Real assets, balance sheets and where the opportunities are hiding

In the previous article I made what some readers thought was a strong claim: in a zero-ERP, high-debt world, contrarian value isn’t one of several valid approaches — it’s the only one where the arithmetic gives a real expected edge. The follow-up question came back almost immediately, by email and on Twitter: fine, but how? How do you actually find those contrarian opportunities when every screen, every newsletter, every TikTok is pointing at the same seven stocks?

The answer is older than the internet, less glamorous than AI, and almost entirely absent from financial media: read the balance sheet. Not the income statement — the balance sheet. The snapshot, not the film. The place where the things the market has been ignoring for years live, because none of them fit in a Bloomberg headline.

First, what counts as a “real asset” — and what doesn’t

Worth being precise because the word “real” gets used in a hundred ways. In this context it means one specific thing: a physical, tangible, finite asset that exists independently of trust in a central bank or a counterparty. Land. Buildings. Machinery. Physical inventory. Oil, gas, copper, gold reserves. Forests. Infrastructure — ports, pipelines, power grids. Things you can touch, photograph, and if the whole financial system blows up tomorrow, will still be there.

What isn’t real, by contrast: cash (it’s a government IOU), bonds (a government IOU with a coupon), shares of asset-light service businesses, and especially — this matters — intangibles on the balance sheet: goodwill, brand, capitalised software, customer relationships. Those are, at heart, promises of future cash flow, not things. In a confidence crisis, promises are worth exactly as much confidence as exists.

The distinction matters because the world is slowly but steadily shifting from an era when intangibles ruled back to one where tangibles do. The last fifteen years belonged to Microsoft, Visa, Adobe — businesses with near-empty balance sheets and huge value built on future cash flows discounted at zero rates. When rates settle at 4–5% and global debt sits where it does, that same model discounts less generously. What discounts better is the stuff the ground doesn’t swallow.

Why the balance sheet matters more now than it has in decades

Ninety percent of the financial analysis you see anywhere obsesses over the income statement: revenue, EBITDA, EPS, margins, growth. The balance sheet is treated as a boring appendix, almost pure accounting. That worked for a generation because in the ZIRP era the only thing that mattered was the future growth of earnings — the balance sheet was beside the point.

Not anymore. In an environment of high rates, sticky inflation and peak debt, two questions become critical — and both are answered on the balance sheet, not the P&L: (1) what does this company actually own, right now? (2) can it survive if the next five years are worse than the last five? If you only look at Tesla’s or Nvidia’s P&L you get excited; if you look at the balance sheet of a European industrial small cap with net cash worth 40% of market cap, you realise the market sometimes leaves €100 bills on the floor and looks the other way.

Five places on the balance sheet where real assets hide

This is the meat of the article. If you keep one thing, keep this. Accounting — by design, not by accident — hides the real value of physical assets. The gap between what the balance sheet says and what those assets actually fetch is the contrarian’s gold mine.

1. Land at historical cost

The worst-kept secret in accounting: under US GAAP, land is carried at the price you paid for it, never depreciated, and almost never revalued. There are companies sitting on land bought in 1962 at 1962 prices. Under IFRS you can revalue with the “revaluation model,” but 90% of companies use the cost model out of laziness or to avoid equity volatility. Result: balance sheets carrying land at values frozen forty years ago, while market prices have multiplied by ten.

Concrete examples: there are US railroads sitting on thousands of hectares of 19th-century right-of-way recorded at almost zero. There are Japanese companies with properties in central Tokyo bought in the 1960s. There are European supermarket chains whose real-estate value, properly revalued, exceeds the entire market cap of the business. The market prices them as mediocre retail operations; you price them as hidden landlords.

2. PP&E below replacement cost

Property, Plant & Equipment is carried at historical cost minus accumulated depreciation. But depreciation is an accounting fiction, not an economic one. A steel mill built in 1985 for $100m sits today on the balance sheet at maybe $20m (after 40 years of depreciation). The cost to build the same mill today might be $500m thanks to inflated materials, labour and permits. The gap — $480m — is hidden real value.

This is one of the classic arguments for cyclical investing: when an industry is at the bottom of its cycle (refining, chemicals, cement, paper, shipping), share prices fall below the replacement cost of the assets. Nobody is going to build a new refinery while existing ones trade at 30% of what it costs to build. That imbalance eventually corrects — supply doesn’t grow, demand eventually does, margins recover.

3. Inventory with hidden LIFO reserves

US-only — IFRS banned LIFO years ago — but big where it applies. LIFO means the most recent, most expensive costs flow through the income statement, leaving old cheap inventory on the balance sheet. In a commodity business like an oil major or a metals processor running LIFO, inventory on the books may be priced at 1990 levels. The footnote in the 10-K reads something like “LIFO reserve: $850m” — that’s $850m of real value not reflected in book equity.

4. Investments in associates and subsidiaries

When a company owns under 20% of another, it sits on the balance sheet as an investment “at cost” or “at fair value.” When ownership is 20-50%, equity-method accounting applies — initial value plus a share of cumulative profits. Here’s where the magic happens: a conglomerate like Berkshire Hathaway, Investor AB, Exor, or any Latin American holding company carries its stakes in listed companies at a number that has very little to do with the current market price of those same shares.

European holding companies have traditionally traded at 20-40% discounts to their net asset value (NAV). If you buy Exor at 75 when NAV is 100, you’re buying Ferrari, Stellantis, CNH and Juventus at 75% of their open-market value. Why does the discount exist? Liquidity, holding costs, distrust of management. But discounts do close — mergers, buybacks, spin-offs — and the contrarian who came in at 75 ends up at 100.

