After I bought my first antiquarian books I had a number for what the collection was worth, and the number was fiction.
It came from a spreadsheet I'd filled in from listings — things I'd seen for sale, half-remembered, some of them years old. No dates, no sources. Every figure in it was an asking price, which is to say an opinion with a number attached. I'd never once checked whether anything actually sold at those prices.
Three rules fixed it. They're specific to books, but the second one applies to anything without a ticker.
Replacement and resale are different questions
For books I record what a dealer would charge me to replace the copy today. That's the insurable number, and for a collector it's usually the one that matters — the scenario you're protecting against is the shelf burning, not a liquidation.
What I'd actually receive if I sold goes in the notes, separately. It's normally far lower, because a sale means fees, shipping, insurance, photography, and the buyer's premium coming out of the same object.
Keeping both is the whole trick. Keep only the replacement figure and you'll overestimate what the collection could ever turn into. Keep only the resale figure and you'll underinsure a shelf you couldn't actually rebuild. They're answers to two different questions and neither is "what it's worth".
Cite or abstain
Every valuation carries a source: the comparable copy, the auction result, the dealer quote — with a date and, where there is one, a link.
If I can't find evidence I'd be willing to show someone, I don't write a number. The entry says insufficient evidence and the book holds at what I paid.
That's the rule I break most often in my head and never on paper. There is always a plausible figure available if you want one badly enough, and the whole value of the record is that it doesn't contain figures I wanted.
Worth being precise about how this fits with the general habit: cost basis is the default, the thing an item falls back to when evidence runs out. Retail replacement is a claim, and like any claim it has to be sourced and dated to earn its place. A replacement figure with a dealer quote behind it beats cost basis. A replacement figure I invented does not.
Asking is not realized
Sold prices and auction hammers are evidence. Listings are not.
The gap between the two is wider than most collectors expect, and it's wide in a consistent direction — asking prices sit well above what things actually change hands for, sometimes by a multiple. It varies by category and by how patient the seller is, which is exactly why the individual dated comparable matters more than any rule of thumb about the spread.
I learned this properly on watches rather than books. I re-valued a collection against realized sales instead of listings, and the total dropped hard enough that I sat with it for a day. Nothing had changed about the objects. The only thing that changed was that the number became true.
What it's actually for
None of this makes a collection worth more. Most of it makes the number smaller, and the first time you do it honestly it stings.
What it buys you is a record that survives you not being there to explain it. Someone opening it later gets the reasoning — what this is, what it cost, what it's roughly worth, how confident that figure is, and where the confidence came from. That's a different artifact from a valuation, and a more useful one.
The rules work fine in a spreadsheet. I ended up building Khalt because mine kept going stale and because the books and the brokerage account never rolled into one number — dated valuations that carry their source, provenance you layer rather than overwrite, photos and invoices attached to the item. Estimated stays labeled estimated.
Nothing in it is public: no feed, nothing to share, no trackers, and it never asks for a serial number or an address. There's a one-click export and a delete that actually deletes — khalt.app/help.
It's free and early, and I'm still looking for a few collectors who'll tell me where it's wrong.