Summary of Key Points
On July 31st, the exchange switched the transmission channel for market data from a “local area network” (where institutional servers were directly connected to the exchange’s internal network) to a “wide area network” (which all users share). This change eliminated the “microsecond-level” speed advantage that some institutions had due to their physical proximity and direct internal network connections. The move sparked widespread discussions in the market about fairness, the differentiation of quantitative trading strategies, changes in trading volume, competition in technical costs, and the boundaries of regulatory policies. Essentially, it represents the regulatory authorities’ effort to address the unfair technical advantages of quantitative trading at the infrastructure level, aiming to promote the healthy development of the industry.
Detailed Analysis
1. Fairness: Ending “Microsecond-Level Advantages,” but Not Absolute Equality
In the past, some institutions placed their servers in the exchange’s data centers, allowing them to access market data a few microseconds faster than ordinary investors (1 microsecond = one millionth of a second). In the A-share market, transactions are based on “time priority,” meaning those who see the data first and place orders first can secure better prices. With the switch to the wide area network, this extreme speed advantage due to physical location is no longer available, reducing the gap between individual investors and institutions. For investors who questioned whether quantitative trading relied on faster internet speeds to gain an edge, this change signals an improvement in fairness.
However, fairness does not mean complete equality: Institutions with sufficient funds can still maintain a certain speed advantage by using better network connections (such as those provided by top-tier companies), optimizing routing to reduce data transmission delays, or placing their servers in third-party data centers near the exchange. They simply cannot return to the “absolutely fast” levels achieved through direct internal network connections.
2. The Quantitative Trading Industry
The quantitative trading sector is divided into “ultra-high frequency” and “mid-to-low frequency” strategies:
- Ultra-high frequency strategies (such as instant order placement and cancellation, arbitrage on market quotes) rely entirely on speed. With the increased latency and network instability, these strategies will need to be retested, and institutions will have to invest in upgrading their infrastructure and systems, which could be challenging.
- Mid-to-low frequency strategies (such as index enhancement and multi-factor stock selection) have decision-making cycles that are measured in minutes or days, so the few milliseconds of latency have little impact on their performance. Therefore, these strategies are less affected by the change.
In the future, competition in the quantitative trading industry will no longer focus on who has the closest data centers or faster internet speeds, but rather on which strategy factors are more effective, algorithms are more optimized, and risk management is better. The space for arbitrage based on hardware speed will become smaller, and institutions with genuine research capabilities will be more competitive.
3. Trading Volume
The decrease in trading volume on August 3rd (to 2.01 trillion yuan from 2.3-2.9 trillion yuan) was not necessarily caused by the switch in data transmission channels. Industry insiders believe that it was influenced by various factors, including market sentiment, fluctuations in technology stocks, and reduced leverage in margin trading. The volume rebounded to over 2.2 trillion yuan on August 4th and 5th, indicating that the change in data transmission channels was not the primary cause.
What is more important to monitor is whether trading volumes will remain below the average levels of July in the coming months, as well as any changes in the trading activity and turnover rates of high-frequency trading stocks (such as small-cap shares). After all, this change only affected how market data is received; the order placement channels remained unchanged, so institutions simply lost their advantage of having real-time access to market data, not their ability to trade.
4. Technical Costs
With the closure of the local area network channels, institutions had to seek alternative third-party data centers near the exchange. For example, the monthly rent for cabinets around the SSE Jinqiao Data Center increased from 7,000 yuan to 10,000 yuan, and availability was also limited. This is particularly challenging for smaller institutions, which may not have the funds to invest in premium network connections and system upgrades. The industry’s reshuffle might not be due to the failure of quantitative trading strategies but rather because the higher operating costs become unaffordable.
5. Regulatory Policies
There are two misconceptions about these changes:
- Quantitative trading is not being banned: Only ultra-high frequency strategies that rely on microsecond-level speed have been affected; mid-to-low frequency quantitative strategies (such as index enhancement) can continue to operate normally.
- The quantitative trading order placement channels have not been closed: The change only impacted the way market data is received, not the order placement processes.
Regulatory efforts are being made in a gradual manner: first requiring reports from automated trading systems, then requiring high-frequency trading servers to register their activities. The goal is to reduce the unfair advantages associated with physical network connections and promote the healthy development of the quantitative trading industry, rather than completely banning it.
Conclusion
This adjustment of data centers represents the regulatory authorities’ use of technical means to level the playing field, shifting the focus of quantitative trading from hardware speed to strategic capabilities. For ordinary investors, fairness has improved; for quantitative trading institutions, ultra-high frequency strategies are facing challenges, while mid-to-low frequency strategies need to place more emphasis on research and development. In the short term, there may be some market volatility, but in the long run, the industry is likely to become healthier.