Lessons learned from Flash Boys

Following the recent dismissal of the so called ‘Flash Boys Case’ made against exchanges for favouring high frequency traders, Annabel Smith explores how Michael Lewis’ original novel has shaped the market... and why it drove asset managers to take legal action.

The idea that high frequency traders were using speed to take advantage of ordinary investors was first brought to the world’s attention with the release of Michael Lewis’ 2014 novel, ‘Flash Boys: Cracking the Money Code.’ A story that follows the epiphany of former RBC electronic traders Brad Katsuyama and Ronan Ryan as they come to realise that the electronification of the market had opened the door for a new ‘predator’ so to speak that used speed to nip in ahead of slower and larger traditional institutions. Its release emphasised the importance of speed in trading and increased the amount of attention paid to how it can be utilised to take advantage of the unequipped ordinary investor.

Beginning with the launch of Spread Networks – a company that at a cost of $300 million laid an ultra-low latency fibre optic cable connecting Chicago and New Jersey, in a bid to sell speed to HFTs and Wall Street banks – Flash Boys tells the story of latency arbitrage and market manipulation. Through high-speed connections – either through co-locations or access to enhanced proprietary data feeds – HFTs and proprietary traders were accused of taking advantage of oblivious traditional institutional and retail investors, using information gathered on the public markets at a faster speed or through access to banks’ private dark pools. According to the book, in 2011 roughly 30% of all stock market trades were occurring off-market, with most of these taking place in dark pools.

It was on Reg NMS in the US that Katsuyama and Ryan based their claims that the market was systemically rigged. Reg National Market System (NMS) was brought in by the Securities and Exchanges Commission (SEC) in the US in 2007, inspired by front-running charges against participants from 2004. It requires the broker to find the best price for the investor in the National Best Bid and Offer (NBBO) meaning the broker must first buy however much of that stock is available at the best price. To gather the picture of the NBBO, exchanges created Securities Information Processors (SIPs) – consolidated data feeds including all bid/ask quotes from every trading venue, however, the IEX founders claimed the system offered a loophole to those firms that were willing to build enhanced SIPs nearer to the exchange’s systems to harvest the information faster.

The aftermath

As the first window into the secret underhand operations of the market, the publishing of the novel has had a significant and lasting impact on how some participants choose to trade and has encouraged regulators to pay greater attention to various practices including payment for order flow in a bid to protect ordinary investors.

“Culturally the novel being published was a milestone at a large US asset manager – especially as we had just reached an inflection point in the industry where electronic trading was being universally being adopted as an efficiency tool and there was an acceptance that this was ‘good technology’ and would allow the industry to scale even more rapidly,” says one buy-sider.

“The ‘Flash Crash’ a few years earlier [a rapid stock market crash that lasted for approximately 30 minutes in 2010] introduced better risk controls, better systems and better oversight of the electronic functions which had been adopted in a limited manner by the larger asset managers. Akin to the recent run on $LUNA [a crash in the price of the digital asset at the start of June that caused $500 billion in losses in the broader crypto market], the event shed light on possible industry weaknesses such as the formation of negative feedback loops due to the fragile interconnectivity of fragmented markets and venues that had appeared as a consequence of the introduction of Reg NMS in the mid 2000s,” they continue.

“I was surprised at how little people knew about what was happening with their orders and how they interacted with proprietary trading activity operated by major banks and the book did at least force people to ask more questions and understand more about what was happening in a complex marketplace.”

Several court cases were immediately spawned based on allegations broadcast within the book: including investigations from the SEC into the dark pools of both Barclays and Credit Suisse. Ultimately, however, the blame fell at the doorstep of the exchanges for allowing the alleged behaviour to take place. In 2014, just a month after the book hit the shelves, five lawsuits were launched that eventually boiled down into a single action against the accused trading venues. 

