The news I teased last week: this project has a new name. Uraninite Edge is now QuantLode. Same person, same dashboard, same newsletter — new home at quantlode.com. Old links redirect.
Here's why, and how the search for a name turned into a small lesson in due diligence.
Why rename at all
Uraninite Edge had two problems.
First, it locked me into uranium. The plan has always been to expand — copper is next, within the first year — and a name built on a uranium mineral can't carry that.
Second, nobody could say it. "Uraninite" got misread, mistyped, and mispronounced by almost everyone I showed it to. A name you have to spell out loud twice is a marketing tax you pay forever.
So I made a list of criteria: works for any critical mineral, short, survives being heard once, works in English and German, sounds like data rather than a blog, and the .com is free.
Everything is taken
The first discovery: every short brandable name you can think of is already registered. Lodex, oredex, orescope, oreiq, minetrics, orebit — I checked over thirty. Almost none of them are actual businesses. They're parked domains, sitting there, waiting for someone to pay squatter prices.
The one that almost was
The second discovery was better. I nearly registered MineralQuant. Web search: clean. Trademark registers in Germany, the EU, and the US: empty. I was ready to buy.
Then a LinkedIn search turned up MineralQuant SpA — a mining-metallurgy consultancy in Santiago, Chile. Twenty-five years old, sells software, and its slogan is literally "Quantitative intelligence for mining." I had seen the "SpA" earlier and assumed Italy. It's the Chilean legal form.
Legally, probably fine — different markets. Practically, a disaster waiting to happen: same industry, nearly identical positioning. Search engines and AI models would blend us together from day one.
The lesson: registers and web searches aren't enough. Check where the professionals actually are.
QuantLode
Quant + lode. A lode is a vein of ore — the thing every miner is actually looking for — with an echo of lodestar. It's mineral-agnostic, five letters longer than "quant," and survives being heard once in both of my target languages.
QuantLode is the umbrella: uranium now, copper next, more later.
For you, nothing changes. Same newsletter, same dashboard, emails now come from the new domain. That's it.
First reality check: the model vs. June
In #3 I promised monthly accountability for the QL Spot Estimate. The first month is in, and it's humbling.
On June 30, the estimate said $78.80. The published month-end value: $85.25. Off by $6.45 — about 7.5%.
That's clearly worse than the ~$2.30 average error from training, but training numbers flatter — the model had seen those months. This was the first real, out-of-sample test. What happened: the estimate is driven almost entirely by the two physical uranium trusts, and in June their shares slid while the survey price climbed. The model faithfully reported what the equity market believed physical uranium was worth — and the market believed something 7.5% below the survey.
That gap is itself information (it's why the dashboard calls the number market-implied), but I'd be lying if I said I was happy with it. Improving the model is now at the top of my list. The full methodology deep-dive — error margins, what it catches, where it fails — is still coming, with this miss as exhibit A.
One data point isn't a track record. That's exactly why I log every month in public.
From here
The rebrand is done, the archive is live, and this issue starts a streak: weekly, Sundays at 09:00 CET.
See you next Sunday.
— Maximilian