Algo trading market structure through the turbulent times of COVID-19
The COVID-19 pandemic has allowed experts to explore the market’s functioning in different time periods and even make some predictions. Moreover, it also allowed the traders to understand the market structure in terms of market fragmentation, algorithmic trading, and hidden liquidity. To study the different stages of the market in the COVID-19 pandemic, we will be going through the NYSE floor closing and reopening against the preceding uncertain period.
We will try to cover what was happening during the period of market uncertainty that took place in early 2020 in a series of blog posts. Here is the first one.
The resulting volatility surrounding the market uncertainty period led many financial commenters to suggest that the presence of algorithmic trading programs worsened the volatility.
Furthermore, some commented that algorithmic trading accelerated the fastest bear market in U.S. history. These comments are not vain, as researches suggest that high-frequency traders (HFTs), a subset of algo trading (AT), tend to withdraw from the market when conditions turn unfavorable. Market researchers also state that withdrawing HFTs from the market is undoubtedly a cause of market instability that increases the volatility and comovement in liquidity.
Similarly, proxies for algo trading worsen price volatility in the market’s uncertain period. Theoretical and empirical evidence provide mixed evidence (Justin Cox and Donovan Woods, Journal of Banking & Finance, 19 Nov, 2021) that algo trading contributes to market illiquidity, price instability, and inefficiency. Many studies believe that algo trading improves market quality, price discovery, and efficiency. However, some of these studies find that these improvements in liquidity come with adverse selection costs to slow traders.
Researchers also indicate that automated algo trading may erode prices’ informational content since they may deter informed investors from producing valuable research and thus mitigate their informational rents. Besides, HFTs are seen to increase market-making costs for institutional investors.
As for the role of algo traders during market uncertainty, it is argued that since they have no affirmative liquidity obligations, traders may withdraw from the market and suspend liquidity provision during unfavorable market conditions. Hence it can be concluded that the lack of obligated liquidity provision among algo traders is consistent with the widespread belief that fast traders suspend liquidity provision when it is most needed, such as during market stress.
Researchers have documented that market-making HFTs, rather than opportunistic HFTs, are the main drivers of liquidity comovement in the market. However, such comovement may also increase market volatility to the extent that fast traders scale back in unison during market uncertainty. It is also seen that algorithmic trading tends to be more passive, implying that low-latency trades supply liquidity. It is also believed that HFTs consume (supply) liquidity when it is plentiful (scarce). Therefore, despite the concern that algorithmic trading contributed to the widespread panic in financial markets only to profit during the COVID-19-induced market uncertainty, algo trading may have actively supplied liquidity and mitigated declines in market quality.
Simultaneously, the increase in market uncertainty should coincide with an increase in hidden liquidity (market fragmentation). Market uncertainty leads to market consolidation as traders seek or prefer venues that offer trading immediacy over venues with low execution costs. Thus, in periods of heightened uncertainty, traders become more urgent and constrained by search and routing frictions and choose to consolidate more of their orders on a few select markets. Therefore, the markets fragmented less during the peak of the COVID-19-induced market uncertainty period. During market turbulence, in terms of hidden liquidity, orders are less aggressively priced, resulting in wider spreads since limit orders traders experience a greater risk of being “picked-off” by informed counterparties during periods of market instability and therefore price orders less aggressively. When spreads widen, traders tend to make use of hidden orders. Thus, although periods of market uncertainty inherently make the costs of consuming liquidity more expensive via a wider bid-ask spread, the economic costs of providing liquidity during market uncertainty are likely mitigated by the presence of hidden limit orders.
Furthermore, traders that tend to hide their orders also place more aggressively priced orders, which should help narrow spreads and reduce illiquidity. Thus, traders may find hidden limit orders appealing to mitigate adverse selection risk during periods of heightened uncertainty. Overall, the market uncertainty period is associated with significant changes in market structure trading dynamics, including hidden liquidity.
This blog post series is aimed to analyze the past and help in making algo trading and HFT market forecasts. We will explain how the NYSE floor closing was affected by the market instability caused by the COVID-19 outbreak in 2020 in our next blog posts.