSF, BHE, GEICO ops, Marmon, Pilot, NetJets, See’s, Lubrizol, Precision Castparts, and dozens more), which carry hundreds of billions more in assets.
Most companies finance growth with debt. Berkshire’s two largest financings are someone else’s money sitting on Berkshire’s balance sheet at near-zero cost. That’s the compounding engine the textbooks rarely explain.
Important caveat — the equity portfolio and the low-cost financings aren’t “deployable” the way cash is. The equity portfolio is already invested. Float backs claim reserves and regulatory insurance capital, so it can’t be pulled for M&A. DTL becomes payable the moment Berkshire sells the appreciated equities. That’s why the “Deployable Subset” column to the left starts from cash above the ~$30B reserve and excludes the equity portfolio, float, and DTL.
Intellectual honesty · what these four numbers aren’t
A caveat — and the Buffett teaching point analysts have missed for 40 years
The four numbers above don’t balance to Berkshire’s full balance sheet. They aren’t supposed to. A sophisticated reader will immediately notice that ~$397B cash + ~$285B equity portfolio = the ~$682B Cash + Equity Stack, while ~$177B float + ~$89B DTL = only ~$266B of liabilities. That leaves a ~$416B gap. The gap is real, and it’s the rest of Berkshire — the wholly-owned operating businesses and book equity that this slice deliberately leaves aside.
What’s missing from the left side. Berkshire’s full assets are much larger than cash + the equity portfolio — they include hundreds of billions in operating businesses (BNSF Railway, Berkshire Hathaway Energy, GEICO’s underwriting platform, Marmon, Pilot, NetJets, See’s, Lubrizol, Precision Castparts, plus dozens more), carried at cost/book and generating their own earnings streams.
What’s missing from the right side. Berkshire’s full liabilities include claim reserves beyond the float number, modest long-term debt at the parent and at BHE/BNSF, deferred revenue, accrued expenses, and other items. And Berkshire’s book equity — what shareholders actually own — is roughly $650 billion as of Q1 2026. That equity is the largest single line on the right side and the reason the four numbers above don’t balance.
So what is the table above? It isolates one specific slice of the balance sheet — the slice Warren Buffett has been explaining for 40 years that most analysts still don’t grasp. A giant pool of low-cost financing (float + DTL) sitting on the right side, funding a giant pool of liquid investments (cash + the public-equity portfolio) on the left side. The financing costs nearly nothing. The investments generate returns. That spread is the compounding engine.
It is not the whole company. The operating businesses matter enormously, and book equity is the foundation. But this slice is the part Buffett wants you to understand — because it’s the part that makes Berkshire’s reported returns on equity look better than they “should” on the equity alone. The assets are larger than what the equity by itself could ever buy, because policyholders and the IRS are quietly funding the difference at essentially zero cost.
If a money manager you’re listening to talks about Berkshire’s cash position without explaining that the cash sits next to ~$266 billion in near-zero-cost liabilities funding ~$285 billion of stocks — they are missing what Buffett has been teaching since the 1980s annual letters. The four numbers above aren’t a balance sheet. They’re the slice of the balance sheet where the magic happens.
SCALE CHECK · AGAINST THE FULL RESOURCE
Berkshire Read cases against the deployable subset
Now that the four sources are on the table — float is properly excluded from the deal-funding pool, and the quarter’s capex is properly deducted — each Berkshire Read case lands against the truly deployable ~$381B subset. The percentages are still tiny. Q1 2026 baselines used throughout (record-high $397.4B cash position).
TMHC acquisition Case 1 · announced May 31, 2026 · all-cash
~$8.5B
~2.2% of deployable subset
The largest residential-construction deal in BRK history is ~2.1% of the cash position alone — BRK could fund it without touching float, without trimming a single public security, and from less than one quarter’s operating cash flow. Against the ~$381B Q1 2026 deployable subset (cash above the $30B reserve + OCF + net sales − capex): ~2.2%.
