Costs to Trade and Liquidity
At its most basic level, the liquidity of an investment defines the ability of a trader to convert a large quantity of the asset into cash at a low cost and with little price impact (Pastor and Stambaugh, 2003). The more liquid an asset, the cheaper it is to trade. All other things being equal, an investor would prefer a liquid to an illiquid asset, especially if the expected holding period is short.
The average cost to trade varies by type of security and timing of the transaction. The basic cost structure is familiar. At one end of the spectrum, trading benchmark US Treasury notes costs little per notional; at the other, moving a large portion of small-cap emerging market securities in the after-market is very expensive. For a very popular (i.e., mid-spectrum) market, Aked and Moroz (2015) found in their internal empirical analysis that a small lot of listed US equity securities cost, on average, 3% for each block of average daily volume. Therefore, to trade 10% of the daily volume of a security, the costs to execute, on average, should be about 0.3%.
Most asset classes—stocks, bonds, currencies, commodities, and derivatives, among others—are viewed as trading frequently, and therefore as being liquid. The illiquid designation is generally reserved for assets such as real estate, timber, art, private equity, and hedge funds that trade less frequently and not on an organized exchange. The liquid–illiquid classification is based on average trading costs over time. As we’ve shown, however, the cost of trading is not fixed. During times of economic stability and abundant financial opportunities attracting liquidity providers is easy, whereas at times of greater uncertainty, finding liquidity providers is difficult. When liquidity is scarce, the cost to trade rises as does the opportunity to profit, sometimes significantly.
What Defines Market Liquidity
Under normal conditions, across all markets, potential liquidity providers require price concessions because they assume, or fear, that investors requiring liquidity have an asymmetric information advantage. Brunnermeier and Pedersen (2009) find that liquidity
- is positively related to volatility.
- has a varying impact based on the commonality across securities.
- can at times co-move across markets.
Illiquidity often coincides with times of high volatility, and the nature of illiquidity varies based on the commonality across securities. Commonality defines the rate at which securities in a market move together. Said another way, markets with high commonality see a broad swath of securities reacting similarly to market events, whereas markets with less commonality show a more diffuse response by securities to the same events. In the cross-section, illiquidity opportunities exist in securities that display some amount of idiosyncrasy from similar securities.
Lower commonality, offering more opportunities for idiosyncratic illiquidity, occurs in markets with greater transparency and access to information. These markets usually have less correlated assets and lower average volatility. In contrast, higher-commonality markets are viewed as riskier because they usually have weaker legal protection and property rights controls. As a generalization, developed markets usually have lower commonality, whereas emerging markets tend to exhibit higher commonality.
During periods of crisis and high volatility, investors will unconditionally sell assets from more opaque markets where differentiation is more difficult. In more transparent markets, however, opportunities for segmentation exist, and liquidity is one of the factors investors take into account.
Measuring the Costs to Trade
Market analysts use a number of traditional liquidity measures to estimate the costs of trading. These fall into two buckets: heuristic/market-based methods and theory-based methods. These metrics can apply to any market, but in this article we apply them to the S&P 500.
Market-based methods include 1) the family of spread characteristics: the bid–ask, effective, and realized spreads; 2) market depth, a measure of the quantity of limit orders at either the bid or the ask price, and 3) turnover statistics, which measure average trading activity as a proportion of total fund assets. These methods have the benefit of simplicity and ease of understanding, but they suffer from a high cost of calculation driven by the need for large quantities of data. We find nothing wrong with these measures, but choose a more theoretical, yet equally useful method.
Theoretical methods attempt to capture the two major asset price effects of illiquidity: price impact and price reversion. Price impact is simply a measure of the linear relationship between trading volumes and asset returns, whereas price reversal captures the transitory effect of prices moving away from, and then back to, fair value over the passage of liquidity events. Two popular theory-based metrics are the Pastor and Stambaugh (2003) and Amihud (2002) illiquidity measures.
We use a variation of the Amihud measure, which has the benefit of capturing both the impact and reversal features. Our augmentation scales Amihud’s illiquidity metric to be meaningful as a cost to trade (CTT) measure. The following equation expresses Amihud’s definition scaled by the 250-day moving average of an average security’s trading volume, MAVOLD250-day :