By Phil Mackintosh, Chief Economist, Nasdaq
With the growth in retail trading this year and plans to deepen the Securities Information Processor (SIP) data that retail uses, it’s time to look in more detail at who exactly is a retail trader, what they trade and how the market works for retail investors.
How big are retail assets?
The Federal Reserve has a report that looks at the balance sheets of households and the owners of corporate securities. Household stock ownership sounds like a plausible way to answer the question, “What is retail?”
Based on that, retail investors own 77% of the market capitalization in total via stocks (held directly), mutual and pension funds. Some would even argue that all three categories are “retail assets,” it’s just that funds are “bundled” and also managed by professional investors.
Chart 1a: Assets from Fed Z1
Having a professional manager makes a big difference in how much regulators typically need to protect those funds. Although it is worth noting that many of the directly held shares are also “managed” by professionals, just Registered Investment Advisors (RIAs) instead of mutual funds.
Unfortunately, data also shows that stock ownership is concentrated in wealthy individuals. In fact, Gallup suggests that after adding direct stock ownership, mutual funds and 401ks together, just under half of all households still have no exposure to the U.S. stock market and the dividends and investment gains they provide. That hurts the long term financial stability of lower-income households.
Chart 1b: Concentration of retail wealth
Source: Nasdaq Economic Research
Regulators have a variety of rules to protect smaller investors
Because sometimes finance and investing is complicated, regulators tend to protect retail investors more than others. In contrast, regulators generally free institutional investors to choose for themselves how to trade. Academic studies suggest that helps both.
Consequently, there are a number of trading rules that apply differently to “large” or “active” traders. But how they do that isn’t entirely consistent.
Table 1: Different rules that protect smaller traders
Source: Nasdaq Economic Research
None of these prescriptive rules consistently capture the same orders or customers, nor do they actually define a retail investor. In fact, most actually catch a combination of retail and institutional trading.
Retail investors are real people
Retail investors come with a variety of wealth and sophistication, but at a high level, they’re expected to be people, not firms or computer-driven trading strategies. So another, more intuitive approach is to look for orders from a “natural person.” Reg NMS doesn’t mention a “natural person,” but other rules, such as SRO and FINRA regulations, suggest it:
- Includes orders on behalf of an individual or family, even if stocks are held in a company or IRA.
- Excludes orders from a trading algorithm or other computerized methodology.
On that basis, some professional investors working for institutions would also qualify as natural persons. So, there are other rules that define, instead, what an institutional investor is. For example, FINRA Rule 4512 defines “institutional accounts” to be a:
- Bank, insurance or investment company;
- RIA;
- Anyone with assets over $50 million (even if they’re a natural person).
The inclusion of natural persons as institutional investors is somewhat circular. This isn’t an error; it just highlights the fact that additional rules are intended to protect the less financially aware retail investors, while limiting the regulatory costs for those with more experience.
Some natural persons are also classified as “accredited investors.” These are natural persons with assets of $1 million or income of $200,000 per year, who are allowed to invest in private companies directly. During 2020, the SEC broadened the rule to include those with FINRA licenses, knowledgeable employees of the private company and family offices with more than $5 million in assets. The SIPs have a similar (but different) way of categorizing a “professional investor,” which affects what they pay for data, which is based on:
- Registered representative with the SEC, CFTC etc.;
- “Investment advisor;”
- Bank employee.
Retail investors mostly use retail brokers and advisors
What an institutional investor needs from their broker is very different from what retail investors need.
Larger institutional orders are orders typically “worked” during the day using sophisticated algorithms to minimize the impact. Brokers have discretion and time to try to find an offsetting natural order to match with.
On the other hand, retail investors’ information and system needs are very different. They each have fewer assets, their trades are smaller, they trade at human speed, pay less in commissions (now mostly zero commissions) and tend to send orders with no discretion (covered orders) using simple limit or market order types. That’s also why the new 606 Rules designed for institutional reporting also used a version of non-covered (non-directed) orders to define what trades to classify as “institutional.”
Because of their different needs, retail investors mostly use dedicated retail brokers who specialize in specific trading systems and settlements.
Not all retail brokers make the same disclosures. Large brokers Fidelity and Vanguard aren’t public companies and many private banks managing retail assets are subsidiaries of larger international firms. But we can use the disclosures of those that do to see that average trades per account (Chart 2a) has increased significantly this year, having been triggered by widespread adoption of zero commissions by retail platforms in late 2019 and boosted by Covid-19 quarantines in March 2020. However, since then, some significant M&A in the sector is making comparisons a little harder.
Chart 2a: How much do retail trade?
Retail disclosures also show that most retail accounts trade infrequently. For many retail brokers, their average customer trades around once-per-month (Chart 2b). But data also shows Interactive Broker’s customers are far more active, while a recent Vanguard study showed their retail customer base traded even less. Only around 20% of Vanguard households trade per year and even the most active quintile of customers average only 13 trades per year.
All of which highlights that retail investors exist on a continuum that ultimately overlaps with less active professional investors.
Chart 2b: How often do retail trade?
However, as a group, it’s been reported that retail trading has increased to as much as 20% of all trading on the U.S. equity market. We should highlight that it represents one side of each trade, but regardless, it represents a significant and growing amount of U.S. liquidity.
Chart 2c: How much do retail trade?
That means most retail orders are:
- Small;
- Covered (not worked);
- Sent via retail brokers or private wealth firms.
Why is that important? It helps find data that hints at how and what retail trade.
Reg NMS requires retail brokers to report publicly execution quality metrics on covered orders (under Rule 605) as well as which brokers they use to execute (Rule 606).
