Many financial institutions use algorithms to make high-risk decisions involving large sums of money — perhaps the highest vote of confidence that any business can give. So what can we learn exactly?
copyright by www.forbes.com
Today, there is a lot of debate about the use of algorithms within business settings. Some feel that their use makes it easier for businesses to systematically shed employees; others feel that businesses put a lot of trust into what they consider to be a black box, unknowable and unaccountable.
Yet all of this discussion seems to ignore the fact that one of the first industries to begin strategically using algorithms, the financial industry, is also one that is notoriously risk-averse when it comes to adopting new technologies. In fact, many financial institutions use algorithms to make high-risk decisions involving large sums of money — perhaps the highest vote of confidence that any business can give.
As the CEO of a company that uses deep learning to create customized marketing algorithms for players in all industries, including finance, I have seen firsthand the impact that an algorithm can have when helping businesses reach their goals and expand their growth. It is no exaggeration to say that algorithms have been a game-changer for the financial industry. They’ve not only made it easier for big institutions to make money, but they’ve also made it possible for individual players to make a name for themselves. Their successes are a testament to the power of machine learning, and they are an example for other industries looking to reinvigorate themselves.
How Algorithms Have Transformed the Financial Industry
The first quantitative hedge funds appeared on the scene in the 1980s, and their influence has only grown since then. One of the most common ways for banks to use algorithms involves setting the parameters for a trade to occur. Traders can create an algorithm that directs the system to purchase a stock when it reaches a certain price or sell if it falls by a certain percentage. While these algorithms are not always powered by machine learning, they are relatively common within the trading world, and they can even be used by those looking to invest in stocks on their own.
More sophisticated versions of these algorithms can incorporate machine learning to take into account any factors that might affect a stock’s price, from world events to changing trends. For instance, last year, JP Morgan launched what it calls its “Deep Neural Network for Algo Execution.” It’s a neural network that combines its existing foreign exchange algorithms into one highly optimized bundle.
While financial institutions use machine learning for its data analysis capabilities to identify potential opportunities in the market, they still often leave it up to humans to choose which opportunities to ultimately pursue. But as the Economist points out, there is also another way that machine learning can be used: to design new investment strategies from scratch, without having to take into account human preferences or prejudices. It used to be that humans would use algorithms to test an existing hypothesis; now, as one investor says, “We start with the data and look for a hypothesis.”
In the financial markets as they exist today, a significant percentage of assets are either being traded by computers without any human input or managed by them. As reported by the Economist, Deutsche Bank estimates that 80% of cash-equity trades and 90% of equity-futures trades are carried out by algorithms. This is astonishing if you think about the sheer volume of trades that are carried out each day. In other words, you would be hard-pressed to find a financial institution that does not use algorithms in some way that could directly impact how much money the business makes on behalf of itself and its clients.[…]