High-Frequency Trading: What's Next?

Let's pause for a moment, if the lightning pace of high-frequency trading permits us to do so. In the U.S. today, high-frequency, electronic, computer-aided trading accounts for as much as 65 percent of all stock-volume activity.

Computers whiz and hum.  Black boxes send out trading orders in thousands and millions of shares, and rout orders to exchanges and "dark pools" all over the globe. Execution occurs in fractions of a second. Algorithms and programs determine what to buy, when to buy, when to sell, when to buy and sell at the same time and on which one of a dozen or more electronic exchanges. Algorithms provide guidance on volume, prices to show, prices to execute, and prices, if only for a few seconds, to report as "bids" or "offers."

High-frequency traders dart in and out of trading positions in seconds. Some firms buy in one venue and sell in another. They earn pennies per share, but generate large profits via big volume--tens of thousands of shares bought and sold in seconds. Tens of millions of shares throughout the day. Many buy or sell shares in one venue and simultaneously sell or buy the related derivative over the counter, in another country, or in exchange thousands of miles away. Execution and profit-generation are confirmed by the flickering of light on computer screens. 

Traders--or actually, their humming black boxes--study patterns, trends and data. They look for discrepancies, distortions, or something out of line. Traders (yes, humans, often quantitative analysts and experts) update computer code and write more algorithms to instruct their firms to deploy more capital to get into and out of trading positions in seconds. Timing is of the essence. 

Many aim to finish the trading day with neutral positions--little or no overnight risk. Some preside over non-stop trading--trading into and out of positions, making markets, and providing bids and offers all over the globe for a continuous 24 hours.

Some trade other "asset classes," as well, exploring similar opportunities in instruments beyond equities, looking to do the same or find discrepancies and trends in foreign currencies, options, convertible bonds, and government bonds. Or they seek to decipher relationships between "asset classes" (convertibles and equities, bonds and mortgages, options and equities, interest-rate derivatives and bonds). Most trade for their own accounts; many trade or execute for customers and clients.

Over the past several years, they have  include such firms with unfamiliar names as Getco, Jump, Allston, Gelber, Jane Street, Sun and Quantlab. They also include hedge funds and clients of hedge funds and asset managers. Occasionally they may include other types of funds (investment funds, pensions and endowments), all looking for an advantage based on rapid execution and "best prices."

In the realm of finance, some say this is exponential progress. Compare to the more docile manner of trading in the early 1970s, when stock certificates were exchanged, counted, bundled, boxed and rolled into the vaults of brokerage houses all over Wall Street from day to day, creating such a paper-work crisis that the industry once days off to recover from the mounds of paper.

Others say this represents a setback for retail investors or value-oriented investors. Is anybody among the throngs of high-frequency traders buying stock to support a company's investment in a new business, investment in sales growth, or investment in expansion to a new region of the country?  Do they care for more than a few seconds about a company's new-product strategy or its business plans for 2013?

Many high-frequency firms rebut that they contribute to capital markets in several major ways:

(a) They provide market liquidity, active markets, and ready buyers and sellers.

(b) They provide "price discovery" with bids and offers updated continuously during the trading day.

(c) They provide "best prices," opportunities for buyers and sellers to search venues to find the best price for a particular stock.


Their detractors argue they hamper markets in many ways:

(a) Unlike the stock specialists in the past, they disappear when markets become too volatile. They balk or refuse to participate at certain times.

(b) They don't always provide honest, good-faith bids and offers. Sometimes they show their hands and wander away before execution.

(c) Skipping from venue to venue (electronic exchange to electronic exchange) with less-than-sincere bids and offers, they often try to trick or deceive markets to gain information advantages--advantages that slice profits from retail- and long-term investors.


And they cause what happened in May, 2010:  the "Flash Crash," when the market fell (Dow Jones) almost 1,000 points (9 percent)--a precipitous, unfathomable, and bizarre collapse in minutes for no explained reason. Perhaps just as odd was the market's subsequent rebound the same day.

It took months for market experts, regulators, and exchange officials to figure out what happened. Some still don't agree. Most worry that flash crashes, in this new, 21st-century trading environment, will appear regularly. Many are concerned about the impact of a market (or markets covering all asset classes and many geographies) on individual investors. What are the virtues and attractions of a marketplace where the better capitalized electronic traders appear to have an ongoing advantage, where these traders get access to the best prices and best execution, and where 1,000-point, unexplained drops in the Dow are regarded as by-products of the game?

Over the past year, we've seen other debacles and unexpected turmoil in equity markets.  BATS, an electronic exchange, widely known for its swiftness in execution and the technology that supports it, botched and then canceled its own IPO offering earlier this year--because of technology glitches.  This summer, Knight Trading, a market-making firm, botched an electronic-trading vehicle that was intended to allow even retail investors to have better electronic access at the New York Stock Exchange. That led to losses over $400 million and several days of its existence in jeopardy.

