Equity Markets and Stock Instruments › Limitations and Practical Considerations

Understanding equity markets requires acknowledging their imperfections and the challenges they present for quantitative practitioners.

Market data quality presents ongoing challenges. Quotes can be delayed, or become stale when an update is missing or the last observation is old. Check the feed’s timestamps, delay policy and session coverage before treating a displayed quote as executable. Historical data requires careful adjustment for splits, dividends, and delistings. Survivorship bias arises when the tested universe is selected using later survival. It can overstate historical returns, particularly when failed securities and their losses are omitted, but its direction depends on the selection rule. Use the point-in-time eligible universe and handle delistings, acquisitions and associated returns consistently with the strategy.

The bid-ask spread represents a hidden cost that can dramatically affect strategy profitability. A strategy that earns 0.1% per trade looks attractive until you realize that the round-trip spread costs 0.2%. This simple relationship can be expressed as:

$$\text{Net Return} = R_{gross} - C_{spread}$$

where:

• $\text{Net Return}$: the actual return after accounting for transaction costs
• $R_{gross}$: the gross strategy return before transaction costs
• $C_{spread}$: the round-trip spread cost (approximately twice the half-spread for a buy then sell)

In our example, if $R_{gross} = 0.1\%$ and $C_{spread} = 0.2\%$, then $\text{Net Return} = -0.1\%$. The strategy loses money despite appearing profitable before costs.

High-frequency strategies must model spread dynamics carefully; even longer-horizon strategies should account for transaction costs in backtesting.

An order can create substantial execution pressure when its size is large relative to available liquidity and trading volume. Buying 100,000 shares may consume several price levels in a shallow book, but share count alone does not determine impact. Execution algorithms can split and schedule trades to manage costs and risk.

Liquidity varies across stocks and through time. An SEC study of small-cap market quality describes Stock D, with a market capitalization of $536.2 million, as tending to have 200 or fewer shares available at spreads below $0.15. That is a single-stock example of shallow depth, not a spread estimate for all small-cap stocks. Even liquid stocks can become illiquid during market stress. During the flash crash of May 6, 2010, market participants withdrew and quoted spreads widened sharply. Price declines and ETF discounts are different measurements from bid-ask spreads.

The metrics discussed in this chapter (P/E ratios, market cap, and dividend yield) provide useful starting points but do not tell the complete story. Two companies with identical P/E ratios may have vastly different growth prospects, risk profiles, and accounting treatments. Quantitative analysis typically combines multiple metrics into composite factors or models. We address this topic in later chapters on factor investing and machine learning applications.