Book Value — When the balance sheet is the answer
Worked example: Berkshire Hathaway — 1965–2023
In March 2009, Bank of America traded at $3.14 per share. Its book value per share — the net assets of the bank per share outstanding — was approximately $20. That was a price-to-book ratio of 0.16 times: investors were willing to pay sixteen cents for every dollar of net assets on the balance sheet. Whether that represented extraordinary value or a rational assessment of the risk embedded in those assets was the entire question. An investor who purchased at those levels and held five years earned roughly eight times their money. An investor who bought Citigroup at a similar discount and sold too early earned far less. The balance sheet told both stories — but only to those who could read what was behind the numbers, not just the ratio itself.
The concept in 60 seconds
Book value is the net worth of a business as recorded on its balance sheet: total assets minus total liabilities. Divided by shares outstanding, it gives book value per share — the accounting value of what equity holders own per share.
Price-to-book (P/B) is the market price divided by book value per share. A P/B of 1.0 means the market is paying exactly what the balance sheet says the equity is worth. A P/B above 1.0 means the market is paying a premium to book — typically reflecting intangible value, earning power, or growth not captured on the balance sheet. A P/B below 1.0 means the market values the equity at less than its recorded net assets.
Return on equity (ROE) is the essential companion metric: net income divided by shareholders' equity. The relationship between P/B and ROE is fundamental. A business earning 20% ROE deserves a high P/B because it creates value well above the cost of the equity deployed. A business earning 5% ROE deserves a low P/B — possibly below 1.0 — because it generates returns that barely cover the cost of capital. P/B without ROE is almost meaningless.
Book value is most useful for: financial businesses (banks, insurers, investment firms) where assets are primarily financial instruments that can be marked to market; asset-heavy businesses (real estate, utilities, basic materials) where the balance sheet reflects real economic value; and as a floor valuation in distressed situations. It is least useful for capital-light businesses where most economic value comes from intangibles — brands, patents, software, customer relationships — that are not captured on the balance sheet at all.
Mental model
Think of P/B as the premium the market assigns to the gap between what accounting records and what actually earns money.
If a business's balance sheet shows $10 per share in net assets and the business earns $2 per share annually — a 20% ROE — the market will pay well above $10 for that share, because $10 of equity is producing $2 per year of income. The premium above book value is the market's estimate of the franchise value: the intangible earning power not captured in the assets themselves.
If a business shows $10 per share in net assets but earns only $0.50 per share — a 5% ROE — the market may pay below book value, because owning $10 of assets that generate only $0.50 per year is not worth $10 if those assets could be liquidated and redeployed elsewhere at higher returns.
The simplest model for a justified P/B multiple comes directly from ROE and the required rate of return. If an investor requires a 10% return and the business earns 20% ROE, the justified P/B is 2.0× (20% ÷ 10%). If the business earns 10% ROE — exactly the required return — the justified P/B is 1.0×. Below 10% ROE, a P/B below 1.0 is mathematically rational.
This model has a critical implication: a low P/B is not automatically cheap. It is only cheap if the ROE is likely to improve — or if the assets could be liquidated at book value and the proceeds redeployed better elsewhere. A persistently low-ROE business deserves a persistently low P/B.
Worked example: Berkshire Hathaway, 1965–2023
Warren Buffett used book value growth as the primary internal yardstick for Berkshire Hathaway for most of the company's history — from 1965 until he retired the metric in 2019. His reasoning: early Berkshire was primarily a collection of operating businesses and equity holdings whose market value was reasonably approximated by book. Tracking book value growth therefore tracked whether the business was genuinely compounding economic value.
Over 58 years from 1965 to 2023, Berkshire's book value per share grew from $19 to approximately $373,000 — a compound annual rate of approximately 19.8% per year. This compared to an S&P 500 total return of approximately 10.2% per year over the same period.
