A common view amongst traders in APAC, especially amongst overnight traders covering APAC, is that liquidity is hard to come by in the non-major markets. The different nuances of these markets create vicious cycles that could grind trading to a halt or kick up a flurry of activity; wider spreads lead to more patient quotes, which lead to longer queues. However, all it takes is a spike in volume caused by a single impatient trader or block to trigger a series of other trades (or even other blocks) from traders who thought they missed out on volumes, only to have the stock fall back to a lull once all the “catching up” is done. The common measure of trade difficulty, quantity as a percentage of ADV, does not reflect the difficulty of having to painstakingly track these volume surges. Also, the ADV contains traded volumes that may not be completely appropriate to compare against an intraday order. We will look into the potential discrepancies in how the ADV is calculated, and also suggest a strategy for capitalizing on volume surges.
The various types of market sessions and trading mechanisms across APAC is problematic for traders managing regional lists, even for a simple task as gauging orders’ liquidity constraints using ADV. Because traders typically conduct the bulk of their transactions during regular market hours, a fair measurement of order difficulty should only consider intraday lit market volumes just like what is normally considered for VWAP benchmarks. Unfortunately, total volumes reported for individual stocks indiscriminately include ALL trades, so when these numbers are used to create the ADV, order difficulty as measured by the order’s quantity divided by the ADV can sometimes be understated by quite a large margin. Table 1 summarizes the magnitude of differences between ADV (based on Bloomberg volumes) and intraday VWAP volume, and the typical source of the differences. In some APAC markets (e.g. Malaysia, Indonesia), a post-close session allows traders to absorb liquidity at the closing price, and the volumes transacted during this session will count towards the stocks’ volumes for the day. However, since these post-close transactions can potentially come from clean up trades/error unwinds that are not consistent from day to day, they should not be considered as a stock’s average volume, which the VWAP volumes exclude. The various types of intraday crosses reported in APAC markets also give a false sense of abundance in liquidity if they are included in ADVs, and this is especially damaging for small and mid cap stocks where block volumes can mask multiple days of inactivity in the market. These results remind us the need to scrutinize the stats and averages that are thrown up on the screen: an order that is supposedly 20% of ADV in Malaysia can potentially end up being over 28% of intraday volume, depending on how the ADV is calculated, and blindly throwing the order in an algo could potentially exert undue pressure on the stock.
Many traders consider trading in blocks as being the preferred way to alleviate market impact, and our study shows that these block prints tend to be clustered, therefore the best course of action following the execution of a block is to stand pat with the balance for a short time for the potential of additional blocks, electronic or upstairs. Chart 2 shows the proportion of blocks in a day that are within the same minute wherever they can be accurately observed (i.e. not in Taiwan or Korea where negotiated blocks typically print after hours). In fully-attributed markets such as Indonesia and the Philippines, we see very high levels of “block clustering”; 44.91% of all blocks for the same stock in Philippines occur within a minute of one another, and this metric is 18.25% in Indonesia. The degree of clustering in these 2 markets are higher than others in APAC suggests that traders are efficiently making use of broker IDs from reported trades to track down liquidity. Surprisingly, clustering is also quite apparent in non-attributed APAC markets such as Hong Kong and Singapore, where over 10% of blocks occur within a minute of each other, and as more time passes the likelihood of another block trading also diminishes. With block liquidity being difficult to source at any specific time during the day, our data shows that a traded block actually serves as a good indicator of additional liquidity to be had.
Over the years, our value of the block studies have also shown that price volatility is heightened by varying degrees around the report of a block trade. Traders can actually take advantage of both these byproducts of blocks – volatility and clustering – to capture additional liquidity around a large trade for potential price and quantity improvements by passively participating in all available venues immediately after a print. Following the positive experiences from the pilot program in the U.S., Liquidnet Surge Capture will also be made available in APAC as an option to soak up additional liquidity subsequent to a block execution. Surge Capture in APAC will be customized to take advantage of the diverse types of venues available in each market, while giving traders extra time within Liquidnet for additional liquidity due to potential clustering. Innovation is key to maximize liquidity in APAC where the ADV, as we have seen, often masks how difficult it is to trade in certain markets.