What will happen next? And to whom? Will there be another collapse of some kind, something unpredicted, unexpected out of nowhere--blamed on high-frequency traders, but inexplicable or puzzling to the public at large?

Where do we go from here? Do we allow the marketplace by itself to resolve these quirks, collapses, and unpredictable swirls? Or should regulators (from Congress to the SEC and CFTC) rush in to take steps, even as they try to understand trading models that are racing a hundred steps ahead of them. 

Attempting to understand their trading schemes (their intents, purposes and profits) can be a mind-boggling exercise for those who contemplate regulation. What are their strategies?  How do the translate strategies into profits? How do they allocate capital? While the black boxes hum away, how do senior managers stay on top of the activity? Perhaps most important, how do they approach and manage risks--risks to their firms and risks to counter-parties and other traders and investors in the market?

No one knows for sure what the right next steps should be--at least in the U.S. Should there be transaction fees or taxes to restrict such activity? Should there be increased capital requirements for participants--as protection against what would likely be yet the next big loss or flash crash?  Should regulatory review boards approve all trading strategies and trading innovation?

Many of the same trading firms preside over or direct trading into what are called "dark pools"--private in-house marketplaces where electronic firms can exchange thousands of shares without having to let public markets see what they are doing. In some ways in the industry, it appears what could happen next--near month or next year--is like wandering into an unknown, uncertain "dark pool.

We're likely at a precipice.

Technology innovation will continue, as long as there are profit opportunities. Some argue profit margins will decline as the number of participants increase, and that in itself could slow down the rush to be the fastest in executing trades on the planet.  With other priorities on their plates (Dodd-Frank and Basel 3, most notably), regulators won't be able to catch up quickly, always hustling from several steps behind, panting while trying to project what is the worst that could possibly happen.

Meanwhile, feeling disadvantaged and sometimes clueless, worn down by the equity-market tumult from the crisis, retail investors seem to have decided to watch this play out while they remain on the sidelines.

Tracy Williams

See also:

CFN:  Dark Days at Knight Capital, 2012
CFN:  Market Volatility, Can You Stand It? 2011
CFN:  Uncertainty in Markets, 2011
CFN:  Here They Come, the Volcker Rules, 2011

Goldman Tweaks Banking's Ladder

Goldman Sachs leads; everybody else follows. Or most everybody else. So it has been for the past couple of decades in how corporate and investment banks structure themselves and recruit, develop, promote and pay for talent. And so it has been in how banks--from Goldman to regional banks involved in corporate banking and foreign banks that set themselves up on American shores--organize banking units.

Traditionally since the 1980s, most corporate and investment banks (including also their trading, cash-management and processing units) recruit finance professionals into the following "programs," "classes" or "levels":

1. Analysts (those with BA or BS degrees) who join a two-three-year program and who virtually learn from scratch financial analysis, accounting, capital markets and banking on the job, while toiling away long hours doing the dirtiest of work for banking teams (spreadsheets, analysis, projections, document preparation, document printing, research, and, yes, mundane errands for managing directors). They often embark on the role without a clue.  To their credit by the end of the first year, many become somewhat astute about accounting, corporate firm valuation, cash-flow analysis, financial projections and the required regulatory steps to issue new corporate bonds.

2.  Associates (those with MBA degrees or analysts who have been promoted to the associate position) who have substantially more responsibility and more client contact and who are expected to have mastered nuances in corporate finance, capital markets and client industries. They come with experience, maturity and ambition. Some were top-rated analysts at banks before they retreated to Wharton, Darden, or Tuck to polish their understanding of finance and markets. They learn quickly that first year that banking is brutal business, not an academic pursuit.

3.  Vice Presidents who are promoted after spending three-four years as associates and who have more management responsibility and greater access and input to client, deals, transactions, and revenue-generating activity. The pressure builds for them, because performance is tied not necessarily to an in-depth knowledge of markets, financial products and balance sheets, but to their ability to generate revenues and minimize risks.

And depending on the bank, fund, or institution:

4.  Executive Directors/Directors/Principals who have more senior roles in the management of client relationships, trading portfolios, and/or business units.

5.  Managing Directors/Partners who, besides having senior roles and significant decision-making responsibility, have extensive experience and a meaningful ownership stake (or simply own a lot of the stock). They are the ones who have spent practically half a career in the business and are known for their contacts, networks, client relationships, experience through hardships and downturns, and uncanny knowledge of just about anything they touch from day to day.

