HFT and the regulation roadblock
It is always the case that there exists a dynamic and complex relationship between regulation and innovation. Moreover, there is an inherent risk always attached to innovation. So the best practice for a forward-looking regulatory body is to mitigate risk and any potential, unpredictable fallout of new business models or financial products. But the bigger question is that can regulation evolve without stifling innovation? This is a pressing question especially in the case of fintech innovations which are redefining the way traditional financial services are offered. So it is important how start-ups and regulators engage around regulation and how regulators need to keep pace with the rate of innovation in the fintech space. But the business in question right now is an established business undertaken by both incumbents and new players with a prowess in technology.
A recent discussion paper by Indian Capital Markets regulator SEBI on algo and HFT trading issued in the first week of August showcases a concerning picture. At a time when Algo/HFT trading has become an integral part of Indian capital markets where all major brokers have invested heavily on Algo/HFT trading infrastructure and putting considerable resources building algorithmic strategies which are an important source of business, bringing measures mentioned in the discussion paper could be disruptive. Go through the link to read the original discussion paper published. Top level observations and impacts:
- Minimum Resting Time for Orders : This would disincentivise traders to place bids/asks resulting in an increase of bid/ask spread and consequently impacting the market liquidity.
- Frequent Batch Auctions: Paper mentions that it existed in Taiwan Stock Exchange and now has switched to continuous limit order book mechanism for regular trading.
- Random Speed Bumps or delays in order processing / matching: This has major structural issues whereby introducing randomization(probability based) algorithm into exchange microstructure can have severe ripple effects. This will block markets for random time intervals leading to opaqueness when price-time priority based order book building mechanism is fair and transparent.
- Randomization of orders received during a period (say 1-2 seconds): Again disturbs the basic principle of price-time priority based order book building mechanism
- Maximum order message-to-trade ratio requirement : Will not this favor only liquid instruments?
- Separate queues for colo orders and non-colo orders (2 queues) : Are we trying to introduce another arbitrage opportunity?
- Review of Tick-by-Tick data feed: Instead of reviewing the accessibility of TBT data, why not democratize colo TBT data to all market participants for free? Colo TBT data patterns expose suspicious activities and bring transparency to the market. It should be available as it is now but can be made available at a low cost.
To conclude, the objective behind the measures is to explore and address concerns relating to market quality, market integrity and fairness due to increased usage of Algorithmic Trading & Co-location in Indian securities market. But will not bringing more complexity in the exchange structure and systems, as discussed above, make it harder for retail investors to understand market dynamics and defeat the premise for bringing in such measures? Also, it would increase governance issues and the expertise required to detect unfair practices on part of the exchange. The need of the hour is for the exchange to innovate and upgrade their survelliance technologies and information systems audit infrastructure to detect unfair HFT/algo trading practices and take a more holistic approach in this space.