What happens between click and fill
When you submit an order, your trading platform sends a TCP packet from your machine to your router. Your router forwards it to your ISP's local exchange. Your ISP hands it to their transit network — shared infrastructure that carries data for many customers simultaneously. That transit network eventually hands your packet to a peering point, where it crosses from one carrier's network to another's. Eventually it reaches the data centre where your broker's matching engine runs — typically Equinix LD4 in London for most retail forex brokers — and the fill is processed. The same path in reverse returns the confirmation to your screen.
Each of those handoffs is a hop. Tracerouting to your broker's server hostname shows every hop and the latency to each one. The total round-trip time is the sum of every hop's delay on both directions of the journey.
Why the path varies
Residential ISPs do not route your traffic on a fixed path. The path from your ISP to your broker's data centre changes based on traffic loads, peering agreements, and what your carrier has decided is the most efficient route that day. A path that goes directly from your ISP to the data centre network on a quiet Tuesday may route through an additional transit carrier during a busy morning when congestion causes your ISP to reroute traffic. You would not notice this as a user checking email. You would notice it as a trader watching fills.
How a proxy changes the path
A proxy node at or near your broker's data centre connects to the broker's matching engine via a short, direct, dedicated path. Instead of your trading data travelling 10-18 hops through variable transit networks, it travels through the residential path to the proxy node — and then just a few hops through the data centre network to the matching engine. The variable transit section is replaced with a consistent direct connection. That is the routing improvement the priority route provides.