Robotti & Company is a New York deep-value boutique founded by Bob Robotti in 1983, specializing in left-for-dead cyclical industries — energy services, building products, shipping — with multi-year holding periods and occasional activist letters. It manages about $650m; Bob is regarded as one of the most steadfast Graham-tradition cyclical value hunters.
This letter says the real bottleneck for AI is not code or money, but physical things like electricity, copper and steel. The manager is optimistic about real-world assets and cautious on popular AI stocks. Key picks: Subsea 7—merging with Saipem to create an offshore-energy leader, with a big dividend for shareholders; Saipem—gains a stronger backlog and fleet synergy; Builders FirstSource—a building-supplies firm down from about $200 to the $70s, seen as cheap despite a large US housing shortage.
The author judges that the true bottleneck of the AI boom is not code or capital, but the delivery capacity of the physical world — shortages in physical inputs such as electricity, copper, steel, and gas turbines. The opportunity lies in discounted physical assets rather than overvalued AI beneficiaries; stance [Optimistic] (optimistic on physical asset demand, neutral on the forecasts themselves).
The author argues that the true uniqueness of the AI boom lies not in software but in its "appetite" for physical resources — demand for physical inputs such as electricity, copper, steel, gas turbines, cooling water, transformers, and transmission lines is surging, while years of underinvestment have left supply in shortage. The physical world's delivery capacity has become the new bottleneck on technology's ambitions.
In its letter, Robotti & Company Advisors notes that today's AI fervor stands alongside the historical railroad, radio, conglomerate, internet, and housing booms — all built on real and durable changes, yet all carrying the same lesson: "the significance of a technology and the returns to its most celebrated stocks are two very different things" — in other words, the significance of a technology and the returns on its most celebrated stocks are two entirely different matters. Within just a few years, AI has gone from a research curiosity to a fundamental driver of markets: capital expenditures are counted in the hundreds of billions of dollars, valuations embed assumptions of decades of flawless execution, and every announcement of a new model, chip, or data center can move hundreds of billions of dollars in market capitalization.
Unlike previous digital revolutions that ran on "bits," AI is desperately hungry for "atoms": demand for electricity strains a grid built for the last century, while copper and steel, gas turbines, cooling water, transformers, and transmission lines that take a decade of permitting and construction to complete are all in short supply. The author therefore offers the letter's most central judgment: "For the first time in a generation, the binding constraint on technology's ambition is not code or capital but the physical world's capacity to deliver." — that is, for the first time in a generation, the constraint on technology's ambitions is not code or capital but the physical world's delivery capacity.
In the author's view, the market's reaction is misplaced: shares of AI's acknowledged beneficiaries have been bid up to levels with almost no room for error, while the physical economy that must be built to support any version of the AI future is broadly ignored — pricing seems to assume that the indifference of the past decade will persist. The author argues that without predicting which model will win, and without judging whether today's spending is excessive, one can already observe demand rising for machinery, materials, energy, and infrastructure while supply cannot respond quickly; the asymmetry of "enthusiasm concentrated in one place while essential demand accumulates in another" is precisely the gap between narrative and reality he describes.
The author believes the decade-plus flow of capital from the physical to the digital has given rise to a contrarian opportunity: the public market has "abandoned" the physical-asset companies that own ships, factories, mines, machinery, and roads, while private infrastructure funds are buying the same kinds of assets at record capital levels and high prices — the most advantageous way to participate is to buy shares of these listed companies at a discount to replacement cost, rather than locking up ten-year private-fund stakes.
On capital allocation, the author points out that investment capital has poured toward the digital rather than the physical, the intangible rather than the tangible, and growth promises rather than supply discipline — the result being that enormous sums are concentrated in a few glamorous assets, which he summarizes as "asset allocation has quietly become asset concentration." On the neglected side are the physical-asset companies that operate what society depends on.
The author believes these overlooked hard assets are producing "one of the most compelling opportunities investors have had in some time." Private physical-asset and infrastructure funds have long since noticed: they continue to raise record capital, acquiring energy, industrial, and real estate assets at negotiated premiums, with valuations propped up by periodic appraisals and capital locked up for as long as a decade. But the author chooses the public market over the private: rather than buying hard assets privately, buy the "discounted fractional interests" of the listed companies that operate those assets — Mr. Market provides daily, liquid pricing for listed companies holding the same assets, and that pricing is often at a discount to replacement cost, "a price no private negotiation would produce." His rationale is that the structure is superior: permanent capital, conservative balance sheets rebuilt in the last downturn, and, in many holdings, owner-operators whose personal wealth stands alongside fund interests.
It is worth flagging that the author reinforces his positioning with phrases such as "a philosophy consistent for forty years" and "capital is increasingly recognizing and wanting to own these assets" — this is typical self-justifying narrative from a position holder; readers should treat it as an investment thesis rather than an objective judgment.
The author explicitly states that he does not forecast the macroeconomy, but instead assembles a picture bottom-up using "grassroots macroeconomics"; his research conclusion is that the market's two loudest narratives — AI and the torrent of capital chasing AI — both ultimately lead to rising demand for the physical economy, and therefore the opportunity lies in cheap, forgotten physical assets.
