On March 31, 2020, we released a study of the impact of the ongoing COVID-19 crisis on US equity market liquidity1. In it, we compared spread costs, market impact costs, intraday market conditions, and market resilience during the crisis and under normal market conditions, and found that volatility, spreads, and market impact increased dramatically in the month of March. This brief update shows how liquidity in the US equity market has begun to recover since then.
Study Period and Data
In this continuation of our earlier study, we present market behavior in three [updated] time periods. First, we use January 2020 to represent a “normal” period of trading under “normal” liquidity conditions; during that time, the implied volatility index of the S&P 500, VIX, averaged just below 14 points. The second period we study is the “COVID-19 crisis” period, March 9th to March 31st, during which the VIX Index more than tripled. Finally, the month of April is included to serve as a measure of the market’s recovery.
As before, we have broken the universe of stocks into liquidity buckets to illustrate the behaviors traders are experiencing. We include the mega-cap constituents of the S&P 100 Index and the small-cap constituents of the Russell 2000 Index in this study, though we consider the full S&P 500 Index in the market impact section for increased robustness. For the stocks in our sample, we study every quote change in the best bid and offer—price or size—and every trade across all US equity exchanges from January 1st to April 30th.
Volatility and Volume
During onset of COVID-19 in the US, volatility rose to a historic high. VIX reached its highest weekly average, 74.92, on March 24, 2020—higher than at the peak of the financial crisis in October 20082—rising from record lows. With increased volatility came increased market volume3, as is very typical when portfolio managers begin reallocating assets to navigate changing conditions. Since then, VIX has declined a bit—and market volume in concert—though both remain higher than in the pre-COVID period.
Figure 1AB
Figure 1AB. The evolution of the forward-looking S&P 500 volatility index (VIX) since the onset of the COVID-19 crisis in the US (A) and average daily notional volume (B) over three corresponding periods.
Market Halts
As reported in our original paper, the US equity market-wide circuit breakers kicked in on four occasions4 in March, and the number of single-stock circuit breakers5 rose to more than 1500 in a single day. Since then, we have seen a decline in the number of halts; the number declined sharply in late March and is slowly returning to normal.
Figure 2
Figure 2. Number of triggers of the Limit Up Limit Down mechanism (single-stock market halts based on price volatility). The number of LULD triggers is typically very low but increased to over 1500 triggers in a single day in March 2020. In April, we see this value slowly returning to its typical low.
Spread and Depth
As COVID-19 arrived and spread in the United States, markets responded with increased volatility, and investors experienced changes as the cost of liquidity increased. The cost of liquidity is related to both the bid-offer spread set by liquidity providers (and in some part by investors placing limit orders) and the market depth6. In this section, we present recovering liquidity conditions7 indicated by spread and depth since the peak of the COVID-19 crisis.
Changes in spread
During the rapid spread of COVID-19 in March, we saw dramatically increased spreads. As is illustrated in Figure 3AB, spreads in S&P 100 stocks climbed to almost 4 times their January values, but they have since recovered quite a bit. In Figure 3B, you can see similar behavior exhibited by Russell 2000 stocks, where spreads also widened in response to increased volatility. Spreads in Russell 2000 stocks have not recovered as much, but we are seeing some improvement.
Figure 3AB
Figure 3AB. Average spread in basis points across stocks in the S&P 100 Index (A) and Russell 2000 Index (B) over three periods in 2020, including the COVID-19 crisis period, shown in navy blue.
Changes in depth
As spreads widen, depth tends to decline, which we did indeed see in March. Since then, depth has recovered somewhat—though less than spreads—in S&P 100 stocks (Figure 4A) but very little in Russell 2000 stocks (Figure 4B).
Figure 4AB
Figure 4AB. Average depth in USD across stocks in the S&P 100 Index (A) and Russell 2000 Index (B) over three periods in 2020, including the COVID-19 crisis period, shown in navy blue.
