Rewriting the Rules for Financial Trading Infrastructure: Going With the Flow

John Knuff

<<<>>>>

By John Knuff (Part 3 in an 8-part series)

<<<>>>>

New Rules

  • The ever accelerating pace, scale, and diversity of markets
  • Agility demands non-stop adaption

The pace of change just keeps on accelerating.

Despite thin markets, peak message rates continue to climb. In the market sell-off of August, 2011, over 5 million messages per second flashed briefly across U.S. equities, futures and options markets.

The peak passed 6 million in September, reaching 6.65 million in October with speculation over the sovereign debt crisis. (Based on Exegy data at http://www.marketdatapeaks.com/)

This was up 137% on the flash crash of May 6, 2010, but without the earlier trauma. Traders only cope with these volumes when colocated with the markets and with agile supply chain partners to support them.

U.S. Equity, Futures and Options Markets: Growth in Peak Messages per Second. Source: Exegy, Inc.

<<<>>>>
Europe too has seen a surge of fast trading, although numbers are still an order of magnitude behind the U.S.

Meanwhile, in Asia ever-faster markets are appearing with the Singapore exchange recently rolling out what it claimed to be the world’s fastest trading engine with latencies for order entry response times below 90 microseconds in colocation. Practice however varies hugely between venues.

Some technology firms are even forecasting 20 microsecond round trips. Meanwhile, ticker handlers and pre-trade risk checks are driving down into nanoseconds with field programmable gate array (FPGA) and 10 gigabit Ethernet technologies.

As latencies converge on zero, the cost to achieve each small advantage increases rapidly with ever shorter half-lives. This has squeezed HFT margins and driven the big money managers into dark pools to minimize slippage against their target price.

As the technology arms race intensifies competition, traders have diversified their strategies, seeking smarter as well as faster solutions to avoid crowded trades.

They are exploring new markets and asset classes using increasingly diverse data sources, and deploying distributed, highly parallel, and data driven strategies to cope with the volumes.

No wonder multitenanted network neutral data centers are so attractive, since they offer the greatest choice of connectivity and e-services providers.

Encouraged by cheap money and the search for yield, portfolio investment flows with emerging markets have taken off.

Where U.S. trading once accounted for over 70% of global equities, it is now around 37%.

Meanwhile Asia Pacific exchanges have grown strongly, with Europe somewhere in between. As real money shifts, the high frequency traders follow, and exchanges in emerging markets have welcomed them with new technology and services.

This has led to increasing correlation and a single global market, operating 24×6.

However, as one U.S. portfolio manager observed, “We’re not just talking here about Chicago traders in Europe or Dutch traders in Asia. We’re now seeing Asian traders arbitraging U.S. and European markets or even other emerging markets, such as Latin America or Eastern Europe.”

A truly global geographical trading landscape is fast becoming a very small and crowded world.

This year’s Automated Trader survey of over 500 global algorithmic traders confirmed strong growth of distributed trading along with colocation and proximity hosting, particularly among day traders.

Over 50% of buy side algorithmic trading firms in the survey had more than one colocation or proximity hub, while 77% of sell side firms had multiple hubs.

Of these most firms had some degree of dynamic placement of orders across the various locations even if it was just a fixed rule.

However, once placed most orders executed independently of what was happening elsewhere at other hubs.

Only around 15% of buy and sell side firms had both distributed and collaborating trading algorithms, which might be deemed the most advanced M2M 2.0 architecture.

The survey confirmed that firms from every region were expanding globally especially in emerging markets.

Multitenanted data centers were used mainly to reduce latency, or else to reduce costs and facilitate the use of e-services.

The survey also confirmed the strong shift to multi-asset trading with equities, equity derivatives, FX and commodities showing the highest demand, but with promising growth across the board.

Competition, however, is never ending.

One European prop trader recalls that five years ago he could run a simple arbitrage between the DAX and Eurostoxx, or Chicago, to make money.

“Then everyone jumped in,” he says, “So now we have to find more complex, global baskets to trade, driven by multi-factor models.”

He notes how quickly sentiment feeds through with statarb algos providing “structure for free.”

While intraday volatility is down, he notes, “when markets do move, we see sharper spikes when all the day-traders rush in to realign things.”

The only way to avoid toxic trades is to monitor lots of complex correlations constantly. Here too people are starting to use hardware solutions just to keep up with the tape.

As the pace of change accelerates while the scale and diversity of markets grow, agile traders must constantly adapt their models and technical strategies.

<<<>>>>

Coming Up in Part 4: Reconstructing the Value Chain.

Investment firms can no longer tolerate long lead time to plumb services into data centers.
<<<>>>>

About Equinix in the Global Financial Markets

Equinix, Inc. (NASDAQ: EQIX) provides global data center services that ensure the vitality of machine-to-machine e-commerce. Some 400+ buy- and sell-side firms have come to value the agility that our network-rich International Business Exchange™ (IBX®) data centers now provide by operating across 38 markets in 12 countries on five continents, including all of the world’s top financial centers. From New York to London or from São Paulo to Singapore, Platform Equinix contains the world’s most robust and mature financial ecosystem with 99.9999% uptime.
Our community includes 60+ innovative trading platforms like Chi-X, EBS, SIX Swiss Exchange, Deutsche Börse and Bloomberg, 150+ financial e-services providers and 680+ networks. These networks in turn link through to thousands of other firms or markets to form a global ecosystem offering fast scalability and customized service across today’s diverse and uncertain global financial system.

<<<>>>>

Learn more at: www.equinix.com/industries/financial-exchange/

<<<>>>>
<<<>>>>