The case – now dubbed the Flash Boys Case – put forward by institutional investors claimed that the exchanges (including Nasdaq, the New York Stock Exchange and BATS global Markets – now part of Cboe) had created a preferential trading environment for high frequency traders (HFTs) that put other investors at a disadvantage. Included in these favourable trading conditions were co-location services that allowed firms to place their servers in close proximity to the exchange’s servers, superior data feeds allowing participants to create a better and faster picture of the market, and the National Best Bid and Offer (NBBO) and complex order types, all offered for a fee.

These latency-focused solutions were designed to improve the speed at which firms could access valuable information and move in and out of orders on venues – which enabled them to get ahead of other investors and profit from the tiny gap in time when they were privy to information that the rest of the market was not. 

The case was thrown out in 2015 on the grounds that the exchanges’ self-regulatory status protected them from private damages lawsuits. However, it reared its ugly head again in 2017 when it was found they in fact had no such immunity. The seven exchanges moved to have it thrown out again in 2019, but this was denied. It was only in March 2022, eight years later, that the Federal Court concluded that the institutional investors could not prove they had suffered harm at the hands of the exchange’s actions, adding that the expert witness testimonial of Dave Lauer was not a reliable methodology. 

“The problem was that the institutional investors really didn’t have any proof that harm was occurring. While there have been some fines here and there I wouldn’t say there is any systemic proof that the markets are ‘rigged’. Electronification was supposed to put the buy-side on an even keel with the sell-side and enable people to trade from anywhere. Theoretically, you didn’t need to be standing on the floor of the New York Stock Exchange to have as much of an advantage as anybody. The problem with that theory is the laws of physics and the issue of speed,” says head of market structure research for Bloomberg Intelligence, Larry Tabb.

“The job of an exchange is price discovery and so they’re less worried about volume and more worried about the tightest possible price that they can provide. They’re also interested in transaction flow because exchanges get paid by the SEC in terms of market data in two ways, one being the number of shares they execute and the other being the aggressiveness of the bids and offers on their market. They’re going to get a larger rebate if the price is tighter versus whether the price is lower or wider and so while I don’t think there’s a collusion to make the markets worse for institutional investors, there are incentives in the exchange infrastructure to try to provide a fast, tight, efficient market.”

Co-location and proprietary products are offered to all institutions and are approved by the SEC, explains one HFT. “The buy-side connects to exchanges via brokers and one of the first questions they usually ask them is do your algos connect to proprietary data feeds,” they say. “Everyone has access to the same thing.”

According to some voices in the market, the court case relating to the publishing of Flash Boys was not driven by feelings of aggrievement from institutional investors but instead at least in part by a set of proactive law firms looking to encourage participants to take part in presumably lucrative lawsuits.

Earlier this month, retail broker Charles Schwab was asked to pay $187 million to settle charges with the SEC that it had misled its robo-advisor clients with regards to fees. The SEC found that from 2015-2018 Schwab did not reveal to clients that the service was directing funds “in a manner that their own internal analyses showed would be less profitable for their clients under most market conditions”.

“If you looked at Bloomberg, there were law firms popping up asking for elite plaintiffs to represent in similar cases to what had happened with Schwab. It’s the same thing for the Flash Boys case, law firms looking for plaintiffs. It’s a uniquely American and uniquely annoying problem,” says the HFT.

Widening the field

While legal action has been somewhat unsuccessful, the publishing of the novel has been effective in driving evolution in the industry. Bringing the events of Flash Boys to the wider market’s attention increased demand for transparency from investors and expanded the range of trading venues available in the market. As Lewis so eloquently put it, the problem was not about removing the hyenas and the vultures from the food chain, it was about giving them fewer chances to kill. 

Among the new venues to come to market is US-based Investors Exchange (IEX), co-founded by Flash Boys stars Katsuyama and Ryan in 2013. In light of what they thought they had discovered, the pair wanted to foster a marketplace that would allow high frequency and traditional institutions to co-exist more harmoniously. Alongside various speed bump mechanisms, the exchange has introduced what is referred to as a D-Limit order or “discretionary limit order”. Limit orders allow an investor to buy or sell a security at a specified price or better. IEX’s order predicts when the market is about to move and then moves limit orders at the bottom of the orderbook – usually belonging to the buy-side – out of the way so they are not run over. 

