What Holds Its Value and What Doesn't

Most advice about resale value is organized by product, which is why it never seems to apply to the thing in front of you. The useful version is organized by cause. Four conditions decide whether something holds its price, and once you can see them, you can estimate an unfamiliar item without looking anything up.

Here is what the data actually shows, and then the rule underneath it.

Depreciation is front-loaded, almost everywhere

The headline number for cars is that the average five-year-old vehicle has lost 41.8% of its sticker price, about $16,571. That figure comes from an iSeeCars analysis of roughly 950,000 five-year-old vehicles sold between March 2025 and February 2026.

The average hides the thing that matters. Depreciation is not spread evenly across those five years, it is concentrated at the front. The same pattern shows up in almost every durable-goods category that has been measured. Whatever the item, the steepest part of the curve is the beginning, which is why buying something one or two years old is usually where the value is, and why selling something you bought six months ago is usually where the pain is.

The spread around that average is enormous. A Porsche 718 Cayman lost 9.6% over the same five years. A Nissan Leaf lost 63.1%. Those are both cars, and the outcomes are not close. Product category tells you very little. Which specific thing you bought tells you most of it.

Furniture falls faster than people expect

Used furniture loses value quickly. Even in solid condition, most pieces resell for well under what they cost new, and the discount deepens with age unless the piece is recognized vintage or solid wood from a known maker.

Material and brand move the number. Solid wood, leather, and a name hold; laminate, polyester, and unbranded do not.

What this means in practice is that a dresser bought eighteen months ago is worth noticeably less than its receipt says, before deducting for wear. Sellers often price from what they paid and wonder why nobody responds, and buyers often have no idea whether the asking price is a good number or a bad one.

The brand gap is widest in equipment

Exercise equipment is the clearest illustration, because the same category contains both outcomes.

A used treadmill from a known brand, such as Peloton or Sole, holds a meaningful share of its retail price for the first couple of years.

A no-name big-box treadmill loses most of its value almost immediately.

The difference is not build quality, or not only that. It is that a buyer can look up what a Peloton Tread is, find what others sold for, and check the motor hours. Nobody can do any of that for an unbranded machine, so the price collapses to whatever a stranger will risk on an unknown.

Worth knowing if you are buying one: high motor hours and outdated software on a smart machine both cut the value, and both are checkable before you hand over money.

Some things go up

Lego is the best-documented case. A peer-reviewed study by Victoria Dobrynskaya and Julia Kishilova examined 2,322 sets sold on the secondary market between 1987 and 2015 and found average annual returns of at least 11%, or 8% after inflation, outperforming gold, large-cap stocks, and bonds over the period.

That is a real finding from a real journal, and it is also widely misread. It describes retired sets held for years, not any Lego box bought today. The mechanism is the point: production stops, demand continues, supply cannot respond.

What the four conditions actually are

Everything above reduces to the same four questions.

Can a buyer verify what it is? A model number, a serial, a brand with a catalog. Verifiable things trade near a knowable price. Unidentifiable things trade at a discount for risk, which is most of why generic equipment collapses.

Is there a liquid market with recent sales? Value is not an intrinsic property, it is the record of what comparable items recently sold for. Thin markets produce wide, arbitrary spreads in both directions.

How fast does the new version improve? Anything whose replacement is meaningfully better every year depreciates on that schedule, regardless of condition. This is why electronics fall and cast iron does not.

Can supply respond to demand? Discontinued, retired, out of production, or slow to manufacture means supply is fixed. This is the Lego mechanism, and the same thing that props up certain cars and instruments.

Score any unfamiliar item on those four and you will be close. Four yes answers and it holds. Four no answers and it is worth a fraction of what the seller is asking.