Alphabet (Google) position Case 3 · published · tripled in Q1 2026 13F
~$16.6B
~4.4% of deployable subset
Berkshire’s first large tech position after Apple — ~58M shares, ~$15.4B cost basis (~$16.6B market value), a top-five-caliber holding. Pro forma for the separate $10B June 2026 placement, total committed capital is ~$25.4B (~6.7% of the deployable subset). Funded comfortably from a fraction of the cash balance. Read Case 3 →
Delta Air Lines position Case 4 · published · new ~6.1% stake, passive 13G
~$2.65B
~0.7% of deployable subset
A contrarian re-entry six years after Buffett exited the airlines — Berkshire’s 14th-largest holding. Less than one percent of what BRK could genuinely put to work in a single quarter, and smaller than a single quarter’s capex line. Read Case 4 →
The punchline. BRK’s Q1 2026 genuinely deployable capital is roughly ~$381 billion — $397.4B in cash & short-term Treasuries (an all-time high, Greg Abel’s debut quarter as CEO) minus Buffett’s ~$30B minimum reserve, plus $10.4B in this quarter’s operating cash flow, plus ~$8B in net public-securities sales (~$24B sold − ~$16B purchased, including new Delta and tripled Alphabet positions), minus the ~$5.0B of quarterly capex that BHE, BNSF, and the MSR subsidiaries consume to keep the operating engine running. Float (another $176.9B) sits behind the insurance subsidiaries and isn’t freely available for deal funding — and Berkshire’s deferred tax liability on unrealized equity gains ($88.7B) sits behind the equity portfolio. Combined, those two low-cost financings on the liability side of the balance sheet total ~$266B — the engine behind the long-run compounding, separate from the deal-funding pool. The TMHC acquisition is ~2.2% of the deployable subset. The Alphabet position is roughly ~4.4% (~6.7% pro forma for the $10B June placement). Scale is still the story Berkshire never wants you to forget — we just want to frame it honestly, capex and all.
Berkshire publishes the inputs to all four sources quarterly — and the quarterly capex that has to come out before any dollar is truly deployable. Each Berkshire Read case lands as a line against the post-capex deployable subset. The point isn’t that individual deals are big — it’s that the river never stops, and the deployable subset alone is enough to make Berkshire’s biggest recent deals look like rounding errors.
CONTEXT · THE FLOW VIEW
BRK’s River of Cash — Quarterly Flow Rollforward
The four-source view above shows the balances sitting on the balance sheet at the end of each quarter. This view shows the flows — the operating cash that came in each quarter and where it went. Two views, one story: the balances are the result; the flows are the engine.
HOW TO READ THESE TWO VIEWS TOGETHER
Stocks above, flows below — both matter, neither alone is enough.
Quarter-end balances (the four-source table above) answer: how much could BRK theoretically pull from each pocket at the end of last quarter? Useful for sizing the war chest. Doesn’t tell you where the money came from or where it went.
Quarterly flow rollforward (the table below) answers: how much operating cash did BRK generate this quarter, and what did Buffett do with it? Useful for seeing the engine in motion. Doesn’t show the accumulated war chest.
Together they show what a real Berkshire Read needs: the standing reserves and the recurring inflows. Each Berkshire Read case is one line against both.
QUARTERLY ROLLFORWARD · LAST 9 QUARTERS
Operating cash flow, uses, and net cash after uses
All figures in $ billions. Approximate — sourced from BRK 10-Q filings; recent quarters directional pending final filing.
Quarter
Operating Cash Flow
Capex (BNSF + BHE)
Share Repurchases
Net Equity Purchases
Named M&A
Net Cash After Uses
Q1 2024
$10.0
(4.4)
(2.6)
(1.9)
—
+1.1
Q2 2024
12.2
(4.7)
(0.3)
(3.8)
—
+3.4
Q3 2024
10.9
(4.6)
0.0
(34.6)*
—
(28.3)
Q4 2024
9.9
(4.8)
0.0
(7.0)*
—
(1.9)
Q1 2025
10.7
(4.5)
0.0
(1.5)*
—
+4.7
Q2 2025
11.5
(4.6)
0.0
(0.5)
—
+6.4
Q3 2025
11.8
(4.7)
0.0
+2.0
—
+9.1
Q4 2025
11.2
(4.8)
0.0
+1.0
—
+7.4
Q1 2026
11.5
(4.7)
0.0
+0.5
—
+7.3
9-quarter total
~$99.7
~(42.8)
~(2.9)
~(45.8)
—
~$8.2
* Q3’24–Q1’25 net equity purchases reflect the well-publicized Berkshire equity-portfolio trimming (notably Apple). Italicized rows (Q2 2025 onward) are directional estimates pending each quarter’s 10-Q filing. Source: BRK 10-Q quarterly filings; figures approximate where exact data not yet confirmed.
