Newtheorum publishes long-form analysis, contrarian research, and grounded commentary on the systems shaping our economies, technologies, and civic life — written by people who actually do the work.
Wall Street's average 2026 EPS estimate sits at $276. Our bottom-up model — drawing on 1,400 company filings, freight data, and 38 earnings call transcripts — lands closer to $198. The gap isn't a rounding error. It's a structural disagreement about how the next 18 months actually unfold.
Elena Marsh walks through the three assumptions that almost no sell-side desk is willing to challenge, and what happens to portfolios if even one of them breaks.
A live look at the pieces generating real discussion — measured by reading depth, not just clicks.
Why gains from generative AI show up in revenue lines a full 18 months before they surface in productivity statistics, and what that means for the next two interest rate cycles.
State-level yield curves are diverging in ways that have nothing to do with federal forecasts. We mapped 47 issues against voter registration data to find the signal in the noise.
Texas and ERCOT's near-miss wasn't a capacity problem — it was a settlement architecture problem. A working proposal that doesn't require new federal authority.
Every piece we publish fits into at least one of these. The categories evolve as the work does — these are the throughlines right now.
Earnings, rates, the dollar, and the structural assumptions hiding inside consensus forecasts.
AI infrastructure, platform economics, and the policy questions nobody wants to ask out loud.
How institutions actually work — and where the next reform windows are most likely to open.
Grounded analysis of the energy transition, beyond the press releases and the panic cycles.
Zoning, infrastructure, and the political economy of how a city actually gets built.
What the labor market data actually shows — separate from what the press releases claim it shows.
Newtheorum is reader-supported. No advertising, no sponsored content, no data sales. The work is paid for by people who think it's worth paying for.
I've cancelled four other subscriptions this year and kept Newtheorum. The pieces actually change how I think about positions I hold, not just how I feel about the news cycle.
What I value most is the restraint. The team will sit on a story for two weeks rather than publish something thin. That's almost extinct in independent media right now.
I send roughly one Newtheorum piece a week to my team with the note "actually read this one." It's the only publication where that note isn't aspirational.