“Things like the D-limit order are not only looking at how to slow trading down, but also to see if venues can protect limit orders that are already existing on the exchange,” says one industry insider. “To their credit the general concept for most exchanges is to maximise liquidity because if you don’t trade, you don’t get paid, you don’t get to get tape revenue and you don’t get all of the market data benefits. In this instance, they’re actually putting an order type in that says when the flag goes up we’re going to move your order away from trading.”

However, the D-limit order is not without its own controversies. It has come under criticism from some who claim it only offers IEX members a delay and therefore puts non-IEX members at a disadvantage to the rest of the market. 

“If IEX was on a fair footing with everybody else who could take a look at the order book and move then it wouldn’t be that controversial,” adds Tabb. “The controversy lies in that moving those D-limit orders does not go over the speed bump for IEX clients, so they have a 350-microsecond advantage over everybody else.”

IEX’s market share hit 3% for the first time in August 2019, although has been somewhat overshadowed by independent Members Exchange (MEMX) which launched into the market in October 2020 with the goal of increasing competition, improving transparency and reducing costs. MEMX entered the market with 0.1% market share only to reach 3% in April earlier this year. IEX declined to comment. 

“That gets back to this issue about speed and market making. MEMX is fast and cheap. They also have a high rebate,” says Tabb. “The ability to be fast and have a high rebate means that MEMX is paying people to trade there and so because they pay you to trade there you can quote more aggressively as well because they’re fast you can quote more aggressively.”

Also launched in 2013 was the Aquis Exchange which, following the same philosophy of IEX, does not allow HFTs to cross the spread in order to prevent them using certain trading strategies on the exchange. Instead, the exchange only allows them to use a specific post-only order type.

“We’re not anti-HFT but there are certain strategies that we believe are done by proprietary trading firms that change the profile of the liquidity pool. If you don’t have HFT and prop traders on your market it’s going to take longer to get your order filled and a lot of people want immediacy,” says Aquis chief executive, Alasdair Haynes. “If someone is always the winner you don’t get competition. We believe there should be several markets out there. Ones that have everybody and others that are specific for the longer-term investor. They can co-habit.”

The events that followed the novel are also most likely the stimulus for European regulators to take more prescriptive action against broker crossing networks (BCNs) scrapped as part of Mifid II in 2018 and to encourage more volumes towards agency matching venues and periodic auctions. 

Payment for order flow

Another practice on which the publishing of Flash Boys shone a spotlight was the controversial process of paying for order flow (PFOF), whereby retail brokers channel their customers’ flow to banks and market makers in return for a fee. Not only do firms have the opportunity to channel this flow to where it might suit them – the US exchange rebate system means some venues charge pay you for providing or taking liquidity – but it also gives these firms access to the information within each order that could then be traded against in the wider market. 

The retail market in the US has swollen in recent years on the back of the pandemic, and the events surrounding the GameStop saga and the meme stock explosion, and the increased attention has brought PFOF into the crosshairs of the regulators. In a recent speech, SEC chair Gary Gensler warned that PFOF presented conflicts of interest and distorted routing decisions. He also suggested it may incentivise brokers to encourage the gamification of the markets as seen with GameStop in a bid to increase trading volumes.

“Exchanges give rebates to traders. High-volume traders benefit more from these arrangements, and retail investors don’t directly benefit from those rebates,” said Gensler. “Just as payment for order flow presents a conflict of interest in the routing of marketable retail orders, exchange rebates may present a similar conflict in the routing of customer limit orders.”

According to 606 reports gathered by the SEC, Citadel Securities forked out the most in payment for order flow in 2020 and 2021 at $2.6 billion, followed by Susquehanna (G1X global execution brokers) which spent a $1.5 billion and Virtu which spent $654 million in the same period.