Since the advent of big data, computer programmes for making sense of huge volumes of information have had an increasing grip on our day-to-day lives. Essentially sets of rules that tell computers how and when to carry out certain operations, these algorithms can precisely target demographics for advertising purposes, they can shape social interactions in lasting ways, they can even drive a car through busy city streets – and they are shaking up the way we do business. James Midgley reports on the profound shifts that increased deployment of ‘algos’ will usher forth in the years ahead.

 

Last month Belgian software company Analizo announced the launch of what it believes to be Europe’s first algorithm-as-a-service.  The firm’s software is intended to place quantitative investing technology in the hands of asset managers and private banks.

 

“Artificial intelligence and algorithmic investing are on the brink of revolutionising asset management,” noted CEO Wouter Verlinden.  “Computers are more adept at making rational and better investment decisions and are able to handle vast amounts of data that a biological brain is simply incapable of processing.”

 

The old adage that one should trade without emotion is now being fulfilled on a scale and in a manner that not so long ago would have been inconceivable.  But while Analizo’s algorithmic service may be the first of its kind, algorithms in trading have been changing the face of the financial world for some time now.

 

At its core, an algorithm is simply a set of instructions telling a computer how to deal with a problem.  An algorithm involves an unambiguous series of rules through which data may be processed, options reasoned and operations carried out – at high speed and with minimal intervention from human handlers.  ‘Algos’ are, in many ways, the building blocks that guide artificial intelligences.

 

Self-driven markets

 

Ever since computerisation of financial markets began in the 1970s, the competition for practical (read ‘lucrative’) algorithmic intervention has only heightened.  And as far back as 2007 a third of all European and US stock trades were driven by automatic programmes – themselves driven by algorithms.

 

Such automated systems have come to dominate commodity futures markets in terms of volume of transactions.  According to a US Commodity Futures Trading Commission report released earlier this year, they account for 63 per cent of crude oil futures, 54 per cent of precious metals and 49 per cent in grains and oilseeds.  Just two years ago, those figures were 54.3, 45.9 and 39 per cent respectively.

 

The systems range from the sophisticated tools of wealth management firms and hedge funds to the ‘bots’ built by enthusiasts on their home computers.  These algorithmic programmes are able to carry out trades at phenomenal speeds, with some even capable of a limited degree of market manipulation.  More often though, it is the predictable nature of the bots which can permit manipulation by other traders.

 

Last year Navinder Singh Sarao, dubbed ‘The Hound of Hounslow’ by the press, pled guilty to wire fraud and ‘spoofing’ – activities which allegedly caused the Dow and the S&P 500 to fall more than six per cent in just a matter of minutes back in 2010.  Automated trading systems make their decisions informed by close examination of a market’s order book; large-scale offers to buy or sell create the impression of buying and selling pressure in the market – and prompt bots to respond accordingly.  Those large-scale offers that urge the bots to action need not, finally, even be carried out.  This ‘spoofing’ means manipulators can buy when prices fall, then do it all again in reverse.  By the same token, algorithmic systems give an illusion of market depth, often withdrawing their buy and sell orders as soon as other traders join the fray.  

 

This kind of movement points to a key change in the markets – their increasingly insular, self-driven nature.  “The whole notion of fundamentals on any given day, for weeks at a time, months at a time, has completely gone out the window these days,” said Doug Duquette, an executive at Chicago-based Vertex Analytics, speaking to the Financial Times.  “You get momentum of algos playing upon algos upon algos, and it will just drive markets to extremes that don’t seem to correlate or line up with fundamentals on any given day or time period.”

 

Despite these and other teething troubles, the digital revolution has been a powerfully democratising force for trading – anyone with an internet connection can now take part.  Only time will tell whether advances in algorithms will create an entirely new kind of technological elitism; automated systems already carry out analysis and place orders with wholly inhuman speed.

