Anyone following the recent headlines coming from Wall Street would be forgiven for thinking that U.S. equity markets have become increasingly alien — if not downright hostile — to investors and issuers alike. Daily trading volume is disappearing into “dark pools,” obscuring prices. Structural changes in market operations are killing the market for initial public offerings by making it less economical for investment banks to take small companies public. High-frequency traders are stoking intraday volatility. Ordinary investors stand no chance against professional traders backed by brawny computers “co-located” in or near stock exchange facilities. Measuring trading times in microseconds, these speed demons chip fractions of a penny here and there from hapless investors with less computing horsepower at their disposal.
Except, of course, when their computers go haywire, which seems to be happening with increasing frequency. This year alone featured the botched BATS Global Markets IPO in March, the flubbed Facebook IPO in May, and the “Knightmare on Wall Street” in August. In the latest of those debacles, a bungled software upgrade sent broker Knight Capital’s computerized trading algorithms on a rogue stock-buying binge, skewing prices for more than 100 stocks listed on the New York Stock Exchange and resulting in a $440 million loss for Knight. The incident was disturbingly reminiscent of the “flash crash,” which sent the Dow Jones Industrial Average down 1,000 points and back up again on May 6, 2010.
All this, many observers contend, has prompted disenchanted investors to pull hundreds of billions of dollars out of stock mutual funds over the past five years. And it has jeopardized America’s place in the world financial order. The moribund IPO market, combined with mergers and acquisitions, has pared the number of publicly traded companies listed on U.S. exchanges in half since 1997, according to James Angel, associate professor of finance at Georgetown University.
U.S. equity markets, in short, seem dysfunctional, if not broken, and many CFOs wouldn’t disagree. “This has to be a contributing factor to the hoards of cash that are just sitting out of equities right now,” says Chad Stone, CFO of Renewable Energy Group.
So why aren’t he and other finance chiefs, who need fair and orderly markets to raise capital for their employers, pressing for stock-market reform?
For starters, CFOs are also quick to acknowledge an alternative explanation for investor unease. That starts with a pair of horrific stock-market crashes in the 2000s, the long-simmering debt crisis in Europe, and the United States’s own long road back from the Great Recession. The scandal-tainted failures of MF Global Holdings and Peregrine Financial Group, the LIBOR rate–manipulation case, and the Standard Chartered Bank money-laundering incident have only added to the sense that the financial markets aren’t a level playing field. “Clearly, trust has been shaken time and time again during the last 10 years,” says Ilya Cantor, CFO of IT services provider Epam Systems.
Besides, some finance chiefs have orchestrated successful IPOs this year, contradicting the idea that going public has become almost impossible. Cantor is one; Epam went public with a $72 million IPO in February. Brightcove, a Boston-based online video hosting company, also completed an IPO in February, raising $55 million. “If you have a quality story in an attractive market, and a product offering you can differentiate from your competition, you can find banks willing to take you public,” says Brightcove executive vice president and CFO Chris Menard.
If the equity markets are not broken, they are nevertheless dramatically different from what they were 20 or even 10 years ago. CFOs who fail to appreciate the evolution risk misjudging its potential impact on their company’s stock, their cost of capital, and their relationships with long-term investors. Here’s a quick rundown on the major changes, and on which you should care about.
1. Brand-name exchanges are now bit players. The New York and Nasdaq stock exchanges are still where most companies go to list their shares, but even more trading takes place now on dozens of broker-dealer trading platforms, electronic trading networks, and those ominously named dark pools. The impact on CFOs? Modest at best. The biggest issue may simply be that there are now more places where computerized trading glitches can occur.
2. Bid-ask spreads are much narrower. Since the Securities and Exchange Commission mandated decimalization in 2001, which resulted in stock prices being quoted in increments of a penny, bid-ask spreads have narrowed and investors have benefited. But decimalization has also squeezed margins for brokerage firms that once relied on those spreads in part to fund research and sales support for stocks of smaller companies. Without that revenue source, critics say, it’s harder for these firms to justify doing IPO work, indirectly raising the cost of equity capital for smaller companies. CFOs acknowledge the issue, but allow that other factors may be contributing to the dearth of IPOs, too.
“I agree with the theory, but I also think we’re seeing more-cautious investors that have been burned by the credit crisis and a tough economy,” says Renewable Energy’s Stone. “In the past, an idea or concept was enough to get investors to buy into your story. Today, they’re looking for a proven business model.”
3. Dark pools are spreading. Dark pools are electronic trading platforms where institutional investors can anonymously trade large blocks of stock, often broken down into 100-share lots. Their goal is to avoid affecting a stock’s price before they can complete their sales or purchases. There are an estimated 50 dark pools now, accounting for 10% to 12% of daily trading volume. Critics argue that by not posting buy and sell offers on “lit” or “displayed” markets, traders in dark pools mask the true supply and demand for stocks, distorting prices. They also contend that dark pools allow some traders to see order flow before others. Last year, dark-pool operator Pipeline Trading Systems paid $1 million to settle SEC charges that it failed to tell customers it was filling most of their orders with one of its own affiliates.
Despite such incidents, Georgetown’s Angel, who also serves on the board of DirectEdge, an electronic stock exchange, discounts the critics’ arguments. For starters, he says, investors don’t need to be able to see every bid and offer to have a statistically valid idea of where a stock is priced. And details of dark-pool trades are made public almost instantaneously once they happen, he adds, usually within milliseconds.
Besides, institutional investors have always tried to camouflage their trading intentions, Angel points out. Under the market’s old structure, for example, it wasn’t uncommon for specialists on the NYSE to keep their entire order book hidden, although the top of the book eventually became public. And institutions would routinely route big orders to the block trading desks of large brokerage firms and ask them to work those orders as discreetly as possible. No one, it’s safe to say, ever announced that they had a million shares of IBM to sell.
Kevin Cronin, global head of equity trading for asset management firm Invesco, has warned that a bigger threat to fair markets is broker-dealer internalization, in which broker-dealers take the other side of customer orders. That represents liquidity other big investors cannot trade against, for the most part, and accounts for a greater percentage of unlit volume than dark pools, he says. Addressing the House Committee on Financial Services on behalf of the Investment Company Institute, Cronin recommended that internalized orders be required to provide “significant” improvement over displayed bid and ask offers, which would reduce their use and so improve price discovery.
4. High-frequency trading is still prevalent, but less so. One of the most hotly contested developments in the equity markets over the past decade has been the debut of high-frequency trading, in which firms use computerized algorithms to buy and sell stocks in the blink of an eye, perhaps millions of times daily, typically closing each trading day with no positions on their books. High-frequency traders generate profits not only by arbitraging price differences across different exchanges and trading platforms, but also by earning “liquidity rebates” from exchanges and other trading platforms in return for sending limit orders their way (see “Inside an Adding Liquidity Only, Post-No-Preference Blind Order,” below). Critics complain that some high-frequency trading algorithms are designed specifically to detect trading of large blocks of stock by submitting, and in many cases quickly withdrawing, multiple orders until they find the price at which orders get filled.
“Their whole job is to sniff around and ping for order flow, and run ahead of it,” says Joseph Saluzzi, founder of Themis Trading and co-author (with Sal Arnuk) of a recent book, Broken Markets: How High Frequency Trading and Predatory Practices on Wall Street Are Destroying Investor Confidence. “All they do is take advantage of liquidity and trade ahead of it. That’s not helping anybody.”
High-frequency traders argue that they improve liquidity for other investors, but critics say that this liquidity comes at a high price: sharp intraday volatility that is scaring off long-term or “rational” investors, ultimately depressing stock prices and raising the cost of equity capital for public companies.
Karen Blasing, CFO of Guidewire Software, which completed a $115 million IPO in January, gives the theory some credence. “I believe high-frequency trading does exacerbate volatility,” she says. “And I think the cost of equity capital is more expensive, particularly for IPOs, as the available window for initial public offerings is significantly influenced by volatility.”
But Bill Stuart, CFO of Synacor, a provider of online content delivery services, isn’t overly concerned, either about the long-term impact of high-frequency trading on his company’s stock price or its long-term investors. “I spend a lot of time talking to institutional investors,” Stuart says. “I don’t think they ignore large swings in the company’s stock if it’s out of the ordinary, but if they can’t see any news driving it, I think they just go on about their business. We’ve come to accept the volatility, and speaking with other CFOs, I think my view is probably consistent with theirs. If you just focus on running the business, the stock price will take care of itself over time. It will eventually go to the appropriate value.”
Whatever ill effects high-frequency trading may have, they may be shrinking. Larry Tabb, founder and CEO of Tabb Group, a capital-markets research and advisory firm, estimates that high-frequency trading will account for only about 51% of U.S. equity trading volume this year, down from 61% as recently as 2009. And HFT profits will likely fall below $1.8 billion this year, down from an estimated $7.2 billion in 2009.
Tabb attributes the falling market share to a marketplace that has become more efficient, making it increasingly difficult for participants to gain advantage over their competitors. “The high-frequency models are interacting with investor models that are more sophisticated and harder to take advantage of,” he says.