5. Net cash above market cap

The rarest of the five, and the most obvious when you find one. There are companies — especially Japanese, Korean and some European industrial small caps — with more net cash (cash minus debt) than their entire market cap. Translation: the market is paying you to take the business off its hands. You pay 80 for the company, find 100 in cash inside. The operating business comes free, with change.

It sounds impossible and yet it exists. Japan had more than a thousand listed companies with net cash above market cap in the late 2010s. The TSE corporate reform of 2023 reduced the count substantially, but several hundred remain. Korea shows similar patterns. And in Europe, in beaten-down sectors like shipping, industrial supplies or specific chemicals, examples appear periodically.

The screens that surface these companies

If you want to build a candidate list in an afternoon, these are the five screens that pay the most. Any reasonable tool (Stockopedia, Koyfin, even a free Yahoo Finance screener) will support them.

• Price / Book < 1 — the classic. Company trades below book value. Even better when the assets are real and have been recently revalued.
• Price / Tangible Book < 1 — strip goodwill and intangibles. More demanding; when it passes, usually signals real hidden value.
• EV / Replacement cost < 0.7 — the most useful for cyclical industries. If rebuilding the asset base costs $1bn and the enterprise value is $600m, the floor is close.
• Net cash / Market cap > 50% — net cash above half of market cap. Instant margin of safety.
• Discount to NAV > 25% — specific to holdings, REITs and closed-end funds. Buying a dollar of assets for 75 cents.

None of these filters alone tells you anything definitive. Each list always has 200-500 names. The actual work — the work that separates an investor from someone who just runs screens — is reading the accounts, one by one, and weeding out the traps.

Where to look in May 2026

These are the sectors and geographies that, as of today, are systematically cheap relative to their real assets. Not individual recommendations — corners of the market where the contrarian finds work:

• European REITs — many trading at 50-70% of NAV. Brexit, high rates, work-from-home and regulatory fear have crushed them. When rates stop or fall, the discount closes.
• Japanese small caps with net cash — fewer than in 2018 after the TSE reform, but hundreds remain. Some with net cash at 60-80% of market cap.
• UK property — off the international radar since Brexit. Houses, offices, retail. Triple-discounted bombshells.
• Mining and energy with proven reserves — reserves carried at cost vs current copper, gold, uranium prices. Massive gap. Cyclical, yes, but asymmetric.
• Shipping — vessels on the books at 30-40% of replacement cost. Hated for a decade. The global order book is empty now and supply isn’t growing.
• Timberland and forestry — forest land valued at decades-old cadastral assessments. Defensive inflation hedge.
• European holding companies at NAV discount — Exor, Investor AB, Wendel, Pargesa, Sofina, Eurazeo. Persistent 20-40% discounts.

The trap you must avoid: value traps

The oldest problem in value investing: cheap can stay cheap, or get cheaper. Many companies that pop up on these screens are cheap because they’re bad — the market is right, they’re not opportunities. A miner with big reserves but operating costs above the copper price is worth zero in the next down cycle. A REIT with properties in emptying postcodes never recovers them. A holding whose CEO destroys capital year after year keeps its NAV discount forever.

How to filter the traps — three quick questions that eliminate 70% of candidates:

• Has the company generated positive free cash flow in at least 3 of the last 5 years? If no, assets are being consumed, not preserved.
• Does management own meaningful stock (insider ownership > 5%)? Without skin in the game, the NAV discount is forever.
• Is there a visible catalyst for the gap to close? Aggressive buybacks, rising dividend, spin-off, hostile bid, management change. Without one, you can wait ten years.

The five-step practical framework

1. Start with the balance sheet, not the income statement. Identify the real assets. Compute tangible book and compare against market cap.
2. Read the footnotes. That’s where LIFO reserves, pending property revaluations, land at historical cost, and stakes at acquisition cost live. The face of the balance sheet lies; the notes tell the truth.
3. Filter with the three anti-trap questions. Positive FCF, insider ownership, visible catalyst. If any fails, pass.
4. Calculate the margin of safety. How far does the asset have to fall before you lose money? 40%+, keep going. 10%, that’s not value, that’s speculation.
5. Diversify across 15-30 names and be patient. Some will stay dormant forever, others will explode upward. The portfolio as a whole, not each individual name, generates the return.

Why almost nobody does this

If the work is so obvious, why doesn’t everyone do it? Because it’s slow, boring and unglamorous. Reading 50 annual reports doesn’t generate content for Twitter. Waiting three years for an industrial small cap’s NAV discount to close doesn’t produce dopamine. Mentioning Exor at dinner impresses nobody. And above all: no professional manager can explain to his client why he owns a Korean timberland operator instead of Nvidia.

That institutional discomfort — the career risk of being wrong while looking different — is exactly what keeps these discounts open. If everyone did this, the discounts wouldn’t exist. They exist because most professionally managed money will never touch these positions, fearing how it looks in a quarterly report. The patient retail investor, with no boss asking every ninety days why returns lag the benchmark, has a huge structural advantage here.

Closing the loop

The previous article left the uncomfortable conclusion: with the ERP near zero and debt where it is, contrarian value is the only strategy with real mathematical edge. This article is the answer to the logical follow-up: how do you actually do it? The answer fits in one sentence — read the balance sheet, find real assets below replacement cost, filter the traps with three questions, be patient.

Meanwhile, the noise. AI. The earnings calls of the magnificent seven. The next FOMC. The macro news that changes every week. All of it happens, all of it gets discounted before you arrive. The only thing that doesn’t get discounted is the patient work of reading balance sheets nobody else wants to read. It’s boring, yes. But it’s where the actual money — the kind that doesn’t evaporate with Powell’s next press conference — is made.