Paper Points to Rise of Odd Lots

Odd lots—which are not reported to the consolidated tape—have grown as a percentage of shares traded in recent years, and could now make up as much as 20 percent of market volume.

That’s according to a recent paper by Cornell University’s Professor Maureen O’Hara, together with Chen Yao and Mao Ye of the University of Illinois.

The paper, which looked at odd lots and their impact on transparency, estimated that executed orders of fewer than 100 shares have grown from about 2.25 percent of total volume at the beginning 2008 to about 4 percent of volume at the end of 2009. In the past two years, they have grown even more.

Experts in the industry say it is very difficult to know how much of current trading comes from odd lots, but some estimates run as high as 20 percent.

“The interesting thing about odd lots is they used to be sort of a backwater,” O’Hara said in an interview with Traders Magazine. “Odd lots now are actually being used via a very different clientele of people.”

Titled “What’s Not There: The Odd-Lot Bias in TAQ Data,” the paper examines how a lack of transparency in odd-lot trades has led to missing data that could skew perceptions of the market.

“Part of what’s happening is algos have now changed how you trade, and so it’s not that uncommon to have an order chopped up into lots of little pieces,” O’Hara said. “Those pieces can all be odd lots, and those things can all fall outside of the consolidated tape.”

The fact that odd lots are not reported to the tape provides an incentive for well informed traders to use odd lots rather than more visible trade sizes, the paper found.

Though the Securities and Exchange Commission recently reaffirmed its policy that odd-lot trades should not be reported to the tape, the authors questioned that move, noting that fragmentation, high-frequency trading and the widespread use of algos have changed markets in fundamental ways.

Last year, NYSE Euronext came out in favor of requiring odd-lot transactions to be reported in consolidated trade data. The exchange operator noted that a significant volume of trading is attributable to odd lots, and reporting those trades would improve the accuracy and reliability of market data provided to the public.

That position represented a turnaround for the exchange, which had previously objected that there were technical hurdles to reporting odd lots to the tape. Now, however, NYSE says any system capacity constraints that formerly precluded inclusion no longer exist.

It was once thought that odd lots were generated primarily by retail investors, but since any trade under 100 shares is considered an odd lot, medium-size trades of expensive stocks often fall in the odd-lots category.

For instance, since Google trades at more than $500 per share, a trade of $50,000 of Google stock would still be considered an odd lot. A full 34 percent of all trades in Google are odd lots that do not report to the tape, the paper found.

“There actually are a number of stocks now that have very high prices,” O’Hara said. “Google is a great example, but Apple’s up there now, and so are others.”

The most popular size for an odd lot is 50 shares, but after that, the next most popular size is for a single share. The authors of the study said this was likely due to high-frequency traders trying to “ping” the market with one-share orders to detect hidden liquidity.

HFT is now the norm in the equity markets, the paper said, as high-frequency traders are now involved in as much as 73 percent of all trades. With the rise of HFT and efforts by institutions to break up their orders to prevent gaming by high-frequency traders, smaller lots are now vital for price discovery.

According to the paper, 30 percent of price discovery is accounted for by trades of 100 shares or less.

O’Hara said the fact that odd lots are correlated with price discovery is important, as it speaks to the fact that small trades now are being used differently than in the past.

“It’s time to revisit this area, both from the point of view of regulators, and also to just understand the new dynamics of the market,” O’Hara said. “Every trade now is slightly different than it used to be.”

One veteran broker who asked not to be named said he had been concerned about odd lots for a number of years. In his opinion, the SEC should start requiring that all trades, no matter how small, get reported to the tape.

“They’re not irrelevant today,” the broker said. “You shouldn’t be able to trade 99 shares all day and not have it represented.”

Alison Crosthwait, managing director for global market structure research at Instinet, agreed that regulators should consider having odd lots printed to the tape. She said the large number of odd lots is a concern, especially if traders are intentionally splitting up orders just to avoid the tape.

But Jim Toes, president and chief executive officer of the Security Traders Association, said any benefits of reporting to the tape should be weighed against the costs involved with changing the system.

“Our organization always encourages cost-benefit analysis be done,” Toes said. “Regarding odd lots, the cost for having this activity being added to the consolidated tape should be compared to the benefit investors will gain with the increased transparency.”

In addition to considering the financial issue, regulators should look to prioritize odd-lot reforms with the many other issues the SEC is currently tackling, according to Toes.

Still, he said the academic paper has already had a big impact on his organization’s members.

“I think the industry is familiar with the report,” Toes said. “And I haven’t heard anybody dispute the data.”