Using 605 reports from wholesalers who fill retail brokers’ orders, we can see that most trades are small (Chart 2d). More than 65% of orders are for trades under 500 shares, representing an average trade less than $12,000. Fewer than 3% of orders are for trades greater than 5,000 shares, which still represents trades averaging less than $150,000.
Chart 2d: How large are retail trades?
So almost all retail trades are well under the $200,000 or 10,000 share levels that protect “small traders” in Table 1 above.
How do retail trade?
We’ve alluded to the fact that retail investors don’t generally work orders. With most orders representing less than the depth on the far-touch, they don’t need to. But that is also important in understanding how retail actually do trade in our modern markets.
Reg NMS Rule 612 requires all limit orders to be sent into the market at decimal price levels. That’s why we typically see quotes on TV and our screens in whole cents.
However, the rule says nothing about prices for “trades.”
Because retail orders (being smaller and less informed) represent fills that are less likely to cause adverse selection, market makers can capture more spread on retail orders than other orders. So much so that for most retail trades, market makers are willing to share some of the NBBO spread with investors. This is done in the U.S. by filling market orders at prices better than the National Best Bid and Offer (NBBO), mostly at sub-decimal prices. A practice known as “price improvement.”
For retail brokers, it’s hard to argue that beating the NBBO isn’t “best ex.” As a result, a majority of marketable retail orders are sent to wholesale market makers for (off-exchange) executions.
In fact, NMS Rule 605 even makes it pretty easy for retail brokers to measure their wholesalers’ performance. Because these orders are “covered” (no discretion), 605 requires the effective spread paid and the on-exchange NBBO at the time to be calculated for all covered orders. This is often converted to a simple ratio: E/Q or effective spread/quoted spread.
Chart 3a: How price improvement works
Source: Nasdaq Economic Research
In fact, metrics from a group of wholesale brokers in January this year show that 85% of all market orders were price-improved. That results in a relatively small proportion of retail trades actually reaching exchanges.
In a similar way, institutional brokers, who serve institutional asset managers, use their dark pools to also try to match trades before searching for liquidity on exchanges.
As a consequence, the U.S. liquidity pie actually looks more like three pies, where the “glue” holding all prices together, and where unmatched trades can still occur, is exchanges.
Chart 3b: A simplified view of typical order routing and venues making up the U.S. liquidity pie(s)
Realizing not being able to trade in sub-decimal prices made it difficult for exchanges to compete for retail flow, the SEC also allowed exchanges to trade (with retail) at sub-penny prices starting in late 2012, via special Retail Price Improvement Programs. In order for brokers to qualify to use these, FINRA required them to confirm that “substantially all” orders being sent also originated from a “natural person.”
What’s the difference between PI and PFOF?
Similar to how rebates reward liquidity providers on an exchange, price improvement (PI) and payment for order flow (PFOF) are both used to reward retail brokers for sending flow to wholesalers. There are also rules to provide transparency on these incentives to further protect retail investors:
- Rule 606 reports provide transparency into PFOF;
- Rule 605 reports quantify PI.
Industry estimates peg PFOF at around $900 million in 2019, with PI even higher. In total, the amount returned to retail brokers and their customers is more than $2 billion annually.
To put those economics in context, the costs of the SIP data that is used to ensure these are all good fills for non-professionals is around $93 million per year, based on public UTP and CTA data. So retail SIP data costs less than 5% of their combined PI and PFOF.
In fact, combined PI and PFOF is also more than double what all U.S. exchanges made for data and colocation in the year. An especially relevant point in light of the SEC’s recent actions to remake the SIP.
What stocks does retail like to trade?
A few years ago, FINRA expanded off-exchange transparency to show what tickers were trading where off-exchange. Those reports show volumes by ticker for each market center and group Alternative Trading Systems (ATSs, including dark pools) and non-ATSs (other facilitations) separately.
From Rule 606 reports, we know that the big wholesalers are also facilitating the majority of retail trades. All those trades are reported to the non-ATS TRF data. Comparing facilitation volumes to market-wide volumes, we can see which tickers have an unusually high proportion of retail activity.
The results in Chart 4 are revealing. The chart shows turnover on the vertical axis, so company trading is adjusted for their stock price and market cap to make them all “equal” for comparing liquidity.
- Something you can’t see from the scatter plot is that microcap stocks (left side of the chart) are, on average, much less liquid. The purple color means you can see they also have a high proportion of retail activity. Both make sense, considering they are not in any major institutional indexes.
- For large-cap stocks, an active retail investor base can significantly increase liquidity versus similar sized stocks (high dots are all purple). As we found, that additional liquidity can lead to tighter spreads, which in turn lower costs of capital.
- Purple dots also represent a pretty high level of non-ATS activity, at around 50% or more of all trading in those tickers. And remember, that’s not including dark pool trading.
Chart 4: The purple stocks have more retail participation
What does this all mean?
There are a few important points here.
First, public data does not easily distinguish retail trading from other trading, but knowing how retail trade, we can get indications from different sources. Some of that data also shows that retail accounts range from very passive to quite active traders.
Second, is that retail is mostly “natural persons,” which means they trade at human speed. That’s in contrast to institutions, which use sophisticated broker algorithms.
In any case, retail trade mostly off-exchange and rely on SIPs only for their post-trade EQ metrics. Those, in turn, show if they are getting better executions than institutions. Mostly they are.
However, not everyone wins in this scenario. Lots of research suggests that gains to spread capture in one market causes equal but opposite declines in other markets. Although retail does better than the SIP, it’s important to consider: Is the price on the SIP right?
