The fixed boundary of a £2,500 threshold for a 'longshot bet' has a high probability of missing much more than it catches, perhaps by a factor of 5 to 1. Anyone trying to exploit prediction markets, or hide insider trading, isn't going to stick to one large transaction when they can split it into multiple smaller ones, making this definition a poor detection mechanism. It’s like setting a tripwire that only triggers for 10% of intruders, letting 90% simply step around it by making two £1,500 trades instead of one £3,000 one.