The relationship between P/B and intrinsic value evolved as the business did. In the 1970s and 1980s, when Berkshire held significant fixed assets and equity portfolios that were largely mark-to-market, book value was a reasonable proxy for intrinsic value. Buffett would typically buy back Berkshire shares when P/B fell below 1.2 times, indicating his view that intrinsic value was meaningfully above book.
By the 2010s, Berkshire had accumulated enormous amounts of goodwill and intangibles from acquisitions, plus operating businesses whose earning power far exceeded their balance sheet values. Book value increasingly understated intrinsic value. Buffett retired the book value benchmark in 2019 precisely because it had become a poor guide: Berkshire's true earning power — its ability to generate cash from See's Candies, BNSF Railroad, GEICO, and dozens of others — was not visible on the balance sheet.
The lesson: book value was the right metric for Berkshire for 40 years, then became the wrong metric as the business shifted from asset-heavy to earnings-driven. Knowing when the tool applies is as important as knowing how to use it.
| Year | Book Value/Share | P/B Ratio | S&P 500 P/B | What drove the divergence |
|---|---|---|---|---|
| 1975 | $95 | ~0.9× | ~1.1× | Berkshire undervalued; early Buffett track record not yet priced |
| 1985 | $1,643 | ~1.4× | ~1.6× | Track record established; equity portfolio growing fast |
| 1995 | $14,426 | ~1.7× | ~2.8× | Berkshire gaining premium but still below S&P on P/B |
| 2005 | $55,187 | ~1.6× | ~2.8× | Consistent; Buffett's buyback floor around 1.2× |
| 2023 | ~$373,000 | ~1.4× | ~4.1× | Book understates intrinsic value; Buffett retired the metric |
Historical pattern
1929–1932: The depression floor. Many businesses traded below book value during the Great Depression — not because book value was wrong, but because investors doubted the assets could generate any returns at all. Benjamin Graham's entire framework of net-net investing — buying businesses at below net current asset value — was born from this era. Stocks trading at 50–70 cents for every dollar of net current assets (current assets minus all liabilities) provided extraordinary margins of safety when assets were ultimately liquidated or companies recovered.
1970s: Book value as the anchor. Throughout most of the 1970s, the S&P 500 traded at or below book value. The 1974 bear market took the index to 0.8 times book. The combination of low P/B, high inflation eroding nominal asset values, and poor corporate ROE created the conditions for exceptional long-term returns for those who bought quality franchises at or below book.
1990s: The P/B premium expands. As the technology economy moved value into intangibles — software, intellectual property, brand equity — P/B multiples disconnected from historical norms. By 1999, the S&P 500 traded at 5 times book. Most of that premium reflected genuine franchise value not on balance sheets. Some reflected speculation. Distinguishing between the two required looking at ROE, not just P/B.
2008–2009: Financial crisis and book value distrust. Bank balance sheets showed large book values derived from mortgage assets and complex structured securities. When those assets proved to be worth far less than book — sometimes cents on the dollar — P/B below 1.0 was not cheap: it was rational given the uncertainty about what the assets were actually worth. The lesson was that for financial businesses, book value is only as trustworthy as the assets underlying it.
Decision framework
Step 1 — Determine whether book value is the right metric. Book value is most meaningful for financial businesses (banks, insurers), capital-intensive industrial businesses (utilities, manufacturers, miners), and asset-heavy real estate companies. It is least meaningful for capital-light businesses (software, consumer brands, professional services) where intangibles drive almost all the value. Applying P/B to the wrong business type produces systematically misleading conclusions.
Step 2 — Calculate return on equity and the justified P/B. Divide trailing net income by average shareholders' equity. This gives ROE. Divide ROE by your required rate of return. The result is the theoretically justified P/B multiple. A business with 15% ROE and a 10% required return justifies a 1.5× P/B. Any premium above or below that justified level requires explanation: either ROE is expected to change, or the assets are carried at values different from their economic worth.
Step 3 — Assess book value quality. Not all book values are created equal. Goodwill and intangibles from acquisitions inflate book value but may not represent real economic assets — write-downs are common. Financial assets marked to model rather than market may carry significant embedded losses. Hard assets like real property may be carried at historical cost and understated relative to current market value. Adjust book value for these distortions before using it as a valuation anchor.
