How the Speed of a Trade Got Down to Nearly the Speed of Light
1Digest
This podcast delves into the world of high-frequency trading (HFT), tracing its evolution from early electronic communication networks like Island ECN to the current nanosecond-level speed races. It examines how HFT has transformed market structure, the technological innovations like co-location and microwave links that drive the speed race, and the distinction between liquidity providers and takers. The discussion also touches upon why traditional banks lagged in HFT adoption, the internal competition within HFT firms, and the physical limitations of speed. Furthermore, the podcast draws parallels between the HFT arms race and the scaling of AI, questioning diminishing returns and competitive pressures. It concludes by referencing studies on market efficiency and the high cost of financial intermediation, highlighting the ongoing impact of technology and competition on financial markets and AI development.
Outlines
Introduction to Masters in Business and Odd Lots Podcasts
Barry Ritholtz introduces the Masters in Business podcast, featuring conversations with influential figures in markets, investing, and business. Joe Wiesenthal and Tracy Alloway discuss their Odd Lots podcast's goal of demystifying market processes.
The Evolution and Impact of High-Frequency Trading (HFT)
The discussion covers the rise of HFT, its shift towards exploiting market structure, and the intense race for speed and proximity to trading servers.
Donald McKenzie: Sociologist of Technology and HFT Research
Sociologist Donald McKenzie is introduced, whose work spans technology's societal impact, including finance and HFT. He explains his qualitative, historical, and people-focused research methods, contrasting them with quantitative economic approaches.
The Sociologist's Toolkit and the Cultural Shift in Trading
McKenzie details his research methodology, involving interviews and access to information. The aesthetic shift in HFT firms from traditional trading floors to tech-like startups is also highlighted.
Genesis of HFT: Island ECN and the Speed Race
The origins of HFT are explored through the story of Island ECN. The extreme speeds involved, from milliseconds to nanoseconds, and the equity trade execution process are explained.
Speed's Impact: From Degree to Kind, and Bank Lag
The podcast examines how extreme speed fundamentally changes trading strategies and skills, shifting from human to machine-centered approaches. Reasons for banks lagging in HFT adoption due to bureaucracy are also discussed.
Competition, Co-location, and Microwave Links in HFT
Internal competition for resources within HFT firms is explored. The evolution of the speed race, including co-location and the dominance of microwave links, is traced.
Market Makers, Liquidity Takers, and HFT Firm Structures
The roles of market makers and liquidity takers in HFT are distinguished. The unique structure of HFT firms, trading their own capital, is contrasted with hedge funds.
Physical Limits, Market Efficiency, and AI Scaling
The physical limits of the speed race are addressed. A study on market efficiency and the cost of financial intermediation is discussed, alongside McKenzie's research on AI scaling laws and diminishing returns, drawing parallels between HFT and AI development.
Keywords
High-Frequency Trading (HFT)
A type of algorithmic trading characterized by high speeds, high turnover rates, and high order-to-trade ratios, using powerful computers and complex algorithms.
Market Structure
The organization and operation of financial markets, including trading venues, order types, and regulations, which HFT strategies often exploit.
Speed Race
The competition among HFT firms to achieve the fastest execution speeds, involving technology optimization, co-location, and network infrastructure.
Co-location
Placing trading servers in the same data center as an exchange's matching engine to minimize latency and reduce data travel distance.
Microwave Links
High-speed data transmission using radio waves, often faster than fiber optics for certain distances, used in the HFT speed race.
Electronic Communications Network (ECN)
An electronic trading system that matches buy and sell orders for securities, with Island ECN being an early facilitator of HFT.
Matching Engine
The core component of an electronic exchange that matches buy and sell orders based on price and time priority to execute trades.
Liquidity Provision
Adding buy or sell orders to an exchange's order book to facilitate trading, primarily done by market-making firms.
Liquidity Taking
Executing trades against existing orders in the order book, removing liquidity, a strategy used by some HFT firms.
Scaling Laws (AI)
Empirical relationships in machine learning where model performance improves predictably with increases in model size, data, and computation.
Q&A
What is the core difference between market-making firms and liquidity-taking firms in high-frequency trading?
Market-making firms continuously place buy and sell orders, providing liquidity. Liquidity-taking firms execute trades against existing orders, removing liquidity.
How has the "speed race" in high-frequency trading evolved over time?
It evolved from milliseconds to microseconds and nanoseconds, involving co-location, faster networks like microwave links, and minimizing latency.
Why did traditional banks struggle to compete with independent firms in high-frequency trading?
Banks had bureaucratic IT and slower decision-making, while independent HFT firms were agile and could implement changes rapidly.
What is the significance of Donald McKenzie's research on AI and scaling laws?
McKenzie studies diminishing returns in AI development, questioning the massive investments when gains may be logarithmic and when qualitative leaps like AGI might occur.
How does the concept of "change in degree becomes a change in kind" apply to high-frequency trading?
As trading speeds drastically decreased, trading shifted from human-centered decision-making to machine-centered execution, requiring different strategies and skills.
What does Thomas Philippon's study on financial market efficiency suggest?
Philippon's research indicates that the cost of financial intermediation has not significantly decreased despite technological advancements, largely due to high compensation and fees in the financial sector.
Show Notes
The average person can enter a stock trade on their computer, hit refresh, and the trade is done. As fast as that seems, there are professional traders moving even faster, executing thousands of trades per second. Over the years, the need for speed got so intense that competing firms would aim to get their own systems closer and closer to the exchange's computers, so as to minimize the length of the wires and get their trades in even faster. How did this happen? And how does this change the nature of trading itself? On this episode, we speak with Donald Mackenzie, a professor of sociology at the University of Edinburgh in Scotland. Professor Mackenzie has been studying the intersection of finance and tech for a long time, and in 2021 wrote the book, Trading at the Speed of Light. We discuss the history of finance technology and look at where the technological arms race is going next.
Subscribe to the Odd Lots Newsletter
Join the conversation: discord.gg/oddlots
See omnystudio.com/listener for privacy information.
Table of contents
United States
