New Ask Claude or ChatGPT about any stock, using Bullpen. Bullpen in Claude or ChatGPT. 7 days free

Wednesday, November 18, 2026

Back to the front page

Research Roundup Vol. LXI

Nvidia faced China workaround enforcement concerns as technology fell 6%; energy rose 4.5%, while the company raised eFY28 revenue growth guidance to 75%.

From X

Undiscovered Compounders on X

$IEHC has tripled this year on its backlog, which went from $13M to $56M+ in under 1.5 years. IEH is a family business in Brooklyn that makes connectors for US missiles and radars. We think a good part of that backlog is the defense giants stocking up early, and that orders will then settle at a plateau. From here, the stock depends on 2 things: how fast IEH’s plant can deliver without letting the primes down, and at what margin. 1. In Q2, IEH booked 2.4x what it shipped + its lead times went from 12 weeks to up to 24. To sell $45M/year, it needs about 205 people vs 180 in mid-2026. Even if orders fell to the bottom of our range tomorrow, the plant would have enough work to run flat out through FY29. The risk is letting the primes down. For more than half of what it sells, IEH is the only supplier, and a prime doesn’t change sources as long as IEH meets its deadlines and quality standards. That’s far from a given. IEH has to hire + train + invest to deliver this whole backlog on time, and the bigger the backlog gets, the bigger the risk. 2. More revenue means nothing without margin, and IEH will need more margin than it has now. On a fixed cost base, volume lifts the margin (in Q1, each extra $1 of sales kept 59c after manufacturing costs, vs 33c on average). But 3 costs will weigh on a backlog sold at fixed prices: - The ramp-up itself. IEH pays its new hires before they’re productive, and Q2’s spread (what’s left of sales once manufacturing is paid for) already fell from 33.1% to 28.3%. - Wages, with a union contract that expires on March 31, 2027. That cost is certain, but it should be manageable. - Above all, gold. It’s the biggest variable: each extra $1k/oz costs about 5% of sales. It also cuts both ways: in the short term, a falling gold price lowers IEH’s costs on the orders it has already priced but not yet made. Our base case still has IEH at $50M of sales by FY29 with a 15% operating margin. So the thesis rests on IEH’s ability to hire and train people quickly while making the investments it needs, and to keep margins high (through cost control and price increases), without upsetting its primes. It’s possible: IEH already did it in 2018-2020, but not on this scale. A ramp-up this big makes for uneven quarters, in both the backlog and the margins. We think the market will misread some of them (as it often does with defense and aerospace small caps), which could give us an entry point. At the current price of $30, our 3 weighted scenarios give $43.7 per share in mid-2031, about 8% a year. That’s too little for us, given the governance risk: - In 2019, IEH found its inventory was overstated by about 1/4 and took almost 4 years to fix its financials. - The family-elected board and the auditor are still the same. - Internal control is still rated ineffective. Before Q2, our plan was to start a position at $20 and build most of it between $20 and $15. The X format forces us to leave out a LOT of nuance, bullish and bearish. The full thesis, 30k+ words, is in the deep dive (link in bio). Then Q2 came in with a record backlog, well above our estimates. It may mean our plateau (the run-rate of orders once the advance orders are absorbed) is too low, in which case we’d gladly redo our model and our entry price. We go through all of it in our Q2 update:

S&P Global

S&P Global’s valuation swung from 7x earnings in 2009 to 50x in 2024 as ratings, indices and proprietary data expanded through acquisitions…

TCEHY · long

Tencent Holdings LTD

Tencent, valued near $500 billion, is a long idea built on its Chinese gaming, social-network and advertising franchises.

AppLovin Deep Dive

AppLovin rebuilt its ad model with AXON 2.0, lifting LTM revenue to about $7 billion and margins to 77% as Unity competition threatens ad…

LXS.DE: a 2028 sale it can force

Lanxess can force Advent to buy at least half its Envalior stake in April 2028, potentially raising about €745 million to reduce net debt.

Build-A-Bear: A Narrower Path

Build-A-Bear's push into prestuffed toys and broader retail distribution tests brand stretchability, but recent results warrant caution.

Publications we follow

Every publication on this site, hand-picked.

130 publications

Newsletters 110

X accounts we follow

Every account on this site, hand-picked.

91 accounts

Investor letters we follow

383 firms

Sectors

Ordered by how often our editor reaches for them.