The author likens the macro environment to "weather": to be used for guidance, not prediction. The portfolio is built bottom-up, asset by asset, anchored on the margin of safety offered by each individual mispricing. "Grassroots macroeconomics" means weaving together the information gleaned from researching individual companies, combined with industry structure and conversations with management teams and industry participants, to form an understanding of the macro. His current macro stance can be summarized as: [optimistic] on demand for physical assets, [neutral] on forecasting itself — the author concedes he has "no edge" in predicting which model or platform will win, nor is he willing to pay an absurd price to find out.
Applied to AI, the author's approach is to avoid the businesses most vulnerable to disruption and to seek out forgotten, often "boring" assets that are nonetheless essential to deploying the new technology: new power generation, transmission, cooling, copper, steel, and the abundant low-cost natural gas that increasingly underpins North American electricity. The logic set out in the letter: without knowing which AI model wins, one knows that all models consume electricity and materials; the key shovels in this gold rush are physical assets.
The author also concedes this is not without headwinds: the investment demands of model developers, chipmakers, data centers, and utilities stack on top of one another to a total of trillions of dollars; capital is also a commodity, and demand on this scale will push up the prices of physical inputs (inflationary pressure) and the price of capital itself — "both are headwinds to valuations." But on the other side, when investment of this magnitude must flow into physical infrastructure, scarcity value accrues to those who already own productive assets with the capacity to build. He therefore emphasizes discipline within the capital cycle: only by enduring consolidation, exits, recapitalizations, and demand shocks (which, the letter notes, are already reflected in cumulative performance) can one reach the eventual repricing. The author likes to say "good things happen to cheap stocks" — constrained capacity meets rising demand, cash flows respond, and valuations get dragged upward.
The author currently identifies two major directions within physical assets: offshore energy, where the supply-demand landscape has clearly improved after the industry shakeout, coupled with the anticipated Subsea 7–Saipem merger; and the North American industrial revival, driven by the triple forces of cheap natural gas, manufacturing reshoring, and AI's electricity demand.
On offshore energy, the author notes that a decade of underinvestment plus a brutal shakeout — a dozen or so competitors liquidated, absorbed, or exited — has left the industry highly consolidated with reduced supply, while demand is returning and growing. The predictable result is accelerating revenue, higher margins, and increased cash flow. These facts, he says, are "appreciated and rewarded even in today's market"; with regained pricing power, broad order backlogs, and a new supply constraint formed by replacement costs far above the market values of surviving fleets, the earnings outlook is clear.
Subsea 7 (action undisclosed; author's stance optimistic): The author calls the Saipem–Subsea 7 merger a "transformative combination" that will create a differentiated global leader: a broad backlog spanning subsea and renewables, an addressable market expanded by intensifying global energy-security concerns, and stronger pricing power; the combined fleet's optimization capabilities and geographic flexibility would also bring additional opportunities. The author specifically notes that Subsea 7 shareholders are expected to receive a substantial dividend as a result of the transaction. He does not, however, disclose in the letter whether he holds or has adjusted the related positions.
Saipem (action undisclosed; author's stance optimistic): As the other party to the merger, the author spends little ink on Saipem's standalone value; his constructive view rests more on the combined entity's global leadership, order backlog, and fleet synergies.
On the project pipeline, large offshore projects continue to grow in Brazil, the Gulf of Mexico, West Africa, and newer basins such as Guyana, Suriname, and Namibia. The author stresses that these projects are driven not by speculative enthusiasm but by the economic necessity of replacing declining production and by many countries' interest in securing energy independence.
The North American industrial revival is the second macro trend the author identifies: abundant stranded low-cost natural gas gives North America's energy-intensive industries a durable structural cost advantage over the rest of the developed world; this advantage is amplified by manufacturing reshoring (which the author describes as part of "Globalization 2.0") and the tech industry's "newly voracious appetite" for electricity. Unremarkable businesses "hidden in plain sight" — fertilizers, chemicals, steel — are expected to benefit persistently over the next decade.
The previous article emphasized the misreading of the energy and materials sectors under the "cyclical stock" label; in fact, housing and building products exhibit the same divergence between narrative and reality. The original letter makes a key point: "The U.S. housing shortage is a demographic fact, not a forecast." The weight of this statement is that it brings the demand side down from "macro forecast" to "established fact."
The data supports this judgment. U.S. housing has long been in structurally deficient supply; estimates of the shortfall by the end of the 2020s vary across institutions, but the direction is consistent:
| Estimating Institution | Housing Deficit (millions of units) | Basis |
|---|---|---|
| Freddie Mac (2020) | 3.8 | Includes owner-occupied and rental |
| NAR (2021) | 5.5 | Adds multigenerational household demand |
| Moody's Analytics (2022) | 2.3 | Calculated from normal vacancy rates |
Even taking the most conservative figure of 2.3 million units, at the current pace of housing starts it would take decades to close the gap. And over the past two decades, U.S. residential starts have consistently run below the implied level required by population growth — precisely the structural long-term trend hidden beneath the "cyclical" label.
