Optimising for latency on a decentralised/distributed exchange is mis-guided and a hangover of legacy finance.

Optimise for liquidity first and foremost, and force traders/arbitrageurs to compete with dynamically priced fees that they themselves set.

If an arb opportunity is presented that is worth $100, then any profit-seeking individual will fight to win it.

In legacy finance this means the fastest connections with trade-centres - the fastest wins. They will only pay their 10-30 BP to the market. Fibre optic companies win

However, instead of charging a fixed-fee, charge a dynamic fee, and enforce a throttle - akin to block space.

Now traders have to outbid each other, the one willing to pay $10 on an arb opportunity will win over one paying $9.

The more demand -> higher the fee.

LPs win.

Conversely, traders who don't want to pay $9, can bid $1 and simply wait patiently. They are price-sensitive.

However, most traders are time-sensitive.

Dynamically priced fees + a throttle add the time domain to liquidity.

If there was a trade opportunity worth $1000, then the same thing will happen. Fees will be bid up to asymptote to $1000 (theoretically)

LPs win.

Therefore designers of exchanges must instead optimise for liquidity. Optimising for latency and low fees is a fool's errand. $RUNE

Queue @SBF_Alameda feedback