The report is out, but doesn't capture the risks of high-frequency trading's spread

February 28, 2011 — 2:58 PM UTC by Dr. Harald Malmgren and Mark Stys

0 Comments

Elizabeth’s note: Harald Malmgren and Mark Stys wrote a column about systemic risks in the wake of the flash crash last spring last spring. Originally published in an academic journal, it was picked up by RIABiz and other mainstream business news publications. One of our readers called it brilliant. This latest column, below, contains their thoughts in reaction to the recommendations on regulatory changes to prevent another flash crash, which was issued by a joint SEC and the CFTC committee Feb. 18.

The “flash crash” on May 6 last year generated growing anxiety among investors that stock trading was somehow spiraling out of control. The SEC struggled to find the cause, or causes, but seemed unable to provide any explanation about what had happened or how it happened. As the days and weeks wore on, it became evident that computerized operations of high frequency trading firms (HFTs) had taken over a dominant role in daily market transactions and had become so fast that trades simply could not be easily tracked.

Eventually the SEC and CFTC collaborated and appointed a group of experts to study what had taken place and make recommendations. The experts focused solely on US trading in equities and equity derivatives. They reported on Feb. 18 and made 14 useful recommendations which may help to reduce the chances of a repetition of what happened on May 6. Here is a link to download the report: http://www.cftc.gov/ucm/groups/public/@aboutcftc/documents/file/jacreport_021811.pdf.

However, even 9 months after the flash crash, these experts still could not explain exactly what had happened on that particular day.

In spite of the good work to date by the SEC, the CFTC and their experts, national regulatory changes made or being considered may already be obsolete. Unfortunately, while US regulators remained focused on domestic trading, markets were continuing to evolve, as HFTs spread out across exchanges around the world, not only taking a growing share of world trading in equities but also spreading into trade in commodities, currencies, and a wide variety of derivatives.

Rapid advances in computerization and communications technologies are enabling trading to take place on many alternative platforms, not only domestically but internationally. HFT enterprises continue to pursue opportunities for worldwide expansion with the objective of interlinking trading without regard to national borders. Ultimately, it is conceivable that worldwide trading in stocks and other asset classes will take place 24/7 just like currency markets function now.

Stock exchange mergers

It should not have been surprising that the London Stock Exchange was seeking to buy the Toronto Stock Exchange or that this news was quickly followed by Germany’s Deutsche Bourse’s intention to acquire
NYSEEuronext, the company that owns the New York Stock Exchange. A wave of market consolidation and globalization in trading has just begun. The purpose of these mergers is twofold. First, they should over time provide increased access and efficiency in trading stocks globally. More important to the exchange owners, the consolidation of trading venues reduces costs for the exchanges and is expected to increase transactions to replace lost exchange revenue that has been dropping for the past three years along with falling volume.

Much of the volume on the exchanges in the US and around the world is now driven by high frequency, statistical arbitrage, and quantitative trading. High frequency trading was the main culprit for the May 6, 2010 ‘flash crash’ and continues to present significant systemic risk to stock markets, as discussed in our previous article published in International Economy Magazine, “The Marginalizing of the Individual Investor.” See: What RIAs need to know about systemic risks in the wake of the flash crash.

High-frequency trading’s spread

Our assessment of the impact of HFTs on the US stock market was translated into numerous languages and republished in various venues, which generated feedback from many new sources globally. While we initially focused on the US because of the dominance in volume and therefore risk, it was also becoming clear thatHFTwas spreading globally, and already becoming important to the exchanges in Europe and Asia, taking increasing market share wherever
HFT
can be connected with local exchanges.

Globally, regulatory changes, particularly Frank-Dodd and Basel III, are scaling back proprietary trading by banks, reducing capital availability from those sources, and increasing the role of alternative trading platforms. As banks have been forced to divest their proprietary trading capability, HFTs have gained in providing transaction volume, but it is questionable if they will ever provide the true two-sided liquidity we have historically expected from the major broker dealers who were market makers in stocks.

For the past decade trading has become more globalized by increased efforts of all investors, including institutional investors and hedge funds, to diversify their portfolios geographically and across virtually all tradeable assets. Computerization has enabled investors to invest in and adjust the allocations of their portfolios in virtually every nook and cranny in financial markets. The markets in stocks, commodities, bonds, and even various alternative investments have morphed into what is becoming a single global marketplace.

Recently, many asset classes have become highly correlated as capital spread to any and all relatively higher yielding assets. An article by William Coaker “Understanding the recent rise in correlations and how you can turn it to your advantage”, originally published in Investment Management Consultants Association’s Investments & Wealth Monitor and re-published by RIABiz.com, provides an excellent discussion on correlations and opportunities they can provide. Diversification has long been considered a means of mitigating risk. But, when returns on differing asset classes move together risk may become elevated, not diminished.