SCALE CHECK · AGAINST QUARTERLY OPERATING CASH FLOW
Berkshire Read cases against one quarter’s OCF
A second lens: each case sized purely against the ~$11.5B quarterly operating cash flow. This is the engine in motion — the recurring number that shows up again next quarter whether Buffett spent a dollar of it or not.
TMHC acquisition Case 1 · announced May 31, 2026 · all-cash
~$8.5B
~74% of 1Q OCF
Largest residential-construction deal in BRK history — yet absorbed by roughly three-quarters of a single quarter’s operating cash. The deal doesn’t even outrun the engine.
Alphabet (Google) position Case 3 · published · newest large BRK position
~$16.6B
~144% of 1Q OCF
Berkshire’s largest tech position after Apple. The open-market stake is roughly one-and-a-half quarters of operating cash — funded alongside Apple trims, not in place of the engine.
Delta Air Lines position Case 4 · published · new ~6.1% passive stake
~$2.65B
~23% of 1Q OCF
A position roughly one-quarter the size of a single quarter’s cash river. Smaller than a single quarter’s capex line.
Berkshire publishes operating cash flow quarterly. Each Berkshire Read case lands as a line on this rollforward. The point isn’t that individual deals are big — it’s that the river never stops.
EDITORIAL CADENCE · WHAT’S NEXT
The Berkshire Read backlog
Every Berkshire-relevant news cycle gets a Case (if it’s a real move) or a Brief (if it’s a pattern test). The franchise publishes when the analysis is ready — not on a calendar schedule, and not on a turnaround clock. We’d rather be late and right than fast and shallow. Six Cases are live; here’s what’s published and what’s in queue.
Published
Six Cases live — Taylor Morrison, Delta, Alphabet, Coca-Cola & more
The franchise now spans an $8.5B acquisition and four equity-stake reads: Taylor Morrison (TMHC), the $8.5B all-cash flagship; Delta Air Lines (DAL), the contrarian airline re-entry; Alphabet (GOOGL), the first tech Read after Apple; and Coca-Cola (KO), the 37-year permanent holding. Each ships with a case study, IC memo, 3-statement model, and deck — all free.
Working journalists are the Berkshire Read franchise’s primary readership. The Institute publishes for the journalist who has 30 seconds to verify the source, 5 minutes to extract a usable pull-quote, and 15 minutes to download a press kit before filing. Below is everything you need.
Press contact · responsive to deadline reporters
Phil takes background, on-the-record, and on-tape conversations. Story-idea hotline runs through the same address.
The Three Masters’ Library is the free companion — the books Buffett, Munger, and Ackman actually read, annotated with why each one matters and in what order, with reader paths for different goals. The bibliography behind every Berkshire Read; yours at no cost. Free.
Want a classic in full? The Reading Room hosts public-domain texts on money and markets, complete and free — starting with Carnegie’s Gospel of Wealth.
See how the Institute thinks — $99
First Principles of Master Investing is the show-the-math synthesis behind every Berkshire Read — how Buffett, Munger, Ackman, and the Outsiders CEOs actually reason, each principle paired with a named case and an intrinsic-value walk. Where the reading list gives you the books, this gives you the method. A window into the analytical method, if you want it — not a step you have to take. $99 introductory ($249).
One-click press kit
TMHC_Press_Kit.zip · ~1.2 MB. Includes all 7 TMHC case files, the 1-page About summary, embargo & republication terms, high-res author photo, and Institute wordmark (SVG + PNG).
Quote, paraphrase, link freely. Wholesale PDF republication requires written approval. Embargo & republication terms (PDF). Preferred attribution: “Phil Baratelli, Baratelli Institute.”
High-res author photo
Phil_Baratelli_NYSE_Bellringer_HighRes.jpg · ~325 KB. Phil ringing the NYSE Closing Bell, November 4, 2003, as Corporate Controller and Treasurer of Armor Holdings, Inc. (NYSE: AH).