How and when people are promoted, paid and awarded bonuses have conventionally been determined by "market benchmarks."  But who sets that "benchmark"?

Often it is Goldman Sachs or occasionally one of its peer firms (JPMorgan, Citi, Morgan Stanley, Merrill Lynch, e.g.).  Sometimes it is set by the industry's efforts to ward off other competing finance sectors or industries.  For example, in the technology boom/bust of the early 2000s, banks felt threatened that dot-coms and other start-ups would seize their more talented recruits.  Some big banks even started a brief trend, as an added incentive, of permitting analysts and associates to invest in private-equity deals the bank had been involved in.  Banks regularly have reshaped, rejuggled and revamped development programs to attract analysts and associates who would otherwise be more interested in technology start-ups, hedge funds, venture-capital outfits, consulting or non-profits.

For years, however, banks have shown a tendency to copy or borrow the programs, titles, bonus payouts, perks and promotion standards of other banks.  Often it has been Goldman that started a trend or Goldman that raised the bar for base pay or bonus ranges.  Other banks--with the help of industry compensation experts and professional recruiters--would peek and try to match the "market" (the market set by Goldman) or at least come somewhat close to it.

That would explain, for example, why compensation offers at most of the top banks, funds, and institutions tend to follow similar trends and tend to be quite similar in amount or package content. If Goldman decides, as it has done so in the past, that MBA associates should be paid bonuses partially in common stock. Then others follow suit. If Goldman changes its mind, as it has done in the past (early 2000s), then others do, too.  Some call it copying; others call it competing to stay in the game.

So whenever Goldman makes a move and tweaks some part of the ladder, the industry watches, waits and then reacts.  Last week, it decided to change elements of its analyst program a wee bit.  The change, announced in the Wall Street Journal, will have a trickling effect on other corporate and investment banks.

The change was not major; it was a modification, not an overhaul of the program. But because it was Goldman, it will be discussed, examined and probably duplicated by others.  It decided not to hire analysts on a contract basis for two years and not to guarantee a base level compensation during that period.  Analysts, therefore, can be dismissed for non-performance, and on the way out, they will not be awarded a thanks-for-your-services bonus.  While on paper or in concept this doesn't appear to have any impact on corporate or investment banking, because it's Goldman, it's news.

What does it all mean anyway?

The bigger story may be that Goldman, like all banks, continues to be under burdensome pressure to manage costs when banking activity is annoyingly volatile and the industry is under siege from regulators and the public. One incremental way to manage banking-personnel costs is to start at the bottom:  Don't hand out gobs of cash to under-performing or disinterested financial analysts, some of whom aren't interested in long-term careers at Goldman anyway and have an eye on applying to business school or joining a Greenwich-based hedge fund.

One immediate effect on the young graduates is that the tweaks will make the tough, grinding, difficult role of an analyst even tougher. The life of the banking analyst is one of 80-100-hour work-weeks, a grueling physical existence where one is on call all the time, even on weekend evenings.  But the pay-offs were always an extraordinary bonus payment, a modicum of prestige, and the opportunity to learn massive amounts of corporate finance and capital markets almost overnight. Goldman's tweak adds pressure. Analysts now won't have the luxury of failing or slipping up, as they (right out of school) adjust to deals, markets, spreadsheets, analysis, research, irascible vice presidents, and merely the real world in general.

Another implication is that if banks tweak the analysts program, they could easily tweak the MBA associates program next. Behind closed doors right now, senior bank managers might be huddling to decide how to manage the costs of another class of new associates from top business schools. They could agree to compensate the new bankers and traders handsomely (as they have always done), but they could decide they must reorganize the program (including compensations and performance expectations) to limit overall costs in other ways:  dismiss "under-performing" associates sooner, reduce the number of associates who are on track to become vice presidents, or offer compensation packages of greater portions of restricted stock. The senior bankers could also decide to encourage associates who perform adequately, but whose hearts aren't in it, to leave.

In its lofty role of setting industry standards and setting the "market" for compensation and promotion, Goldman knows by now its incremental steps will probably be copied.  It's not only Morgan Stanley that might follow suit, but the modest-size boutiques, the regional corporate banks, or the West Coast investment-management firm, too, will adjust.

Many say the roles, compensation and lure of corporate and investment bankers have hardly changed in the past several years. But with these kinds of tweaks and revisions, that new day could be dawning.

Tracy Williams

See also:

CFN:  Forced Ranking:  Does it hurt? July-2012
CFN:  The Dreaded Performance Review--2011
CFN:  Becoming a Top Performer--2009
CFN:  Is I-banking still Hot?--2011
CFN:  How Does Goldman Do It?--2010
CFN:  Mastering Technical Skills in Banking--2010