Hard assets

The last time most asset classes became closely correlated was in the first half of 2007. At that time, major central banks were intensely worried about what they described as the “correlation of 1.0.” They feared that this was historically unstable and suggested market breakdowns in the future. Only a few months later cracks began to appear in world financial markets. While we anticipate asset class correlations may return to more normal levels in the future, we think that the transition from present close correlations to a more normal distribution of risk with not be smooth. More likely, getting there will be a roller coaster ride.

In recent months the US stock market has been characterized by extraordinarily low volatility along with increasing correlation, not only among stocks, but between commodities, bonds, and other asset classes. At times, hard asset investments became inversely correlated with the value of the US dollar. When the US dollar fell, investors around the world sought hard assets as an alternative while at the same time many foreign investors sought US stocks because they looked cheaper and cheaper. When the dollar rebounded, the stock market fell. Today, it looks to us that these tight correlations may be starting to break down. We can already see recent tight spreads in the fixed income markets breaking up and spreads widening between stronger and weaker borrowers in the Eurozone and in the US domestic public and private bond markets, alongside shifting risk spreads between equities and bonds.

As more and more investors adopt a broader asset allocation we have seen more and more asset classes migrate towards a 1.0 or -1.0 correlation. These correlations could easily be reversed at the speed of light – by a computer trading algorithm working across multiple asset classes and exchanges, and then suddenly reverse again. These types of relationships have held and reversed before. For example, we can look at the relationships between the US dollar and gold. The correlation between the US dollar and gold has been inversely correlated from 2002 to the present, with some temporary major reversals in 2005 and late 2010.

The increased use of exchange traded funds (ETFs) by institutional investors and hedge funds has lead to new opportunities for rapid entry into and exit from a security that is expected to act as a proxy not just of an industry, sector or index, but for an asset class or sub class as a whole. It was ETF liquidations that were a major factor in the May 6th flash crash. The flexibility of ETFs combined with their acceptance by large traders has created an arbitrage opportunity between hard assets and the ETF. A good example can be seen in the arbitrage between investment in precious metals and their comparative ETFs.

While the growing role of high frequency trading has gathered attention in the US stock exchanges, the migration into other asset classes and other national markets causes deeper concern. An article by Jeff Cox of CNBC, “Ready for ‘Splash Crash’, the Ultimate Market Meltdown?” discussed fear of a potential multi asset crash based on such algorithmic trading and keyed on fears of typically highly leveraged commodity market participants. But, algorithmic trading has already spread beyond equities and commodities to currencies and fixed income as well. While the variance in exchange structures may make the programming of the computerized HFT, statistical arbitrage, and quantitative trades more challenging, and in some cases less time sensitive, it has also created integrated and computerized cross asset class trading that poses new challenges for risk management.

Breakneck speeds

In parallel with the next wave of market consolidation and rapid expansion of HFT and other forms of algorithmic trading, the Federal Reserve and other central banks have become major factors not only in currencies and interest rate markets, but through interventions in those markets they have also affected equities. European sovereign debt seems to be resurfacing, threatening banks, debt markets, and economic growth in Europe. The burgeoning debt management challenges of our federal, state and local governments are now raising questions about whether economic recovery might be stymied by public spending cuts and tax hikes. The recent wave of political and social unrest that has spread across North Africa and into the Middle East has added geopolitical risks that few investors had foreseen just a few months ago. All of these challenges could dramatically alter world capital flows and asset correlations.

The point is that this convergence of market exchanges, correlation, technologies, regulations and alterations in central bank decisions, and unforeseen geopolitical events could take us in directions we never imagined — and at breakneck speeds. Investment managers need to take into account the changing ways in which markets function. As computerized trading’s speed increases and enables cross asset class and cross border trading, one must consider the effects on correlation and diversification. There are continuously growing opportunities for diversification. But diversification is not without risks during period where correlations change rapidly, causing our assets and our stomachs to go up, down and in circles – much like riding a roller coaster.

Dr. Harald Malmgren is the global economic strategist for Bluemont Capital Advisors, LLC , the chief executive of Malmgren Global and currently the Chairman of the Cordell Hull Institute in Washington, DC. He is an internationally recognized expert on world trade and investment flows who has worked for four US Presidents.

Mark Stys, chief investment officer for Bluemont Capital Advisors, LLC, is a former bond trader with Fidelity Capital Markets and later became head of Fixed Income and International Strategy.



Share your thoughts and opinions with the author or other readers.

Submit your comments: