The Future (6 page)

Overall, the technology-enhanced integration of the global economy is lifting the relative economic strength of developing and emerging countries. This year (2013) the GDP of this group of countries (as measured by their purchasing power) will surpass the combined
GDP of advanced economies for the first time in the modern era. The potential incapacity of these countries to maintain political and social stability and to deal with governance and corruption challenges may yet interrupt this trend. But the technological drivers of their ascent are powerful and are likely to prevail in consolidating and increasing a dramatic and truly fundamental change
in the balance of global economic power. Already, in the aftermath of the Great Recession, it is the emerging economies that have become the principal engines of global growth. As a group they are
growing much faster than the developed countries.
Some analysts doubt the sustainability of these growth rates. But whatever their rate of growth, it is only a matter of time before these economies experience the same hemorrhaging of jobs to intelligent machines that is well under way in the West.

M
OST PEOPLE AND
political leaders in advanced industrial countries still attribute the disappearance of middle-income jobs simply to offshoring, without focusing on the underlying cause: the emergent reality of Earth Inc., and the deep interconnection between outsourcing and robosourcing. This misdiagnosis has led in turn to divisive debates over proposals to cut wages, impose trade restrictions, drastically change the social compact between old and young and rich and poor, and cut taxes on
wealthy investors to encourage them to build more factories in the West.

These distracting and almost pointless arguments over labor policies are echoed in similarly misguided debates over the impact of national policies on financial flows in the age of Earth Inc. The nature and volume of capital movements in the ever more tightly interconnected global economy are being transformed by supercomputers and sophisticated software algorithms that now handle the vast majority of financial transactions with a destructive emphasis on extremely short-term horizons. One consequence of this change is a new level of volatility and contagion in the global economy
as a whole
. Major market disruptions are occurring with greater frequency and are reverberating more widely throughout the world.

The sudden disruption in credit markets that began in 2008, and the global recession it triggered,
resulted in the loss of 27 million jobs worldwide. When the period of weak recovery began one year later, global output started to increase again but the number of jobs restored—particularly in industrial countries—lagged far behind. Many economists attributed the jobless nature of the recovery to a new eagerness by employers to introduce new technology instead of hiring back more people.

Exotic, computer-driven “manufactured financial products” like the ones that led to the Great Recession now represent capital flows with a
notional value twenty-three times larger than the entire global GDP. These so-called derivatives are now traded every day in volumes forty times larger than all of the
daily trades in all of the world’s stock markets put together. Indeed, even when the larger market in bonds is added to the market in stocks, the estimated value of derivatives is now
thirteen times larger than the combined value of every stock and every bond on Earth.

The popular image of trading floors is still one where people yell at one another while making hand signals, but human beings have a much smaller role in the flows of capital in global markets now that they are dominated by high-speed, high-frequency trades made by supercomputers. In the United States, high-speed, high-frequency trading represented more than 60 percent of all trades in 2009. By 2012, in Europe as well as the U.S.,
it represented more than 60 percent of all trades. Indeed, stock exchanges now compete with one another with propositions like one from the London Exchange, which recently advertised its
ability to complete a transaction in 124 microseconds (millionths of a second). More advanced algorithms will soon make trades in nanoseconds (billionths of a second), which
according to some experts will further increase the risks of market disruptions.

An academic expert in automated trading at the University of Bristol, John Cartlidge, said recently that the result of the increasing speed of trades “is that we now live in a world dominated by a global
financial market of which we have virtually no sound theoretical understanding.” In the first week of October 2012, a single “
mystery algorithm” accounted for 10 percent of the bandwidth allowed for trading on the U.S. stock market—and 4 percent of all traffic in stock quotes. Experts suspected the motivation was to slow down data speeds in order to enhance the advantage for the high-speed computer trader.

Advantages in the speed of information flow have played an important role in markets for at least 200 years, since the Rothschild bank used carrier pigeons to get early word of Napoleon’s defeat at Waterloo,
enabling them to make a fortune by shorting French bonds. Fifty years later, an American investor chartered faster sailboats to gain earlier knowledge of key battles in the U.S. Civil War and
make a similar fortune by shorting bonds from the Confederacy. But the emphasis on speed has now reached absurd levels. Trading firms routinely place their
supercomputers adjacent to their trading floors—because even at the speed of light, the amount of time it takes for the information to cross the street from another building would confer a competitive disadvantage.

A few years ago, a business friend in Silicon Valley told me about an opportunity to invest in an unusual project to build a straight-as-an-arrow fiber optic cable from Chicago’s trading center in the inner Loop to the New York Stock Exchange’s trading center in Mahwah, New Jersey. The inherent value of the project—since completed—came from its promise to shave three milliseconds off the time it took
to transmit information over the 825 miles (from 16.3 to 13.3 milliseconds). Traders at the other end of the cable gain such a significant advantage with a three-millisecond head start over their competitors that access to this new cable is being sold at premium prices. An even newer microwave system with even faster data speeds (though less reliability in bad weather) is now being built along the same route.

The melting of the North Polar Ice Cap has led to the start of a new project to connect markets in Tokyo and New York with faster financial information flows via a fiber cable along the bottom of the Arctic Ocean. Three other projects have commenced to link Japan and Europe under the Arctic, and a new transatlantic cable being built for another $300 million is expected to increase the speed of data flows between New York and London by 5.2 milliseconds.

The spending of $300 million to save a few milliseconds in the flow of information is but one tiny example of how much of the wealth that used to be allocated to inherently productive activities is now diverted instead toward
what many economists call the financialization of the economy. The share of the American economy now devoted to the financial sector has doubled from around 4 percent in
1980 to more than 8 percent at present.

