Share on Facebook: http://on.fb.me/1xTAE6b How much do you trust a computer algorithm with your money? Today's stock market trading is carried out by ultra-fast computers, scanning for market trends and making decisions in fractions of a second. Lately, these market trends include real world news stories. But in the future, could stock market algorithms create FAKE news stories and social media posts, just to gain an advantage? How do YOU think social networks could affect the stock market in the future? Let us know in the comments below! -------------------------------------------------------- Subscribe to Fw:Thinking: http://www.youtube.com/subscription_center?add_user=fwthinking For the audio podcast, blog and more, visit the Fw:Thinking website: http://www.fwthinking.com Fw:Thinking on Twitter: http://www.twitter.com/fwthinking Jonathan Stickland on Twitter: http://www.twitter.com/jonstrickland Fw:Thinking on Facebook: http://www.facebook.com/FWThinking01 Fw:Thinking on Google+: https://plus.google.com/u/0/108500616405453822675/ [TRANSCRIPT]: Planning for your retirement? The robot has your money. Remember all those movies back in the eighties with all the guys on the floor of the New York Stock Exchange making crazy hand signals like an over caffeinated third base coach as they traded stock? Well today these little Gordon Gekkos have been replaced by computer based algorithms that can make trades lightning fast. They identify market trends and pounce on them. And increasingly those market trends include real-world news, which these algorithms can jump on faster than as humanly possible. And they're serious business. I mean, just check out their names. You've got Sniper, Stealth, Guerrilla...that's guerrilla like Rambo, not like Planet of the Apes... The name of the game in high-frequency trading is speed. And these algorithms are constantly in a dog fight to identify market trends before the others. They're so fast at identifying trends that sometimes they act a little too quickly. That's what happened back in May, 2010 during what is now known as the Flash Crash - where the Dow lost one thousand points in just two-and-a-half seconds! To gain an edge, often these algorithms will try to fake each other out. They'll fake a handoff, or they'll create a smoke screen, or they'll do some...other sports metaphor thingy. At any rate, one of the common concepts they'll use is called 'quote stuffing.' This is when they create a fake bid and they cause other algorithms to look the other way for a couple of nanoseconds, and then they'll go and pounce on an emerging trend before anyone else can. But what if in the future, instead of creating a smoke screen with fake bids, they started to create it through fake news. Now we've already seen that people have been able to inflate stock prices by creating fake news stories to drive the prices up. All in an effort just to unload some stock they don't want anymore. Now we've also seen that computer software can write certain types of stories faster and better than humans can. Stories like financial news coverage, or sports scores. So, imagine this scenario. Say an algorithm is scanning every major news outlet in the world plus Twitter posts. If it detects frequent occurrences of words like major earthquake and the name of a particular city, it automatically starts unloading shares of companies with headquarters in that city. Now, knowing this, what's there to stop people from programming algorithms to put out fake news that looks totally real? A fake report of a top-notch CEO resigning unexpectedly could get retweeted a million times. That could cause the stock price to plunge until people confirmed the news to be false. Now this all may sound pretty scary. But really, it's just another signal to noise ratio problem. It's just now, we have to depend upon computers to help us separate the signal from the noise, even when some of those computers are the ones creating all the bunk in the first place. Algorithms aren't gonna go anywhere, so we might as well cozy up to the good ones and hope that they're smarter than the bad ones.

• published: 05 Feb 2014
• views: 55968