Liquidity is the ability to buy and sell a financial security at a price reflecting its value. It is important because it helps investors trade in an out of stocks more cheaply.
It’s also important because less expensive trading should usually lead to higher valuations.
But how exactly should we measure liquidity and how has it changed over time?
Is liquidity the depth of the market, or is it the average trade size?
Those who say liquidity has fallen often point to the smaller quantity on the NBBO and the smaller average trade sizes.
Chart 1: Over time, the average trade size has fallen along with spreads (bps)
However, those data points are complicated by the fact that over the same time, spreads have fallen while ADV has significantly increased (Chart 1). Consistent with our findings in V is for volume, we should expect the depth at these new narrower spreads to fall, which in turn will reduce average trade quantities. That finding suggests an apples-to-apples comparison requires us to look at the cost of liquidity at a consistent level, which we will do below.
Is liquidity shares traded?
Importantly, Chart 1 also shows that cheaper trading (spreads fell over 90%) led to an explosion of liquidity. Shares traded (ADV) is around 10x higher now than in the early 2000’s. From that perspective, the market is much more liquid today. Put another way, an order that is 30% of ADV actually represents a much larger proportion of shares outstanding, than it did 20 years ago.
Is liquidity more about value traded or turnover?
To really compare this measure of liquidity it makes sense to compare the U.S. to other markets.
According to a combination of Nasdaq, Credit Suisse and World Federation of Exchanges data:
- The U.S. equity market trades close to $325 billion in notional value each day (of which roughly 28% is made up of ETPs), with a market cap of $35 trillion. That gives an average corporate stock turnover of around 1.7x.
- In contrast, Europe has a market cap of around $18 trillion, which is about 50% of U.S. market cap, but trades just $37 billion each day. That’s much lower turnover—around 0.5x
- In fact, on these metrics Asia is more liquid than Europe. It trades $85 billion each day on $28 trillion market cap, a turnover averaging 0.8x.
Based on this metric, in comparison, or standalone, U.S. liquidity looks very good.
Chart 2: U.S. value traded is nearly 9X the amount traded in Europe and 4X the amount in Asia
Is liquidity about the cost to do the same trade?
There is an entire industry dedicated to collecting and computing trading costs. The data overwhelmingly show that transaction costs are falling, and that U.S. costs are the lowest in the world.
For example the Research Affiliates estimates in the chart below show a drop in costs for large trades of almost 70% in the past 60 years. That’s before you realize that “average daily volume” has significantly increased, especially over the past 10 years, meaning this is measuring much larger trades in later years than at the start.
On this measure liquidity has improved.
Chart 3: Costs for large trades have fallen
Depth of book – better or worse?
Let’s come back to the NBBO, as it represents the costs of instantaneous liquidity (rather than large worked orders), and try to account for narrower spreads vs. depth problem.
Data from the tick pilot found that when spreads widened, the cost of trading to that level was the same or more, even though the depth on the NBBO increased, thanks to the lack of fills at smaller increments.
Put another way, with tighter spreads, you can afford to sweep multiple levels of lit liquidity to get to the same cost as the old, wider, quotes. In fact, you can sweep many more than 6 cents from mid for the same cost as an old 1/8th (12½-cent) spread.
Citadel recently published an interesting study looking at depth in exactly this way over the last decade. They computed the cost of sweeping the book to instantly complete trades of $1 million, $10 million and $100 million. They did this for a large cap and small cap portfolio trades. What they found was:
- The cost of liquidity was essentially unchanged over the decade.
- Although the cost did vary, it was mostly due to volatility spikes.
- Inappropriate tick size also hurt specific stocks.
- Small cap costs were around 5x higher than large cap, for the same notional (Chart 4), which is probably pretty good, given a $10 million trade in small cap will buy you much more ownership in each company.
- Costs for trades up to $10 million were only marginally more than the actual NBBO spread, for large cap and small cap.
- However a $100 million trade required more stocks to trade through multiple levels on some stocks, and was closer to 10x more expensive in small cap stocks.
Chart 4: How the costs of sweeping the book to fill a portfolio trade increases with trade size
Interestingly, the same can’t be said for futures.
Despite U.S. futures trading $310 billion each day, around 32% more than stocks, data from Credit Suisse shows the depth on the quote has fallen dramatically since 2017, even though ticks have not changed (much—technically, as the index increased in value, the tick has fallen in basis points).
Chart 5: The value on the top three levels of the S&P 500 futures has fallen since 2017
Despite automation, indexes and ETFs, stock liquidity is stronger than ever before
Most of this data shows the cost of instant liquidity has been consistent or improved, while trading costs on worked orders have fallen.
That’s despite the growth of index funds and ETFs, and despite the automation of the market and High frequency traders. Also despite the fact that spreads have compressed as markets have automated.
We get it, change (and robots) are scary. But markets are cheaper and more efficient now than ever before.
There might only be 500 shares on the bid, but there is a lot more liquidity behind.
