Show HN: I reverse-engineered car lease math against three real dealer documents

2 pointsposted 5 hours ago
by amirjavid

5 Comments

amirjavid

5 hours ago

I'm the builder. The motivation: lease math is technically disclosed on every dealer document — MF is on page 3, fee breakdown is itemized — but it's not in a format where you can act on it at signing.

A few things from these three deals that surprised me:

1. The Georgia TAVT has a feedback loop: each $1 of cash down reduces the taxable depreciation base, which reduces the tax, which reduces NCC by ~$1.077, not $1.00. Most calculators don't handle this.

2. The Louisiana dealer was $11 over the statutory doc fee cap. It's small but it's illegal — and it's the kind of thing buried in a wall of line items.

3. The EV9 deal (Oregon) was actually clean — dealer at buy rate, $0 doc fee, $0 add-ons. I wanted to include a "good deal" in the verification set because if the tool only ever finds problems, that's its own bias.

The tool is at quotedefender.com — upload a screenshot of your lease quote and it runs the same analysis. Happy to answer questions about the lease math or any of the three specific deals.

sokka_h2otribe

4 hours ago

I got halfway through. I'm tired of AI writing sometimes here. <<That's what most calculators get wrong>>. Not, that's $11, it's NOT very significant.

AI slop ATM seems to indicate heavy importance on all your sentences, but not all of your sentences are that important. It makes it harder for me to skim personally, and ultimately, it's harder to edit because it "sounds" ok as it is (even when it's not.)

But, in my experience, the AI writing slop [negative experience] is everywhere here.

altairprime

2 hours ago

In the general case beyond this post specifically, drop an email to the mods when a submission is particularly badly written (and note why: if it seems AI-generated, if it sees manipulative emphasis on marketing words, if it’s low-content rabble-rousing like most GitHub issues links, etc); they seem to care about writing quality, and can take more direct action than we can as users.

amirjavid

25 minutes ago

Fair point. I've stripped the 'filler' and updated the post to focus strictly on the raw math like that $1.077 Georgia tax feedback loop. Hopefully the technical data is more useful to the community than my initial draft.

amirjavid

4 hours ago

Good call. It’s easy to let the writing get clinical when the math is this dry, but I hear you on the skim-ability. I’ve updated the post to lead with the actual findings rather than the 'important-sounding' filler.