First, dismantle the category system. Have a single category for all bidders, including motorcyclists and commercial vehicle fleet owners.
Then, levy the actual premium to be paid according to a pre-determined value of the vehicle. Start with the cheapest vehicle, say, a basic motorcycle, and assign it a par value of 0.1. That means the buyer of such a bike pays 10% of the quota premium. A bigger bike might be 0.3 (30%), a light commercial vehicle 1.0 (100%), a heavy commercial vehicle and mass-market car 1.8 (180%), a premium car 2.5 (250%), a luxury car 3.5 (350%) and a supercar 7.0 (700%).
The actual multiples can be worked out by simulating bids based on a single-category scenario, using historical strike prices of all the categories over a 10-year cycle. The average strike price for any given supply year is thus arrived at. Say, it is $25,000 next year. A Yamaha X1 Cub buyer will thus pay $2500 for his COE, a Toyota Corolla buyer will pay $45,000, a Lamborghini Huracan buyer $175,000.
The multiples need not be the ones suggested above. They can be split infinitesimally. Traditionally, we might use a vehicle’s open market value (OMV) to determine the multiple. But if OMV is deemed too variable, a separate table of vehicle values can be devised. This can be based on historical OMVs or posted prices of vehicles on international markets. Averages can be arrived at, which will then determine the multiple each vehicle will attract for a COE premium. If done well, this system could even render the tiered-ARF system and motor loans curb redundant.
Lastly, make all COEs non-transferable.
This overhaul will result in more equity, more transparency and a COE system that is more market-driven. It will also prevent speculation, collusion, and players gaming the system.