Stock Holders Face $10T Warning: Factory Paid, Product Pending

Episode 780 on YouTube brings hosts Mindy Jensen and Scott Trench together with valuation expert Aswath Damodaran to unpack where cash flows in the big technology funding wave. The backbone is a simple factory split. NVDA, the power suppliers, the builders of data centers and the landlords of land collect cash today. Sellers of products such as Meta, Alphabet and Amazon hope to earn years from now. For a holder of stocks, that split is everything. Cash entering the factory side is certain; the product side runs on hope. Damodaran tells it without drama. He says funding choices should start from who gets paid and when. That lens frames all the math that follows.

Since ChatGPT arrived in November 2022, between $2T and $2.5T has flowed into this arena. Not even four years have passed. Damodaran calls it the largest build in history. The fortunes poured into railroad lines and the early internet cheer look modest beside this scale. Cash went to giant setups running new versions, to power and to buildings. Stock markets cheered the flow because balance sheets on the factory side swelled fast. As funding appetite grew, expectations grew too. Nobody asked about brakes. Looking back today, the picture is clear: registers filled, bills moved to the future. A funding wave of this scale lifted stock valuations as well and built an expectation that is hard to reverse.

The present sales picture is harsh. Total yearly revenue from all products stays near $250B on generous math. Anthropic and OpenAI lead subscriptions and usage. In his Substack piece, Damodaran reports Anthropic reaching a $65B yearly pace by the end of July 2026, with OpenAI near a $40B tempo. CNBC confirms the same $65B figure and points to roughly sevenfold growth within a year. So a $2T buried factory faces $250B in sales. Unless that ratio heals, a write-down is unavoidable. The harsh math in the same piece adds more: a talked-about $2T listing needs $1.2T in sales across ten years at a 30 percent profit margin. Stock prices follow hope here, not that math.

A durable balance needs yearly sales between $8T and $10T. Tax and profit ladders push the bar that high. The noted finance writer refuses the famous B-word and speaks of foam instead. He poses three questions: is it possible, is it plausible, is it probable. Today he grades all three low. The reverse math on the BiggerPocketsMoney page points the same way: about $30T of value, $29.9T of enterprise value, $450B of free cash and a 66-times multiple. For that structure to work, free cash must compound at 32.6 percent yearly. So math speaks, not hope. For anyone holding stocks, the message is plain: the bar sits high while time runs short. Stock valuations track present cheer, not that bar.

Stock pricing: factory or product

Ad cash keeps this race alive. Alphabet and Meta fund the bet from profit pools in search and social. The bill still lands on holders of stocks; dividends and buybacks shrink. The real fragility grows beneath the balance sheet. Axios flags CoreWeave debt: five-year protection costs above 1100 basis points, about $1.1M a year to insure $10M. Market voices call it a Darwin moment. The new clouds count as the riskiest slice of the trade. So loss sits not only on giant balance sheets but across credit markets. For stock holders the lesson is clear: profit pools look deep, yet debt chains run long.

For anyone funding through an index, the picture is even denser. This cluster holds between 39 and 40 percent of the S&P 500. So a passive-looking portfolio leans on the same bet. History warns of the timing trap: those who left in 2008 fear and never returned missed a giant climb. A horizon of six to nine years rides through such waves. But advice differs for two years to retirement: look to bonds near 4 to 5.5 percent. Trimming stock weight to protect sleep beats extra yield at that age. Funding horizon and risk appetite belong in one sentence. The personal schedule comes before the market schedule.

An ATB study puts numbers on S&P 500 crowding. The top ten stocks hold about 40 percent. The seven giants hold about 34 percent. In an equal-weight setup the same seven fall near 1.4 percent. The gap is huge. Equal weight buys low and sells high on autopilot, gathering laggards and trimming runners. The sleep test runs through here: crowding that ruins nights turns into bad calls by day. Spreading single-cluster risk across a portfolio keeps funding discipline intact. For a chooser of stocks this table guides. Tying weight to a cap and writing it into the schedule removes emotion.

