hard · Market Microstructure hft
Exchange A's co-location facility offers a round-trip latency of 15 microseconds. Exchange B offers 80 microseconds. A market maker's quote on Exchange B becomes stale when Exchange A's price moves.
What is the maximum time window during which the market maker on B has not yet received the signal to update before an HFT co-located on A can snipe the stale quote, and what is the market maker's most effective microstructure defense?
- 65-microsecond window; the market maker should simply widen its posted spread on Exchange B permanently to offset this latency disadvantage relative to the faster Exchange A.
- 65-microsecond window; the market maker should also co-locate at Exchange A to receive the price signal in 15 µs and cancel-replace the Exchange B quote before the HFT can snipe it.
- 80-microsecond window; the market maker must receive the price signal in 80 µs before reacting, so achieving true latency parity with the faster Exchange A is structurally impossible.
- 15-microsecond window; because the HFT co-located at Exchange A fires in only 15 µs, the market maker on Exchange B enjoys a full 65-microsecond cushion before any sniping order can arrive.
Sign up free to see the explanation and track your rank →
More Market Microstructure hft practice
- A trader places a large sell order for 50,000 shares at $50.01 only to cancel it immediate
- A high-frequency trading firm detects a price change on the NYSE and executes a trade on B
- An HFT firm's co-located server receives a direct data feed… — What is the primary risk th
- A high-frequency trader places a buy order for 10,000 shares… — Which prohibited practice
- A high-frequency trader (HFT) notices that the price of a stock has just risen on the New
- How does the algorithm's reaction contribute to a potential flash crash?
- A trader wanting to buy 10,000 shares cheaply places a 50,000-share sell order at $50.01
- As the price of the underlying stock begins to fall rapidly, what is the delta-hedging res