Part of this startling increase reflected the large investments that financed the information technology explosion up to April of 2000, and part represented the rapid growth in mortgages that accompanied the buildup of the housing bubble up to 2008. Yet even after the bursting of the dot-com bubble, and later, the housing bubble, the financial services sector continued to gain a larger share of GDP. The driving force behind this historic shift has been the application of powerful supercomputers and algorithms to the manufacture of exotic financial derivatives—and the capitulation by government in the face of lobbying by the financial
services industry for the relaxation of regulatory standards that used to impede the marketing of such instruments.

An estimated 82 percent of derivatives are exotic instruments based on interest rates, almost 11 percent are based on foreign exchange contracts, and roughly
6 percent are based on credit derivatives. Less than one percent are based on the
value of actual commodities. But the overall flows are so incredibly large that, to pick one example, the value of oil derivatives traded on a typical day is an astonishing
fourteen times the value of all the actual barrels of oil traded on that same day.

In theory, these high-volume, high-frequency computer-driven flows are justified by the assertion that they improve the liquidity and efficiency of markets. Many economists and bankers hold the view that the large flows of capital represented by derivatives actually add to the stability of markets and do not increase systemic risk, in part
because banks hold collateral equal to a large percentage of what they are trading.

Others, however, point out that this view is based on the now obsolete assumption that more liquidity is always better—an assumption that is in turn based on two theories about markets that are part of the long discredited “standard model”: that markets tend toward equilibrium (they don’t), and that “perfect information” is
implicitly reflected in the collective behavior found in the market (it isn’t). Nobel Prize–winning economist
Joseph Stiglitz says that high-speed trading produces only “fake liquidity.”

Unlike trades in the stock and bond markets, derivatives trades are almost completely and totally unregulated. That adds to the risk of increased volatility in markets, especially when the daily volume of electronic transfers of capital now exceeds the
combined total of all of the reserves in the central banks of all advanced countries. In practice, the progressive displacement of human decision making from the process and the explosion in the trading of artificial financial instruments in volumes that dwarf the transactions of real value in the global economy has contributed to the increased frequency of major dislocations in the role of capital as a reliable and efficient factor of production. Some of the artificial instruments now being traded in high volumes are difficult to distinguish from gambling.

There are two factors that explain the underlying reason why the management of global capital flows by supercomputers in microsecond intervals creates new systemic risks in markets: extreme complexity and tight coupling. And they work in combination. First, the complexity of the system sometimes produces large and troublesome anomalies caused by a form of “algorithmic harmonics” (essentially computer programs reacting to
one another’s simultaneous operations rather than underlying market realities). This complexity means that the problems thus introduced into the system’s operation can be extremely difficult for any actual human being to understand without taking a considerable amount of time to get to the bottom of what has gone wrong. Second, the tight coupling of multiple supercomputers ensures that no one will have the luxury of time to figure out what’s gone wrong, much less the time to address it.

One example: on May 6, 2010, the value of the New York Stock Exchange fell a thousand points and rebounded almost as much—
all in the time span of sixteen minutes—for no apparent reason. There was no market-sensitive news of the kind normally associated with a sharp drop of that magnitude in such a short timespan. As
The New York Times
reported the following day, “Accenture fell more than 90 percent to a penny,
P&G plunged to $39.37 from more than $60 within minutes.” The
Times
quoted one trader as saying “
it was almost like ‘the Twilight Zone.’ ”

It required five months of intensive work by specialists to understand what had happened to cause this so-called Flash Crash, which they eventually found was the result of the complex interaction between automatic trading algorithms used by a large number of supercomputers in a way that created, in effect, an
algorithmic echo chamber that caused prices to suddenly crash.

When one of these experts, Joseph Stiglitz, recommended remedies to prevent the recurrence of the Flash Crash, he suggested a new rule to require that
offers to buy or sell must remain open for
one second
. The captains of finance in charge of the financial companies with the most at stake in the current pattern of business, however, reacted with horror to this proposal, claiming that the one-second requirement
would bring the global economy to its knees. The proposal was then rejected.

The global market crisis of 2008 was primarily caused by a particular kind of derivatives: securitized subprime mortgages hedged with an
exotic form of insurance that turned out to be illusory. Supercomputers sliced and diced the subprime mortgages into derivatives that were so complex that no human being could possibly understand them. And once again, the robosourcing of these exotic financial instruments aided and abetted the marketing of those same products to buyers throughout the global economy.

When the actual quality and real value of the mortgages in question were belatedly examined, they were suddenly repriced on a mass basis—triggering the credit crisis and bursting the housing bubble in the United States. The fact that they had been linked to a complex web of other computer-driven financial transactions (collateralized debt obligations, or CDOs) led to the credit crisis, a massive disruption in the availability of capital as a basic factor of production in the global economy—essentially, a global run on banks. This led, in turn, to the Great Recession, the effects of which we are still struggling to escape.

Incidentally, after the manufacturing of these derivatives gained momentum and scale, virtually the only remaining role for human beings in the process resulted from the legal requirement for a signature on each underlying mortgage by someone with the responsibility for reviewing the integrity of each mortgage that had been sliced, diced, securitized, and rubber-stamped with a AAA rating by
corrupted and captive ratings agencies, then sold around the world.

As the subsequent lawsuits revealed, this requirement for signatures by actual human beings could not keep up with the speed of the supercomputers—so low-wage employees were hired to forge the signatures of loan officers a hundred times a minute, without the slightest attention to the substance and meaning of the documents they were signing—
a practice that’s been popularly labeled “robosigning.” Though no robots were involved, the very term illustrates the intertwining of robosourcing and outsourcing.