Changes in Normalized Spread
In our earlier liquidity analysis, we introduced a discussion of how tick size impacts the relationship between spread and depth and introduced the idea of a normalized spread. We mentioned that during the liquidity regime created by COVID-19, both the bid-offer spread (cost) and depth (availability) had changed, and that we could not directly compare the periods before and during the epidemic. We consider our normalized spread8 metric a more accurate measure of market “liquidity.” Compared to January levels, liquidity costs are still up 322% for S&P 100 and 302% for Russell 2000 stocks. Compared to March levels, these costs are down 52.9% for S&P 100 stocks and 25.8% for Russell 2000 stocks.
Figure 5AB
Figure 5AB. Normalized average spread in bps across stocks in the S&P 100 Index (A) and Russell 2000 Index (B) over three periods in 2020, including the COVID-19 crisis period, shown in navy blue. The spread has been normalized to account for purchase or sale of a fixed amount of liquidity (the average available at the touch in January) in order to compare periods more effectively.
Intraday Dynamics
The charts above show some improvement, though it is clear we have not yet returned to “normal” trading conditions as volatility remains high. In this section, we share updates to measures of intraday liquidity dynamics that continue to illustrate shades of recovery.
Intraday volume
As is shown in Figure 6AB, there was little change to the intraday volume distribution—for mega cap and small cap stocks alike—during the COVID-related period of heightened market volatility, except for minor adjustments in the opening and closing 15 minutes of the trading day. In April, we see the curves remaining stable.
Figure 6AB
Figure 6AB. Average percentage of total daily volume traded in each 15-minute time bin throughout the trading day for S&P 100 (A) and Russell 2000 (B) stocks. Both include January 2020 in blue, the COVID-19 crisis period in navy blue, and April in green.
Intraday spread and depth
Figure 7AB compares intraday spreads from January to the COVID-19 crisis period and April. Time-weighted average spreads for each 15-minute period within the trading day are illustrated, for both S&P 100 stocks (7A) and Russell 2000 stocks (7B). As expected from the daily spread data presented earlier, bid-offer spreads show recovery in each period of day. Figure 7AB illustrates increased spread recovery in the first fifteen minutes over the rest of the day.
Figure 7AB
Figure 7AB. Average spread in basis points in each 15-minute time bin throughout the trading day for S&P 100 (A) and Russell 2000 (B) constituent stocks. Both include January 2020 in blue, the COVID-19 crisis period in navy blue, and April in green.
Figure 8AB shows the corresponding changes to intraday depth, where available displayed liquidity has improved disproportionately. As we observed in the last paper, depth is related to the tick size; as tick size becomes a constraint for bid offer spreads the depth starts to grow. Since spreads remain higher than normal, the tick size is no longer a constraint for most stocks and the NBBO depth remains constant throughout the day.
Figure 8AB
Figure 8AB. Average available depth at the NBBO in USD in each 15-minute time bin throughout the trading day for S&P 100 (A) and Russell 2000 (B) stocks. Both include January 2020 in blue, the COVID-19 crisis period in navy blue, and April in green.
Here again, it is important to consider the normalized spread when taking a fixed amount of liquidity in order to analyze the true scale of change in cost. Figure 9AB illustrates the spread normalized by order size to estimate the real cost of liquidity intraday, comparing conditions during the COVID-19 crisis to normal trading conditions in January and again to April. As before, for S&P 100 stocks, we fixed the traded depth at $100,000 as was typically available to trade at the touch in January; Figure 9A indicates the spread cost associated with trading this typically available size in these stocks, then and now, for a fairer comparison. For Russell 2000 stocks, we similarly fixed the traded depth at $10,000, shown in Figure 9B.
As expected, there is some improvement in normalized spread throughout the trading day. For Russell 2000 stocks normalized spreads remain very high, over 300 basis points in the first fifteen minutes of the day.
Figure 9AB
Figure 9AB. Spread in basis points in each 15-minute time bin throughout the trading day, normalized to account for purchase or sale of a fixed amount of liquidity, $100,000 for S&P 100 stocks (A) and $10,000 for Russell 2000 (B) stocks (the average amount available at the touch in January). Both include January 2020 in blue, the COVID-19 crisis period in navy blue, and April in green.
Realized Market Impact
As in our original liquidity investigation, we want to present not only the factors affecting spread costs but those affecting the cost of larger orders, sized as a percent of the total volume traded on a given day. Ongoing participation as traders access more than the depth available at the NBBO and spread orders over a longer horizon impacts prices over time. Here, we share changes in realized market impact9 for longer-horizon orders, comparing January, March, and April.
Dividing the data into 30-minute bins, we study the effect of large order sizes traded at varying speeds. For each stock, date, and bin, we calculate the total trade imbalance, summing the dollar volume of buy-initiated trades, subtracting sell-initiated trades, and dividing by total volume to determine participation rate. These participation rates represent institutional orders of corresponding size. For example, an observation with a $1M trade imbalance and a negative 10% participation rate serves as a proxy for a sell order of $1M executed at a 10% participation rate.
Recovery in market impact
Market impact and normalized spread follow similar patterns. Both are proportional to volatility and inversely proportional to (the square root) of volume, and we see that reflected in our empirical analysis. Just as spread costs have recovered, realized market impact has started to recover in the month of April, as shown in Figure 10AB. But market impact remains very high compared to January, with larger differences at higher participation rates.
Figure 10AB
Figure 10AB. Realized market impact cost for S&P 500 constituent stocks (A) and Russell 2000 (B). The horizontal axis represents the average participation rate within each group of orders (trade imbalance divided by total volume in each 30-minute bin), and the vertical axis represents the price impact (side-adjusted VWAP performance versus arrival midpoint in each 30-minute bin).
Conclusion
As illustrated, liquidity in the US equities market is starting to improve—as we expect with volatility declining from its March peak—but costs remain quite high compared to January. While single-stock circuit breakers have returned to normal levels, recovery in liquidity is slow. Recovery is more pronounced for highly liquid stocks, but even there, liquidity in the first fifteen minutes of the day remains low and costs high. Liquidity remains particularly low in the first fifteen minutes of the day for less liquid stocks, where normalized spreads are almost 150 basis points higher than in the rest of the day.
4 The market-wide circuit breaker kicks in if the S&P 500 Index declines by 7% or more; in that case, market activity across all stocks is halted for 15 minutes. During the peak of the COVID-19 crisis, it was triggered four times on March 9th, 12th, 16th and 18th.
5 On April 5, 2011, national securities exchanges and the Financial Industry Regulatory Authority (FINRA) filed a proposal to establish a “limit up limit down” (LULD) mechanism as a reaction to the Flash Crash. The rule went into effect market-wide in 2013, pausing trading in a stock for five minutes if the price goes above or below an average reference price in the immediately preceding five- minute period by 5%, 10%, or 20%, depending on the liquidity of the stock and time of day.
6 As before, bid-offer spread is the distance between the national best bid (NBB) and national best offer (NBO) across exchanges. Depth is the total size posted in limit orders to buy and sell in USD across all exchanges at the NBB and NBO. For each stock we analyze a time-weighted average of spread and depth throughout each trading day.
7 As before, we remove the first five minutes of the trading day in calculating daily statistics to improve robustness of metrics, as the first five minutes of the day contain disproportionately more volatility and less volume than later periods. Market-wide trading halts have been removed from all calculations.
8 To estimate a normalized cost of liquidity we use the square root rule, projecting the cost of liquidity increasing by the square root of the increase in liquidity required. In other words, if an investor wants four times the liquidity available at the NBB or NBO, the investor will have to pay twice the bid-offer spread.
9 In calculating participation rate and order size, we use TAQ data including quotes and trades across exchanges from January 1, 2020 through April 30, 2020 for Russell 3000 stocks. We classify each trade as buyer-initiated or seller-initiated using the Lee-Ready algorithm—trades occurring at the NBO are buyer-initiated and trades occurring at the NBB are seller-initiated. [Inferring Trade Direction from Intraday Data; Charles M. C. Lee, Mark J. Ready; TheJournalofFinance Vol. 46 (2): 733-746.]
.svg)
