In 2010, a financial telecommunications company spent $300 million to lay an 800-mile fiber-optic cable straight through the Allegheny Mountains from Chicago to New Jersey. They didn't build it around the mountains—they tunneled straight through rock in total secrecy.

The entire point of this engineering feat? To cut signal transmission time by just 3 milliseconds.

Two years later, that $300 million cable was effectively obsolete, replaced by microwave networks cutting microseconds off the transmission time.

Why this 2010 speed race still matters for crypto traders today

Most traders spend their time analyzing exchange maker/taker fees, funding rates, or slippage. But the largest tax active traders pay on continuous order books is latency arbitrage.

On traditional central limit order books (CLOBs), trades are matched continuously in real-time down to the microsecond. When a major market movement occurs on one reference venue (like Binance or Coinbase), there is a tiny millisecond window where older quotes sit stale on secondary exchanges.

High-frequency traders (HFTs) spend millions on co-located servers, direct exchange feeds, and ultra-low-latency hardware to win the race to hit those stale quotes before the market maker can update them.

How this impacts retail traders directly

You might think, "I'm not fighting HFT bots directly, so why do I care?"

You care because market makers aren't stupid. To protect themselves from getting constantly picked off by speed-arbitrage bots during fast market moves, liquidity providers are forced to do one thing: widen their bid-ask spreads.

Every quote you see on a continuous order book is wider than it naturally needs to be because market makers are actively pricing in the risk of losing speed races. Retail traders end up subsidizing the latency arms race through wider spreads and worse execution on every market order.

The Mechanism Problem

Switching to a different continuous exchange doesn't solve this because the flaw is baked into the matching engine architecture itself: continuous-time matching inherently rewards microsecond speed over price discovery.