Benzinga reports the valuation expert fully closed his stake in the largest component maker. The reason is plain: the sales needed for a factory of this scale defy sense. Damodaran bought cheap: MSFT in 2013, Meta after the data scandal, NVDA near $1.80 on a split-adjusted basis in 2018. He closed Tesla over politics. Five of the seven giants stay in the portfolio. The rest look fair or slightly rich. So no blanket exit, only a careful stance. For a holder of stocks the message is plain: value bought cheap earns a farewell once rich. That approach leaves no room for feelings in portfolio care; balance sheets and prices talk, excitement stays quiet.

Portfolio lessons: weight, caps and patience

Selling discipline runs on autopilot. Once a single name passes the 15 percent portfolio cap, trimming starts. No sale enters the mind before richness hits the 25 to 30 percent band. With California taxes the total load nears 26 to 27 percent, so an early sale costs dearly. Hence the patient rule: no trigger pulled before the bar breaks. No grief over profit left behind. Looking back and sulking ruins discipline in stock funding. Damodaran never mourns cash left on the table. A funding choice faces forward; yesterday price never excuses today. That cool head forms the quiet engine of long-horizon portfolio success.

A listener question teaches a full lesson. One saver holds about 70 percent in Musk names and about 85 percent in the seven giants. The advice is staged exit: a slice sold every six months, through limit orders, after homework. The watch list holds BYD, Palantir and MercadoLibre. Buys sit far below present prices. No rush, only plan. That tempo keeps a crowded holder of stocks safe. Leaning on one cluster steals sleep; staged thinning restores calm. In funding, patience beats speed. That discipline lights a path for everyone swept into the S&P wave. A limit order goes on the book, feelings stay aside; portfolios run this way. On balance-sheet days, plan talks instead of panic.

The race between the index and a fifty-stock global portfolio teaches too. The index is easy and cheap; a custom portfolio takes work but educates. The first rule is to do no harm: never bury 80 percent in one name. Spread runs across fields and regions. The stock count stays near fifty. The holder knows every name and reads every balance sheet. Beating the S&P index is not required; the goal is calm sleep and steady yield. A portfolio should let its owner sleep at night. Such plainness outlasts tangled setups. Joined with valuation discipline, the outcome shifts. In funding, few names with close watch beat many names with thin calm; balance wins. Reading balance sheets pays more than chasing charts.

Balance-sheet reality: rental or sale

Circular cash flows spark the sharpest debate. The xAI side rents finished centers to the Anthropic side. Reuters reports Anthropic pledging $200B of spending on Google cloud and components. A parallel structure links MSFT and OpenAI. Rental means the factory is leased out, not products sold. The figures from Moody carried by DataCenterDynamics widen the frame: $785B for 2026, about $1T on the 2027 horizon. The AWS side runs above a $15B pace, the MSFT side holds $37B with 123 percent growth. Payback talk spans 12 to 24 months. For a holder of stocks the question stands: can rental income replace product sales. Debt and rental loads swell on balance sheets while the answer lags.

The ceiling math is the final knot. Through a consultant-pay lens, tool-only use caps near $2T. McKinsey-style advisor fees anchor that sum. Dreams of replacing workers wave a $26T pool, yet systems stall when no buyers remain. A credible ceiling sits near $10T. Funding tempo never cools: $700B to $800B of fresh cash yearly toward $3T to $4T. SPGlobal points to $1.3T by 2027. All six giants show negative free operating cash for 2026-27, recovery near 2029. Debt, rentals and special vehicles pile onto balance sheets. TaipeiTimes expects OpenAI near a $70B pace by year end, above the $50B September figure. An Anthropic listing is sidebar talk this fall; an OpenAI one next year. Counting stays disputed. Yahoo writes NVDA near $6T, the first firm at that line. For stock markets one question stands: which arrives first, price or sales.