What to Do
While there’s no hard proof that the changes in the equity markets are as pernicious as their critics contend, there is a broad consensus that some tweaking could be in order. To counter high-frequency traders using algorithms to ferret out and trade ahead of block trades, Invesco’s Cronin has suggested that regulators may want to impose a fee on canceled orders above a certain ratio of executed transactions. He has also asked whether it makes sense to continue to allow liquidity rebates.
The SEC has been looking into equity market structure issues since January 2010, when it issued a 74-page concept release soliciting public comments. It has taken a few steps since then, such as implementing “circuit breakers” designed to prevent extreme volatility of the sort experienced in the 2010 flash crash. And in September, it handed down a $5 million fine against the NYSE for purportedly giving some customers improper early access to trading information, a charge the exchange’s parent company, NYSE Euronext, neither admitted nor denied.
The SEC also approved a new rule this July that requires national securities exchanges and the Financial Industry Regulatory Authority to create a marketwide consolidated audit trail that will make it easier for regulators to monitor and analyze trading activity. But that’s expected to take years.
James Angel, the Georgetown professor, isn’t terribly worried. “I’ve been a student of the equity markets for 30 years now, and there’s always been a lot of grumbling about market structure,” he says. “In the old days, it was guys grousing about those specialists on the NYSE ripping them off, or somebody else manipulating prices, or the fact that the tape was running so slow that no one knew what the real prices were. When you look at measurable dimensions of market quality, such as execution costs and speed of execution, the market structure looks really good.”
Good, but perhaps not good enough.
Randy Myers is a contributing editor at CFO.
High-Frequency Trading: A Brief History
The seeds of a stock market conducive to high-frequency trading were planted in 1971 with the debut of Nasdaq as the world’s first electronic stock market, but grew mightily in 1997 when the Securities and Exchange Commission issued new order-handling rules for that market. Among other things, the rules forced market makers to begin publicly disseminating prices available on competing electronic trading networks, or ECNs, and to display limit orders that fell inside their own bid-ask spread. As these new order-handling rules helped ECNs attract business, ECNs grew in number and began competing for trading volume by offering liquidity rebates, or what came to be known as “maker-taker” pricing. Such schemes rewarded traders who added liquidity to their exchange by posting limit orders, while still charging those who took liquidity via market orders.
Finally, in 2005, the SEC passed Regulation NMS, which among other things prohibited exchanges from executing a trade at a price inferior to one quoted elsewhere. This meant that high-frequency traders could be assured of getting a trade at the best price if they were the first to post it. Soon, they began stepping up their use of high-speed computers, sometimes using complex computer-managed trading algorithms to arbitrage pricing between multiple markets and securities. The most aggressive co-located their computers near or in exchange-owned facilities to gain a few precious microseconds over their competitors. — R.M.
Inside an Adding Liquidity Only, Post-No-Preference Blind Order
The ability of high- frequency traders to exploit price discrepancies across multiple trading platforms is attributable in part to the complex and varied kinds of orders they can now place. NYSE Euronext, to cite just one example, offers more than 30 order types at its wholly owned electronic trading network NYSE Arca, some of which can be combined with others to create even more specialized orders.
Tim Quast, president of Modern IR, a consultancy that helps companies understand and communicate to shareholders how their stock is being traded, gives this hypothetical — and simplified — example of how a high-frequency trading firm might use one of those order types to take advantage of an M&A announcement:
XYZ Co. discloses that it is buying a competitor in a deal that will immediately benefit revenues, earnings, and market share. The trading firm quickly places an “Adding Liquidity Only, Post-No-Preference Blind Order” to try to earn rebates or trading credits by selling shares of XYZ. This type of order allows the firm to simultaneously post an undisplayed bid order for the same stock; that order will continue to fill as long as other buy orders overlap with it. This, Quast says, allows the trading firm to draw shares from other market centers and quickly earn trading liquidity rebates.
Simultaneously, the firm uses margin to aggressively buy futures contracts on a basket of exchange-traded funds that contain XYZ in their underlying stock portfolios. The firm’s algorithms create the appearance of demand for the shares of XYZ, which changes the price of the future contracts.
In its equity trades, the firm never owns more than 100 shares at once, but it manages to trade a total of 450,900 shares that day, earning a rebate on 90% of that volume. The rebate is 30 cents per 100 shares traded, yielding a gain of $121,500. Ideally, the firm also earns a profit on its futures transactions, where it also collects payment-for-order-flow credits. — R.M.