The State Paid Up to 13.3% Less for Fuel in July: Why Retail Prices Are Not the Same Benchmark
Georgia’s consolidated public procurement prices for major fuel grades fell 12–13% in July. Regular gasoline was priced at GEL 2.85 per litre while reported retail prices were above GEL 3.50. The gap is real, but it is not a direct measure of retailer profit.
The public contract repriced sharply in July
Under Georgia’s consolidated fuel procurement framework, July prices fell by roughly 12–13% across the listed grades. Euro regular gasoline moved to GEL 2.85 per litre, premium to GEL 2.87, super to GEL 3.17 and Euro diesel to GEL 3.20.
The formula uses the prior month’s Platts benchmark, the GEL exchange rate, excise and VAT, with an adjustment when the monthly change exceeds 3%. The June decline in the benchmark therefore fed into the July contract price.
The retail gap is visible
At the time of the report, regular gasoline at consumer stations was approximately GEL 3.53–3.65 per litre. That placed the public regular price roughly 68–80 tetri lower.
It is tempting to label the difference as retail margin. That would be analytically wrong because the two prices cover different contracts, volumes, payment conditions, delivery obligations, loyalty discounts, timing and service networks.
Bulk procurement buys different economics
A consolidated buyer aggregates predictable demand, accepts contract rules and can create lower selling and credit risk for suppliers. Retail stations must finance land, staff, convenience services, small transactions, local inventory and geographic coverage.
Public contracts can also allocate products by supplier and use a lagged formula. A spot retail comparison on one day may therefore reflect different benchmark periods.
Modelled annual fuel bill at 15,000 km (GEL)
Model, not measurement: 8.5 L/100 km petrol, 5.0 L/100 km hybrid. Retail price GEL 3.59 (midpoint of the reported range); public procurement price GEL 2.85.
What the gap is worth to an ordinary owner
A model, with its assumptions stated: 15,000 km a year, 8.5 L/100 km for a petrol car at the 2.0-litre median engine size in our data, and 5.0 L/100 km for a comparable hybrid. At the reported retail regular price of GEL 3.59 — the midpoint of the GEL 3.53–3.65 range — the petrol car burns GEL 4,577 of fuel a year and the hybrid GEL 2,693, a difference of GEL 1,885. Fuelled at the July public-procurement price of GEL 2.85, the same petrol car would cost GEL 3,634, or GEL 943 less than at the pump.
Set that against what the market asks. The median hybrid listing on AutoBridge is $13,500 against $12,800 for petrol. Those are unadjusted medians of two different groups — the hybrid set is newer, with a median model year of 2020 against 2019 — so the $700 gap is not a like-for-like powertrain premium and cannot be turned into a payback period. What the model does say is that at 15,000 km a year the fuel-cost difference is GEL 1,885. The arithmetic is not in dispute; the assumptions are. At 8,000 km a year the modelled saving falls to about GEL 1,005, or 53% of the figure above.
What the gap is worth to an ordinary owner
A model, with its assumptions stated: 15,000 km a year, 8.5 L/100 km for a petrol car at the 2.0-litre median engine size in our data, and 5.0 L/100 km for a comparable hybrid. At the reported retail regular price of GEL 3.59 — the midpoint of the GEL 3.53–3.65 range — the petrol car burns GEL 4,577 of fuel a year and the hybrid GEL 2,693, a difference of GEL 1,885. Fuelled at the July public-procurement price of GEL 2.85, the same petrol car would cost GEL 3,634, or GEL 943 less than at the pump.
Set that against what the market asks. The median hybrid listing on AutoBridge is $13,500 against $12,800 for petrol. Those are unadjusted medians of two different groups — the hybrid set is newer, with a median model year of 2020 against 2019 — so the $700 gap is not a like-for-like powertrain premium and cannot be turned into a payback period. What the model does say is that at 15,000 km a year the fuel-cost difference is GEL 1,885. The arithmetic is not in dispute; the assumptions are. At 8,000 km a year the modelled saving falls to about GEL 1,005, or 53% of the figure above.
Modelled annual fuel bill at 15,000 km (GEL)
Model, not measurement: 8.5 L/100 km petrol, 5.0 L/100 km hybrid. Retail price GEL 3.59 (midpoint of the reported range); public procurement price GEL 2.85.
The useful question is pass-through speed
Instead of treating the gap as profit, measure how quickly changes in import cost and benchmarks appear in public procurement and retail prices. Track the size and duration of the spread by grade.
A persistent spread that widens while costs fall may justify competition analysis. A temporary spread caused by formula timing or inventory replacement needs a different explanation.
A transparent monthly dashboard
Publish public contract price, lowest and median posted retail price, loyalty-card price where verifiable, Platts benchmark, GEL exchange rate and taxes. Put all dates on the chart.
This would let drivers see whether the state and retail market are moving in the same direction without pretending that they buy the same product under the same conditions.
The public–retail price gap is not a retailer-margin estimate. A margin calculation would require wholesale acquisition, operating cost, discounts and inventory-timing data. Fuel-cost figures in this report are a model with stated assumptions (15,000 km a year; 8.5 L/100 km petrol; 5.0 L/100 km hybrid), not a measurement of any individual vehicle. AutoBridge figures are from the canonical listing set, 1 March – 27 July 2026.
Methodology
Public records and official statistical releases listed below, combined with the AutoBridge listing database under the canonical analytics filter (passenger cars, not hidden, published from 1 March 2026, asking price of $2,000 or more).
Reported state-procurement and retail fuel prices, plus unadjusted powertrain medians from 26,633 AutoBridge listings with a known fuel type. The annual fuel-cost figures are a model, not a measurement.
Public sources through 26 July 2026; AutoBridge snapshot of 27 July 2026 covering listings published from 1 March 2026. July is a partial month.