About This Research
TernQED explores how reductions in latency create value in electronic markets.
The Core Question
Latency—the delay between when market information is generated and when it arrives at a participant—acts as friction on coordination. The market’s state arrives late and unevenly across participants, driving safety buffers (spreads, inventory limits, hedging, margin) and races to synchronize.
But how much value does reducing this friction actually create? And who captures it?
Six Mechanisms
This research maps latency’s effects through six fundamental mechanisms:
- Adverse Selection Reduction - Faster quote updates reduce toxic fills
- Inventory Risk Reduction - Quicker reactions lower position risk
- Coordination Cost Reduction - Tighter synchronization requires smaller buffers
- Arbitrage Capture - Speed advantages capture price dislocations
- Information Asymmetry Exploitation - Earlier signals enable informed trading
- Queue Priority Capture - Faster submission wins queue position
Why Equities + Crypto?
We focus on equities and crypto because their contrasting structures separate what is fundamental versus institutional:
- Equities: Regulated, centralized, sub-microsecond competition
- Crypto: Unregulated, fragmented, millisecond competition
Same physics, different constraints. This contrast reveals which effects are universal.
Evidence Standard
Everything published here must be auditable:
- Claims traceable to sources
- Confidence labels on all assertions
- Falsification paths provided
- Corrections when wrong
This site is built with an evidence gate—an automated system that blocks publication of unsupported claims.
Built With AI Agents
TernQED uses AI agents for:
- Research aggregation (Monday automation)
- Writing assistance (interactive helper)
- Claim verification (evidence gate)
- Content finalization (Friday automation)
Human-AI collaboration: AI handles research and verification. Humans handle synthesis and insight.
Open Questions
This is ongoing research. Key open questions:
- How do we quantify latency value across different strategies?
- What is the optimal latency-jitter tradeoff?
- How do welfare effects differ across market structures?
- Can we predict latency value from first principles?
Contact
Questions? Corrections? Suggestions for research angles?
[Add contact information]