Cost basis and margin for card vendors, including the awkward case of trades

Ask a card vendor what their margin is and you will usually get a percentage that sounds plausible and is not derived from anything. This is not carelessness. It is that cost, in this business, is genuinely awkward: stock arrives in collections rather than units, prices move constantly, and a large share of acquisitions are trades where no money changes hands at all.

The awkwardness is worth pushing through, because margin is the number that tells you whether the way you buy is working, and every shortcut around it produces a number that flatters you.

Cost is what you paid, recorded when you paid it

A card's cost basis is what it cost you to acquire, fixed at the moment of acquisition. It is not what the card is worth now, not what you could rebuy it for, and not what a price guide says. Those are all useful numbers and none of them is cost.

This matters because the alternative — looking up a cost when you need one — quietly destroys your history. Repricing stock today would change what last year's sales appear to have earned, and a business whose past profit moves whenever the market moves cannot be reasoned about at all.

So: capture cost at intake, store it with the card, and treat it as a historical fact rather than a live figure. The practical implication is that acquisition is the moment that matters. Cost information is never easier to record than when the money is leaving your hand, and it is essentially unrecoverable later.

Splitting a collection price across cards

You rarely buy one card. You buy a box for a single number, and then sell its contents one at a time over months. To know what any individual sale earned, that one number has to become many.

The standard approach is to allocate in proportion to value: work out roughly what each card is worth, add those up, and give each card the share of the purchase price that matches its share of the total. A card representing a fifth of the collection's value carries a fifth of what you paid.

It is approximate, and it is enormously better than the alternatives. Two shortcuts to avoid:

• Splitting the price evenly across the card count. This assigns the same cost to a chase card and a common, which makes the good cards look wildly profitable and the bulk look like a disaster.
• Assigning the whole cost to the first cards that sell, so later sales appear to be pure profit. This front-loads all the pain and then tells you a comforting lie for the rest of the collection's life.

Bulk deserves a note. If a collection was bought largely for a few cards, it is defensible to allocate almost nothing to the bulk — but then the bulk shows up as nearly pure profit when it sells, which overstates how good that side of the business is. Giving bulk a small nominal cost keeps both halves of the picture honest.

A trade is a sale and a purchase at the prices you agreed

A customer trades you a card, and the two of you agree what each card on the table is worth. That agreement is the price. Your books should treat the trade as what it is: you sold one card and bought another, and you paid for the second with the first.

The card you hand over counts as sold at its agreed price. Its profit is that price minus what the card cost you, exactly as if a customer had paid cash for it, and it shows on the day of the trade.

The card you take in enters your stock at its own agreed value, because that is what you paid for it — in cards rather than cash. When it sells later, its margin is measured from that figure, so the gain you made on the first card is never counted a second time.

Cash that balances a trade is part of the same deal. If the customer adds cash, it helped pay for your card, so it is already inside that card's sale price. If you add cash, it is part of what the incoming card cost you. Either way it still shows in your cash figures on the day the money moved.

This puts weight on the agreed value. It is only a real price if you would sell the card at it. Agree a card in above what it would fetch, and the trade shows profit today that comes back as a thinner margin, or a loss, when that card sells. A trade valued honestly reports honestly.

Accounting standards for inventory and barter differ by jurisdiction, and if you are filing anything on the basis of these numbers you should confirm the treatment with an accountant where you operate. The reasoning above is about getting a decision-useful picture, which is a related but not identical goal.

"No cost recorded" is not the same as "free"

Some stock genuinely has no recorded cost — it predates your records, or it arrived in a way nobody wrote down. The temptation is to enter zero and move on.

Resist it, because zero is a claim. A card recorded at zero cost reports infinite margin and inflates every total it touches, and once the distinction is flattened it cannot be recovered. Keep "unknown" as its own state. Then a report can tell you that a figure rests on two hundred cards with no cost behind it, which is exactly the caveat you want attached to a margin you are about to make a decision on.

Which margin to look at

Margin per card tells you whether a particular buy was good. It is nearly useless as a business metric, because it says nothing about how long the money was tied up. A card bought at a third of market and sold two years later at market has a magnificent margin and was a poor use of capital.

The pairing worth watching is margin alongside how quickly stock turns over. Together they tell you what a given amount of money earns you in a period, which is the actual question. It is also the number that settles arguments about whether to hold out for a better price — holding costs something, and once you can see the turnover figure, that cost stops being invisible.

Correcting a cost you got wrong

You will eventually find a card whose recorded cost is simply wrong — a typo, a misplaced decimal, a collection allocated before you noticed what was in it. Fixing the card is easy. The question is what happens to sales that already used the wrong figure.

Leaving them alone preserves history but means your reports permanently contain a number you know to be false. Rewriting them makes the reports true but changes periods you may already have reported on or filed against.

A workable rule is to correct within a recent window and leave anything older alone, with a note. That keeps recent reporting accurate, keeps closed periods stable, and — importantly — makes the choice visible rather than implicit. Whichever rule you pick, write it down, because the worst version is correcting inconsistently and then not knowing which of your historical numbers were touched.