 

The smart revolution

 

While trading presents an opportunity to see these algorithms and their effects in the wild, such software has come to permeate the structures of almost every business and industry one might care to mention.  “Algorithms are where the real value lies,” exclaimed Peter Sondergaard, Senior Vice President of Gartner and Global Head of Research, in the opening keynote at the 2015 Gartner Symposium/ITxpo in Orlando.  “Algorithms define action,” he stressed, hoping to usher in “algorithmic business”.

 

Those developments today see around a quarter of the world’s largest 300 businesses scurrying to fill a new position – that of the Digital Director, whose role is managing data and the algorithms that sift through it.  Likewise, algorithms underpin the advent of the Internet of Things (IoT) and Industry 4.0.  The International Data Corporation forecasts worldwide spending on IoT to reach US$800 billion this year, growing to US$1.4 trillion by 2021.

 

Some of the greatest disruption appears in industries where algorithms allow outsourcing of resources.  Uber, essentially the world’s largest taxi company, possesses no vehicles of its own.  Media giant Facebook generates no new content, while retailer Alibaba owns no inventory. Airbnb is the largest provider of accommodation but owns no real estate, and the largest communications companies (Skype, WhatsApp, Viber, Facebook Messenger) have no infrastructure to speak of.  This disruptive trend – dubbed  ‘Uberization’ with a mixture of playfulness and worry – is only set to accelerate, given the obvious streamlining benefits, plus many other features.  Algorithms power the matching processes between customer and outsourced product (drivers, say) as well as creating encryption, social connection, targeted advertising and much more besides.

  

Algorithms can revolutionise and streamline any industry that has the potential for wasted resources.  In the energy industry, companies such as UK’s PassivSystems is looking to leverage them to produce ‘smart’ systems for managing energy services.  “Our smart home energy asset management services enable us to reduce lifetime operating costs for consumers, making our services compelling for equipment manufacturers, financial institutions and energy companies,” says CEO Colin Calder.  The company’s cloud-based platform ‘PassivEnergy’ uses advanced weather-adjusted algorithms and machine learning to optimise performance of solar panels, home batteries and heating.  

 

As the size and complexity of utilities and other networks grow, their operation tends towards optimisation for average usage, with little room for adaptive improvisation.  Algorithms, if properly designed, can allow segments of networks to operate independently and cleverly thanks to machine learning.  Intelligent Traffic Systems are one such example, with Siemens leading the charge to upgrade transportation communication solutions across Africa.  The company’s products use complex algorithms to dramatically improve traffic flow, as well as including number plate recognition cameras and speed-over-distance calculations to maintain safety and security.

 

Dumb data, clever algorithms

 

“Data is inherently dumb,” stated Peter Sondergaard back in 2015.  “It doesn’t actually do anything unless you know how to use it; how to act with it.” 

 

Most people generate a staggering amount of data in a single day.  They book tickets for transport, do their shopping online, click links on social media, schedule dates on Tinder and other apps.  The scale of that data is only growing as more devices become ‘smart’ and as we incrementally approach IoT and Industry 4.0.

  

The world’s largest public transport app and data company Moovit recently announced a partnership with engineering consultancy Atkins aimed at accelerating the transition to smart transport systems.  The collaboration will combine infrastructure experience and big data; Moovit’s app gathers some 500 million data points every day.  “Moovit believes the future of urban mobility lies in the quality of data provided to cities and municipalities so they can plan and implement smart, efficient transit systems that meet the demands of ever-increasing populations,” stated Josh Wine, Moovit’s Chief Revenue Officer.

 

While ventures such as this undoubtedly improve efficiency, save time and money, and generally improve quality of life, there may be a darker side to algorithms.  Following the release of documents by Edward Snowden which showed that spy agencies had been collecting millions of webcam images (unbeknown to the people photographed), IBM Resilient CTO Bruce Schneier asked in the Guardian, “Is it really okay for a computer to monitor you online, and for that data collection and analysis only to count as a potential privacy invasion when a person sees it?”  The combination of surveillance data and sophisticated algorithms represents a danger to privacy which, eventually, may threaten to destroy the very concept of privacy, especially as we welcome an increasing number of IoT-enabled smart devices into our homes.

