Artificial Intelligence is Taking Over Investment Banking
In the hallowed hallways of Morgan Stanley’s office in Canary Wharf, men in crisp tailored suits pass each other, briskly chatting away on their smartphones, while attractive young women in black pencil skirts and long legs scamper through the halls with take-away salads from Chop’d, dressing on the side. Walking these very halls has been the aspiration of many, the full-time MBA candidate who was looking for a career change and “wanted to get into investment banking”. Just what is investment banking, you might ask? Well, the money-making part of the operation is called “the front office”. Let’s just say that if you meet a 9 at All Bar One in Canary Wharf and she asks you if you’re in a front office role, just say yes. The front office is where all the money is made, where all the movies are based on, where all the
cocaine caffeine-fueled late nights happen. The “front office” in your typical investment bank falls into one of the three following distinct areas:
- Sales and Trading
- Corporate Finance
As we type this, artificial intelligence and business process automation are being used to deconstruct each of these areas, brick by brick, in order to cut costs and make things more efficient. For the Managing Directors who control the investment bank, everything is about hitting their numbers so they can get that big fat bonus. Here’s how artificial intelligence is making that happen.
Artificial Intelligence in Sales and Trading
The most alpha role in the front office is that of the trader. They are the big swinging dicks of an investment bank, and they can just about get away with murder and not lose their jobs. As it turns out though, it was the nerds that took them down. Just take a look at an excerpt from this article by MIT Technology Review that came out a few weeks ago:
At its height back in 2000, the U.S. cash equities trading desk at Goldman Sachs’s New York headquarters employed 600 traders, buying and selling stock on the orders of the investment bank’s large clients. Today there are just two equity traders left.
Let’s let that sink in for a bit. Those dudes (yes, they were probably all dudes) were taking home an average salary of $500K a year. Now they’ve been forced to “move on to more value added activities”. The days of Liar’s Poker have all but died. Those 600 traders have now been replaced by technology, and 200 computer engineers (or “nerds” as they were once called by the traders whose jobs they replaced). Across Goldman Sachs, over 30% of their staff are now computer engineers. And it’s not just Goldman. Last week, JP Morgan hired a “Global Head of Machine Learning” from Microsoft. The guy is one the world’s foremost NLP specialists, and doesn’t have any background in finance whatsoever.
It’s important to note here that traders in an investment bank who trade on the banks’ dollar used to be known as “proprietary traders” or ” prop traders”. They already left the investment banks a while ago to start or join hedge funds because of tightening regulations. We’ve written all about hedge funds that are using artificial intelligence for stock trading and talked about the spectacular returns of those that do. We’re not talking about traders who generate alpha, we’re talking about traders who service clients of the investment bank and help them conduct transactions. Sure, maybe you still have some sales guys left on the desk who aren’t based out of Mumbai and who are dealing with clients, but how long before they’re replaced by a Mya or an Amelia?
Artificial Intelligence in Corporate Finance
The biggest chunk of work in this part of the bank is mergers and acquisitions or M&A along with IPOs. While you’ll always have the “rainmakers” at the Managing Director level who facilitate the deals, all the grunt work to prepare the deal books is performed by well-paid analysts. We’ve already seen how artificial intelligence is starting to attack accounting. A lot of the work performed by analysts in preparation for a corporate event is mundane data gathering and filling out worksheets. Then this output is passed on to the Vice Presidents who whip it into presentation decks that can be used to pitch the deal to various stakeholders. Do you know how much time could be saved if you had a tool that could answer any question you had about a publicly traded company?
A company we wrote about before called AlphaSense, using artificial intelligence technology called natural language processing (NLP) to comb data sources such as conference call transcripts, investor relations presentations, and even your own content which you can upload. You can search across millions of documents with a few clicks and get your answer in about 3 seconds for 35,000+ global companies.
What you can now do is take all those spreadsheets that your analysts spend 16-hour days filling out and then plug each field into the Alphasense API. Change the name of the company to which ever publicly traded company you want to acquire. Click refresh. No more overpaid developed markets analysts. No more “John in Mumbai”. Sure, it doesn’t work for private companies but we’re pretty sure you can link it to any accounting system out there that private companies use.
The other key role that Corporate Finance teams perform is that of IPO offerings. In the same MIT Technology Review article mentioned earlier, the CFO and previous CIO of Goldman Sachs, Marty Chavez, made the following statement at a Harvard symposium last month:
Goldman has already mapped 146 distinct steps taken in any initial public offering of stock, and many are “begging to be automated”
Just how long will it be until the entire IPO process is automated such that you fill out a lengthy online application for an IPO which is then taken all the way through the 142 steps until the day your shares start trading using technologies like blockchain to make sure everything is audited? We’re already seeing crowdfunding for equities being introduced. How long will it be until the “stock market” just encompasses all these secondary markets that are being created by fintech companies such that your entry into the entire interlinked and highly regulated system begins with the issuance of a Dun and Bradstreet number?
UPDATE 2/27/2017: JPMorgan Software Does in Seconds What Took Lawyers 360,000 Hours The program, called COIN, for Contract Intelligence, does the mind-numbing job of interpreting commercial-loan agreements that, until the project went online in June, consumed 360,000 hours of work each year by lawyers and loan officers. The software reviews documents in seconds, is less error-prone and never asks for vacation.
Artificial Intelligence in Research
A very basic interview question in finance is whether or not you know the difference between “active funds” and “passive funds”. The difference is that passive funds simply track a stock benchmark like the S&P500. That’s not that difficult to do and fees are therefore cheap for “passive” products. “Active” products are where a very well paid asset manager tries to beat the benchmark. Get this. 86% of the time they can’t beat the benchmark and they still get paid that salary. This is why all the money is now flowing from active investment products to passive investment products:
When the portfolio managers who built all those isht active products made portfolio allocation decisions, do you know where they were getting all those genius ideas from? In many cases, these ideas were coming from the research teams at investment banks who provide “paid research”. Nowadays, basic financial commentary is being automated with ease.
Just recently, Narrative Science announced a partnership with Factset in which they will distribute AI generated financial commentary. And you know what? You can be sure that AI algorithm wasn’t in the bathroom last night doing blow with the CEO of the company it wrote writing about. The analysts putting out all that research have a conflict of interest because many times the companies they write about are clients of the investment bank. The technology behind this is called natural language generation and its a technology that is so compelling, Gartner forecasts that by next year 20% of all content will be written by software.
It’s not just the “front office” getting overhauled by automation and AI, it’s also the back office as well. Companies like R3 are using blockchain technology to make all those jobs they just finished moving to Mumbai and the Philippines become all but obsolete. The key takeaway is this. If you’re thinking about “getting into investment banking” in the front office or the back office, you better bring some mad coding skills to the table.
Pure-play disruptive tech stocks are not only hard to find, but investing in them is risky business. That's why we created “The Nanalyze Disruptive Tech Portfolio Report,” which lists 20 disruptive tech stocks we love so much we’ve invested in them ourselves. Find out which tech stocks we love, like, and avoid in this special report, now available for all Nanalyze Premium annual subscribers.