Which cars hold their value best (and how we measure it)

Two cars with the same sticker price can diverge by thousands of euros after three years on the road. One quietly bleeds value while the other barely flinches, yet at the moment of purchase they look almost identical on the forecourt. To make that gap measurable rather than anecdotal, VIN Analyser computes a Value Retention Index for every model we track, distilling years of listing data into a single comparable number you can act on.

What the index actually measures

The index expresses the share of original transaction value a vehicle retains at a fixed point in its life, normalised against its segment. A score of 100 means the model holds value exactly as well as the median car in its class; anything above means it outperforms the field, and anything below means it lags the typical example of a similar car.

We anchor the measurement at a standard three-year, 45,000-kilometre reference point because that is where most first-cycle ownership ends and where the depreciation curve has already revealed its character. Anchoring everything to the same point is what makes two scores genuinely comparable rather than apples against oranges.

• Score 120+: top decile retention, scarce on the used market and priced accordingly.
• Score 95-110: solid, predictable resale with low downside risk.
• Score 80-95: average decay, the bulk of mainstream models sits here.
• Score below 80: accelerated value loss, often a signal of oversupply or reliability concerns.

How the score is computed

We regress observed resale prices against age, mileage, trim and region, then strip out the noise so the index reflects intrinsic retention rather than local market quirks. The result is a number you can compare across brands without mentally adjusting for segment, currency swings, or a single unusual listing skewing the picture.

Crucially, the index is rebuilt continuously rather than fixed once and forgotten. Every fresh batch of listing data nudges the curve, so a model whose retention is quietly weakening shows up in the score before it becomes common knowledge on the forecourt.

Why we normalise by segment

A city car and a large SUV depreciate on completely different curves, so raw percentages mislead. By benchmarking each vehicle against its own class, the index isolates the part of retention the model controls and filters out the part the whole segment shares. A diesel estate that loses 45% in three years might be a star in a weak segment or a laggard in a strong one, and only the normalised score tells you which.

What can move a model's score

Retention is not static. A facelift, a recall, a shift in fuel policy, or a sudden glut of ex-fleet cars hitting the market can all bend the curve. The index absorbs these as they happen rather than treating last year's reputation as this year's fact.

A retention score is not a reward for a good badge; it is a measurement of how the market actually treats a car once the showroom shine is gone.

Reading the index before you buy

A high retention score cuts both ways. It protects you on resale but means you pay more up front, because the market has already priced the strength in. The sweet spot for value buyers is often a mid-range score paired with a depreciation curve that has already absorbed its steepest drop, letting someone else pay for the new-car premium while you inherit the flatter part of the curve.

It also pays to read the index alongside the confidence band that travels with it. A score of 104 built on thousands of comparable listings is a far firmer signal than the same 104 inferred from a thin, niche dataset, and we never disguise one as the other.

On every report we plot the index against the model's price history so you can see the trade-off directly rather than guessing. The number is a starting point for negotiation, not a verdict, and reading it well can be the difference between a car that quietly funds its own replacement and one that costs you twice.