What Sellers Actually Give Up
One in four cars that sold this spring cut its asking price first — and the haircut has a measurable shape.
The distance between first ask and final listing price
When a car sells, it rarely sells at the number the seller typed in on day one. That gap — between the opening ask and the price the listing exits the market at — is one of the more honest signals in any used-car market. In Georgia's import market this spring, the signal is pointing in one direction.
Among 8,825 sold or archived listings with full price history, tracked since April 1, 27.6% had their asking price cut at least once before they left the market. That means more than one in four sellers decided, at some point, that the market would not meet their original number. The median markdown among those who cut was 5.7%, which translates to roughly $700 off the original ask at the current active-listing median of $12,000.
That $700 figure is also a floor. AutoBridge's listing-history data captures advertised price revisions — the edits sellers make in the system. It does not capture what happens in the car park or across a dealer's desk. The final cash price, agreed verbally or settled over a handshake, is almost always lower than the last listed price. The 5.7% visible markdown is therefore the measurable portion of a discount that runs deeper in practice.
Price revision direction — active listings (cuts vs raises)
What the active inventory is already showing
The sold cohort describes what already happened. The active inventory describes what is building. Of 17,416 active listings with multiple price versions on record, 5.5% have already registered at least one price cut — versus only 1.4% that have raised their ask. That is a cut-to-raise ratio of roughly 4:1.
A 4:1 ratio is the arithmetic signature of a buyer's market. Sellers are not testing the ceiling; they are searching for the floor. The direction is one-way. When nearly four times as many active listings are moving their prices down as up, the aggregate price level in that inventory is drifting lower even before any individual transaction closes.
The active median ask sits at $12,000. Given the directional skew already visible in listed-price behavior, that figure should be read as a ceiling on the distribution rather than a midpoint.
What the active inventory is already showing
The sold cohort describes what already happened. The active inventory describes what is building. Of 17,416 active listings with multiple price versions on record, 5.5% have already registered at least one price cut — versus only 1.4% that have raised their ask. That is a cut-to-raise ratio of roughly 4:1.
A 4:1 ratio is the arithmetic signature of a buyer's market. Sellers are not testing the ceiling; they are searching for the floor. The direction is one-way. When nearly four times as many active listings are moving their prices down as up, the aggregate price level in that inventory is drifting lower even before any individual transaction closes.
The active median ask sits at $12,000. Given the directional skew already visible in listed-price behavior, that figure should be read as a ceiling on the distribution rather than a midpoint.
Price revision direction — active listings (cuts vs raises)
The frozen inventory backdrop
These markdown patterns do not exist in isolation. Prior AutoBridge analysis identified approximately $159 million of ask-capital parked in listings that have been on the market for more than 120 days. That graveyard cohort is, by definition, inventory where no price discovery has yet occurred — listings that have neither sold nor been meaningfully revised downward.
Graveyard inventory suppresses the visible markdown rate. If a car sits at its original price for five months without selling and then gets archived, it may not register as a cut at all — it simply disappears from the market having never closed. The 27.6% cutter share is therefore drawn from the listings that did move, and the true share of sellers who started too high is almost certainly larger once frozen inventory is included.
The excise law reset and where prices have moved
The April 2 excise law — which introduced a base rate of 4.5 GEL per cubic centimeter for vehicles older than six years, with a 60% reduction for left-hand-drive hybrids and exemption for EVs — reset the cost structure for a significant portion of the import pipeline. Caucasus Auto Import has publicly noted that mid-range used prices are down 10 to 20% since April across affected segments.
A 10 to 20% segment decline sits well above the 5.7% median visible markdown in our listing data. The reconciliation is straightforward: many sellers have not yet moved their listed prices to reflect the new clearing level. The gap between where prices are advertised and where the market is actually clearing is, in part, the $159 million sitting in graveyard listings, and in part the unobserved final-price negotiation that leaves no trace in listing-history records.
The cost of starting too high
The markdown curve has a practical interpretation for sellers. A car that exits the market with a 5.7% price cut has already absorbed that loss in nominal terms, plus the carrying cost of however many days it took to reach the revision decision. Sellers who price to the post-excise market reality at listing time avoid both the markdown and the delay. Sellers who start at a pre-April reference price are, in effect, paying for price discovery with their own time and a measurable discount at the end of it.
The data does not show how long it takes to reach the first cut, but the logic is straightforward: early price accuracy is cheaper than late correction. The 72.4% of sellers who sold without a visible cut either priced correctly at the outset or negotiated a sufficient final discount without ever revising the listed number. Either way, the market cleared without a public markdown.
Methodology
AutoBridge listing-history database (versioned price change records), sold and archived cohort listed since 2026-04-01, snapshot date 2026-06-22; n=8,825 listings with at least two price versions · Active-listing dataset: 17,416 listings with multiple price versions as of 2026-06-22; cut/raise counts from direction of most recent price change versus first recorded price · Graveyard figure ($159M in 120+ day listings) and segment price decline context (10-20% since April) sourced from prior AutoBridge analysis and public comments by Caucasus Auto Import · Limitation: listing-history data captures advertised price revisions only; final transaction prices are unobserved and typically lower than the last listed price, meaning realized buyer discounts exceed the 5.7% median markdown shown here
Over 128,000 unique listings published since January 1, 2026.