The original letter specifically names Builders FirstSource, whose stock currently trades in the $70 range — seemingly weak, but in fact an "inevitable cyclical pullback." The company is a representative of post-financial-crisis industry consolidation: from a regional player in the early 2010s, it grew through successive acquisitions into a nationwide distribution giant. Consolidation brought not only market share but operating leverage — its EBITDA margin in the last upcycle nearly doubled from its pre-crisis level.
The problem is that the market tends to assign cyclical valuations when such a company's profits are at record highs, and when the share price pulls back, it forgets that the company has "changed the game." At just over $70, versus the high of roughly $200 in 2021, the stock is down more than 60% — precisely the "ugly duckling" state the original letter describes. What has not changed in the fundamentals: the distribution network has regional monopoly characteristics, SKU-management capability forms a moat, and the long-term demand from the housing shortage has not disappeared amid macro rate fluctuations. Every "inevitable intermediate drawdown" is a re-entry opportunity for the compounding of a consolidator's value.
The original letter uses a brief parallel construction to identify the commonality across several holdings — "supply constrained, demand growing, valuations below what private buyers are willing to pay." This is no coincidence; it is the unifying logic of the physical economy's great cycle:
The commonality of these assets: public-market sentiment still resides in the old framework, while private buyers are already willing to pay a significant premium for "supply certainty." For example, North American energy M&A has been frequent over the past two years, with the per-share values implied by transaction prices generally 20–30% above buyers' public share prices — precisely the confirmation of private value in "one hand cash, one hand stock" transactions.
The original letter closes by quoting Rudiger Dornbusch's famous words as its final note — in effect, a distillation of its operating philosophy. Historical experience shows:
These cases all confirm the same pattern: fundamental improvement typically accumulates in a "quiet period" for seven to eight years, while market perception completes its leap from neglect to overvaluation within a few months. The market is currently in the tail end of this "narrative lag" — the underlying asset values of the physical economy have improved, but market sentiment has not yet fully acknowledged it; this is precisely the window for positioning.
The original letter summarizes a strategy unchanged since 1983: hold physical assets through shares of listed companies, buy at a discount, hold patiently, share interests with the investee, and participate actively when necessary. Compared with private-market investing, this has measurable advantages:
| Dimension | Public-Market Ownership (this strategy) | Direct Private-Market Holding |
|---|---|---|
| Information transparency | Periodic reports + public disclosure; data traceable | Relies on counterparty-provided information; information asymmetry |
| Liquidity | Can exit at any time (though unwilling); no redemption lock-up | Typically locked for 5–10 years |
| Fee structure | Low management fees; no performance tiers | Generally 2/20 or higher |
| Participation cost | Small amounts suffice; flexible portfolio | High threshold; concentrated risk |
| Value discovery | Market prices deviate short-term but revert long-term | Always valued at book or DCF |
The key point is that "buy low" opportunities in the public market do not disappear because information is transparent, because market sentiment is often more extreme than rumor. When a building-products distributor is still labeled a "cyclical stock" and sold even as its demand is structurally rising, that is a typical manifestation of this mismatch.
The closing words of thanks are no mere courtesy. In investing, "letting time share part of the work" requires aligned conviction on both sides. The client is willing to accept short-term drawdowns, while the manager maintains discipline and curiosity. This relationship is itself a scarce asset — it allows the manager to hold heavy positions amid pessimism, and allows the client to earn above-average returns while waiting.
The "practice since 1983" mentioned in the letter is not a static slogan but a dynamic calibration: insist on buying real assets, insist on discounted margins of safety, insist on aligning interests with operators. When economic reality and narrative finally converge, the patient will be rewarded twice: with appreciation of the assets themselves, plus the gain from discount repair.
| Position | Direction | Author's Stance | Key Data |
|---|---|---|---|
| Subsea 7 | Not stated | The merger is a "transformative combination" that will create a differentiated global leader; shareholders can expect substantial dividends | The merged entity covers a broad backlog across subsea and renewable energy, with enhanced pricing power |
| Saipem | Not stated | As the other party to the merger, sees the merged global leader status, order backlog, and fleet synergies favorably | Fleet optimization and geographic flexibility bring additional opportunities |
| Builders FirstSource | Not stated | "Ugly duckling" status; fundamentals are unchanged, and each pullback is a re-entry point for the value compounding of an integrating company | Stock price in the $70 range, down over 60% from the ~$200 high in 2021; EBITDA margin nearly doubled versus pre-financial-crisis levels |
| Metals & materials (copper, etc.) | Hold / Monitor | Supply is constrained and demand is growing, naturally lifting the central price level | Copper mining capex has been weak since 2013, and new mine development takes 7–10 years |
| Canadian oil & gas | Hold / Monitor | No longer a "high-cost marginal project," but public markets still rely on the old mental model | Private buyer transaction prices at a 20–30% premium to buyer public share prices |