Step 4 — Check for value-destroying capital allocation. A business trading below book value may be signaling that capital is being misallocated — management is deploying equity into low-return projects that destroy value. In this case, a buyback at below book value is immediately accretive to remaining shareholders. Management's willingness to buy back stock below book value — or to pay dividends rather than reinvest at low returns — is a meaningful signal about how aligned they are with shareholders.
Step 5 — For financial businesses, examine asset quality, not just the ratio. For banks and insurers, P/B below 1.0 demands investigation of the loan book or investment portfolio. What are the credit metrics? What is the non-performing loan ratio? What assumptions are embedded in the investment portfolio's fair value marks? A bank at 0.6 times book with clean, performing assets in a rising rate environment is a very different proposition from a bank at 0.6 times book with deteriorating credit metrics in a recession.
Common mistakes
Applying P/B to capital-light businesses. Microsoft, Google, and most consumer brand companies trade at 8–20 times book value. This does not mean they are expensive. It means their balance sheets massively understate the value of their intellectual property, brand equity, customer relationships, and proprietary technology — none of which appear at fair value on the balance sheet. Using P/B to evaluate these businesses generates entirely wrong conclusions, systematically labeling them as overvalued.
Treating goodwill as equivalent to tangible assets. Goodwill arises from acquisitions — it is the premium paid above the fair value of identifiable assets. It sits on the balance sheet indefinitely (unless impaired) but represents no physical asset, no cash flow in itself, and no liquidation value. A business with $50 per share in book value, $35 of which is goodwill, has a very different balance sheet from one with $50 per share of tangible assets. Always calculate tangible book value (book value minus goodwill and other intangibles) alongside total book value.
Assuming low P/B means distress will reverse. A business persistently trading below book value is often doing so for structural reasons: low ROE, poor capital allocation, industry overcapacity, or an eroding competitive position. The market is saying: these assets as currently deployed are not worth their accounting value. Without a credible thesis for how ROE will improve — management change, industry restructuring, asset sale — a low P/B is a description of a problem, not a solution to one.
Ignoring the cost of book value growth. A management team can grow book value simply by retaining earnings — even at low rates of return. Book value per share growth of 8% per year sounds good; if ROE is 8% and all earnings are retained, that is the mechanical result. It does not mean the management team is creating economic value — they are simply reinvesting at the cost of capital with no excess returns. Book value growth only creates value if the ROE on reinvested capital exceeds the required rate of return.
How VI Stack uses this
Book value and P/B are the fourth valuation lens in Gate 3 — The Forensics. The weight given to this lens depends heavily on the business type.
For financial businesses — banks, insurers, asset managers — P/B is the primary valuation anchor, used alongside ROE and the ratio of tangible book to total book. A bank at 1.0× tangible book with 12% ROE is assessed differently from a bank at 1.0× tangible book with 6% ROE.
For industrial businesses with significant hard assets, P/B is used as a cross-check against FCF yield and EV/EBITDA. The replacement cost of assets — what it would cost to build the business from scratch — is compared to book value to assess whether the balance sheet is overstated or understated relative to economic reality.
For capital-light businesses, P/B is explicitly noted as not applicable as a primary metric. The analysis focuses instead on franchise value — the earning power not captured on the balance sheet — and the return on incremental capital rather than return on total equity.
In Gate 4 — The Pitch — any use of P/B in a target price must specify: (1) which assets have been adjusted from book value and why; (2) what ROE trend justifies the applied multiple; (3) for financial businesses, the tangible book value and asset quality assessment.
What's next
val-05 — Owner Earnings closes the Valuation series by returning to Buffett's most refined concept: the true earning power of a business, after all the accounting is stripped away. Owner earnings synthesizes the lessons of P/E, EV/EBITDA, FCF, and book value into a single, integrated view of what a business is genuinely worth to its owners — and why that number is the right foundation for a long-term investment decision.
vistack.io