 

Algorithms are already affecting our individual lives to an astonishing degree.  Our love lives, for instance, may be decided by matching algorithms in dating applications (and subsequently our family lives, children and so on).  Dating company eHarmony is currently designing a twist on its scoring and matching systems – one that takes them to employment.  The programme produces a ‘Fit Score’ based on skills, culture and values, and personality – and to what extent these match those desirable in a particular role.  The drawback to processes such as these is that anything not accounted for by the algorithms is ignored; people are reduced to a handful of statistics, and history holds plenty of warnings for such an activity.    

 

Some are putting convenience to one side and asking some hard, ethical questions of algorithmic technologies.  Former Google Design Ethicist Tristan Harris has launched his ‘Time Well Spent’ initiative, claiming that, “Technology is hijacking our minds.” Harris and others are bringing attention to the increasingly sophisticated algorithmically-driven design choices which are keeping us hooked.  Facebook shows us a never-ending reel of media tailored to our tastes; Netflix keeps us watching with smart recommendations; Amazon persuades us to buy the things we never even knew we needed; and so on.  True consumer choice is being eroded more – and more cleverly – as time goes on.  Advertisers can target not only particular demographics (as they might by selecting a time to air a TV commercial) but, through social media, can reach precise subgroups with precise product formulations.

 

Regime change ahead

 

The algorithm-powered digital revolution is only just getting underway, with tremendous potential disruptions looming on the horizon.  Last year Google’s AlphaGo AI defeated Go grandmaster Lee Sedol 4-1; as AIs such as this gain mainstream traction, humans may begin to feel increasingly redundant.  

 

Back in 2013, research conducted by Oxford University and Deloitte estimated that over 35 per cent of jobs in the UK are at high risk thanks to automation.  The same research suggested that nearly half of all jobs in the US could be lost within two decades.  This includes some surprising job categories.  New York-based Associated Press already makes use of algorithms to generate simple financial stories – at a potential rate of 2,000 per minute.  IBM’s culinary AI Chef Watson continues to try to persuade users of the complementary flavours of a ‘chicken daiquiri cocktail’ or ‘jellyfish cassoulet’.  Back in 2014, Deep Knowledge Ventures, a Hong Kong VC firm, even went so far as to appoint an algorithm (called ‘VITAL’) to its board of directors.

 

In the world of finance, algorithms power the bête noire of blockchain-based currencies such as Bitcoin, as well as smart contract platforms like Ethereum.  There, complex algorithms provide the mechanisms (and cryptographic protection) by which decentralised ledgers are agreed upon and confirmed by independent nodes or ‘blocks’.  At the same time, the growing potential of accessible quantum computing means that entirely new classes of algorithms may soon be put into action – ones capable of solving exponential problems currently beyond the capabilities of classical computing (practically speaking).  Those same algorithms might also mean the end of cryptography as we know it – since quantum computing would ‘solve’ such security measures in seconds.  That includes the security of the blockchain, too.  

 

Finally, some are raising alarm bells over the potential for algorithms to not only reflect cultural values, but actually inculcate them as well.  Google’s driverless car is literally algorithm-driven – and it has to make tough decisions about what to hit if hitting something is inevitable.  “We do not really have a consensus on morals,” Dr Sandra Wachter, a researcher in data ethics at the Oxford Internet Institute, reminded Alphr.  Through exposure and permeation, experts are suggesting that algorithms stand a real chance of altering cultures; the same mechanism that persuades consumers to buy something through advertising might just as well, over time, instil a homogenised (and predominantly Western) point of view on global audiences.  

 

The revolution represented by the power of algorithms will arrive sooner than we might think – indeed, in many quarters it is already well underway.  Only time will tell the true extent of the changes to businesses and industries held in store – and, moreover, to the very fabric of our daily lives.