How to read a vehicle market report

The anatomy of a report

Vehicle market reports come from several kinds of publisher — auction and remarketing companies, valuation guide providers, listing marketplaces, industry associations and lenders. Each measures a different slice of activity, which is why two credible reports can point different directions in the same month without either being wrong.

Whatever the source, a usable report tells you five things: what was measured, over which population, across what period, compared against what, and how it was adjusted. If a headline number arrives without those five, it is a marketing figure rather than a measurement.

What an index level actually means

An index is a relative measure. A level of 100 means "equal to the base period" and nothing more — it is not a dollar amount and it is not a quality score. Two immediate consequences follow.

First, levels from different publishers are not comparable, because their base periods and constituents differ. Comparing one publisher's 104 to another's 98 is meaningless. Second, the change is usually more informative than the level — but only when you are explicit about the comparison period. Month-over-month, quarter-over-quarter and year-over-year answer different questions, and a report that shows only the flattering one is telling you something about itself.

Notice that the same underlying series is simultaneously falling and rising depending on the window. Both readings are correct. Anyone quoting only one of them has chosen a story.

Seasonal adjustment

As covered in seasonality, the used market has a repeating annual shape. Seasonal adjustment attempts to remove it so that the remaining movement reflects genuine change. Adjusted and unadjusted versions of the same series can tell opposite stories in the same month — an unadjusted series falling in November may be perfectly normal, while an adjusted series falling in November is a signal.

Always establish which you are looking at. If the report does not say, assume unadjusted, and compare year-over-year instead so the calendar cancels out.

Mix effects: the big one

Mix is the most under-appreciated concept in vehicle market data and the source of most confident wrong conclusions.

An average transaction price can rise while every individual vehicle falls in value, if the vehicles being sold that month were newer, better equipped or from a more expensive segment. This happens routinely, because what arrives into the used market is not constant — lease maturities, fleet returns and repossession volume all vary in composition, not just in quantity.

Better reports control for this by tracking constant-quality or like-for-like series: the same age band, same segment, adjusted for mileage and condition. A report that quotes raw average sale price with no mix control is measuring what sold, not what things are worth. Both are legitimate measurements; only one answers "did values change".

An average transaction price can rise while every individual vehicle falls in value. Ask whether prices changed or the mix of what sold changed — before anything else.

Sample basis and coverage

Every report measures a population, and that population has edges:

• Wholesale auction data reflects trade-to-trade transactions. It moves earliest and is the best leading indicator, but it is not retail and it over-represents fleet, lease and repossession supply.
• Retail listing data reflects asking prices, which are not transaction prices. Listings show intent and inventory; they overstate value in a softening market because sellers reduce prices slowly.
• Retail transaction data is the most relevant to a consumer-facing decision and the hardest to collect broadly.
• Valuation guide values are themselves modelled outputs, derived from some blend of the above with editorial judgement applied.

Also check geographic and segment coverage. A national number is an average of regions that may be moving in opposite directions — see regional demand. And check sample size at the level you care about: a national index may rest on a large sample while the specific model, trim and region you are actually asking about rests on a handful of sales.

Lag, revision and staleness

Data takes time to collect, clean and publish. A monthly report describes a period that has already ended, sometimes weeks earlier. In a stable market this hardly matters. In a fast-moving one it matters enormously, and it is precisely when the market is moving fast that people reach for the report.

Two related habits. First, note whether figures are subject to revision, and whether previous periods were revised in the current release — a quietly revised prior month changes what the current move means. Second, keep one high-frequency, local observation of your own alongside the published series: the number of comparable vehicles listed within your radius, checked weekly. It is cruder than any report and it is current, which sometimes matters more.

Supply metrics beside price

Price is the headline; supply and velocity are the explanation. The metrics worth tracking alongside a value index:

• Days to turn — how long inventory takes to sell. Rising days usually precede falling prices.
• Weeks of supply — inventory relative to the current selling rate. The cleanest single read on whether a segment is tight or loose.
• Conversion or sale rate at wholesale — the share of vehicles offered that actually sell. A falling conversion rate with flat prices means sellers are holding out, and prices typically follow.
• New vehicle inventory and incentives — new supply competes directly with late-model used, and heavy new incentives pull the top of the used market down.

Read together, these usually tell you what is about to happen to the index before the index reports it.

A reading checklist

1. What population was measured — wholesale, listings, or retail transactions?
2. What period does it cover, and when was it published?
3. Is it seasonally adjusted?
4. Is it mix-adjusted or constant-quality?
5. What is the comparison — month, quarter or year?
6. What is the sample size at the level I actually care about?
7. Does the geography match my market?
8. What are supply and velocity doing?
9. Were prior periods revised in this release?
10. Does the publisher have a commercial interest in the direction of the number?

Ten questions, most answerable in two minutes from the methodology note. That habit is the entire difference between using market data and being used by it.