By Danny Watkins, founder and CEO, Market EarlyBird
Twitter is where the world’s news breaks today
Studies suggest that social media in general, and Twitter in particular, carry breaking news stories up to 10 minutes before they break on the wires. There are anecdotal accounts of corporate earnings reports appearing on Twitter four hours before an official announcement is published on the regulatory news service. The August 2017 terrorist attack in Barcelona is an example of a news event where a credible Twitter account (in this case LiveSquawk) reported the event nine minutes before CNBC.
This is most relevant to Financial Services, as news, whether national or international, about a company or an individual, is one of the major factors that drives the price of financial instruments. Indeed, in 2014, the SEC approved Twitter for the announcement of corporate actions in the USA.
The prices of infrequently traded instruments such as corporate bonds and other illiquid derivatives don’t vary much over time, so the impact of a single Tweet or announcement may not be significant. However sovereign bonds, commodities, currencies and equities are highly sensitive to announcements including weather, conflict, fiscal policy, politics, natural disasters, product news and corporate actions.
Prices are also affected by ‘fake news’. Whether a particular statement or assertion by an individual politician, journalist or commentator is complete, accurate and true – or not, it may still affect company stocks and currency exchange rates. Take for example the announcement on October 10th, 2017 by Dow Jones that Google was going to acquire Apple for $9bn and move into Apple’s headquarters. Although this was quickly denied and the announcement pulled almost immediately, Apple’s stock price spiked more than $2, returning to normal within a few minutes.
Twitter isn’t allowed on the trading floor – in theory
Clearly it’s becoming increasingly important that news carried on Twitter is available in real-time to traders, analysts and portfolio managers.
But Twitter is generally banned from trading floors, and often entire financial firms. The Twitter feed itself is probably blocked at the firewall, so even if someone manages to install a Twitter client or access Twitter in a browser, they won’t receive any data. Twitter is blocked because it not only has a broadcast mechanism, but it supports closed groups and provides an encrypted direct messaging system. It’s generally acknowledged that it’s not a good idea to allow outbound tweeting from anyone other than the approved marketing channel. But as we’ve seen in recent times with LIBOR and FX rates, if an exchange rate or index can be fixed, then someone will try to do it – and an encrypted, private messaging system built into an app that everyone has access to seems like the perfect mechanism to use. So it’s prudent for the firm to ensure services such as these are blocked.
Clearly removing Twitter’s capability to broadcast or message among groups or with other individuals external to the firm is a necessary move to ensure regulatory compliance and mitigate the risk of scandal, huge fines and reputational damage to the firm. But blocking it entirely denies those very traders whose job it is to act on breaking news access to the fastest and most diverse source of news on the planet.
What do the traders do? They complain, but in most instances they live without Twitter. Most of them are in the same boat, so there may be no disadvantage in the market.
Now imagine a scenario where the price of oil suddenly spikes without any obvious news or event that might have caused it. In almost every trading firm, someone on the commodities trading desk will turn on their mobile phone, connect to the Internet via an external mobile phone network, and access Twitter to see if they can find out what’s caused the market to suddenly move. The first time they may observe the firm’s rules and actually leave the trading room before turning on their phone.
But events like this are not uncommon. You can imagine that the third or fourth time, no-one will bother leaving the trading floor before firing up their mobile. If an instrument price changes inexplicably it has become accepted practice to access Twitter on the mobile. So the compliance ban on Twitter on the trading floor is circumvented, and the head of desk or the head of trading turns a blind eye. Compliance may not know. They may suspect. Or they themselves may be complicit. But it happens. We’ve had this confirmed at several of the banks we know.
Of course once accessing Twitter on a mobile phone on the trading floor is ‘accepted practice’ then its use becomes more common and more widespread. So when a trader wants to know what’s going on in the market it’s just as easy to Tweet their friend to ask. And it’s easy to send a Tweet asking them to buy some stocks at a particular price or to collude on a rate submitted to an index –opening the firm to the risk of multi-million dollar fines and having its reputation trashed on the front page of the Financial Times and the Wall Street Journal. Is this a risk worth running? The heads of trading and compliance are the people who really must make that call – their jobs, and potentially their liberty, are on the line.
Then there’s another issue. MiFID II extends transaction reporting and best execution requirements from equities to most other financial instruments in many situations. If traders are covertly using Twitter – even if only consuming rather than Tweeting – they will be getting information from it, and they may use that information to trade. What happens when that trade becomes the subject of an investigation? When it comes to Trade Reconstruction, there won’t be any information about what was seen on Twitter, so it may appear that the trader had no legitimate basis to make the trade, or even worse it may appear they had acted on insider information.
What are the options?
A curated feed
Several companies, including some of the major market data providers, offer a read-only Twitter client for finance – but none of them directly address the compliance issues. Instead, they get around them by providing a heavily curated feed where they select the Twitter accounts to be streamed. This is marginally better than having no Twitter access, but no financial firm can use this to generate alpha because every other user at every other firm sees the same Tweets at the same time.
An aggregated news feed
A number of companies offer aggregated news feeds that include relevant Tweets alongside other sources of news and then highlight major stories. The disadvantage of this type of service is that it generally doesn’t cover breaking news, as most often they wait until they have multiple sources before publishing the story – which will normally mean that even if the story broke first on Twitter, recipients of these aggregated feeds won’t see it promptly. Some of these suppliers even use employees to research, collate and verify stories before publishing them. While that may result in greater accuracy, it certainly adds a delay before the story is available to their clients. This may be adequate for analysts and investment managers, but it’s probably far too slow for trading applications. Besides the market reacts to the news, not just validated news.
One or two companies offer data mining for Twitter content – using AI to collect together data from Tweets about news stories that are relevant to a particular equity, commodity, bond or currency pair, and then alert their users to the story, and the likely impact on the market. These alerts can even be integrated into trading systems so that trades are executed automatically. However, this type of service is unlikely to publish analysis from individual commentators and doesn’t offer the kind of breaking news and in-depth analysis that many traders are looking for from Twitter.
Follow anyone with compliant, read-only Twitter
Most financial firms would prefer a fully-featured Twitter service delivered to their traders and analysts. After all, a trader or portfolio manager can only generate alpha if they follow a news source, company or analyst that’s informed and publishes relevant content but which isn’t followed by most of their competitors.
A head of desk at one major bank said, “We believe we have the best analysts in the world, and we want them to be able to choose who they follow on Twitter.” And, of course, the system has to provide comprehensive searching and filtering capabilities, so that users can find those relevant sources that provide reliable, credible breaking news and in-depth analysis.
But with the risks that Twitter presents, the only way this can be permitted is if there’s a compliance capability built into the system that records every Tweet seen by each user, recording the display-time for later review and potentially for incorporation into a Trade Reconstruction. Banks don’t have a surfeit of compliance people, so they are unable to allocate staff to continually review the Tweets that their users are viewing. This means the compliance system needs to be able to alert on suspicious behaviour by users and itself have an extensive searching and filtering capabilities.
As far as we know, there’s only one system available on the market that offers all of these capabilities – and our customers are taking advantage of it.