30 September 2014

67447

Dan Barnes - Information Corporation

47 | posts 60,335 | views 8 | comments

Future Finance

Finextra and Oracle have gathered together some of the industry's top thought leaders to discuss, debate and analyse the key trends and issues within transaction banking, regulations and retail banking. This group will focus on upcoming regulations, new service offerings and industry debate shaping the new financial services landscape with regular blog posts, video interviews, webcasts debates and surveys.

'Flash Boys' in a nutshell

16 April 2014  |  2937 views  |  5

The PR bluster around the new Michael Lewis book ‘Flash Boys’ is deliberately confusing, yet almost no-one seems to have read it. With the FBI investigating high-frequency trading (HFT), the issues that the book raises are clearly very serious. This is what you need to know.

1) HFT strategy is under the spotlight, specifically latency arbitrage. What is it?

Latency arbitrage takes advantage of the delay that other traders face in sending and receiving data and orders. An HFT firm using latency arbitrage gets data faster than other traders so it can tell what is being bought and sold at what price ahead of other traders. Secondly it trades faster than them - many thousands of times faster in some cases – so if you had to get to 30 different markets in order to buy 1,500,000 IBM shares, it would know you were trying to buy those shares, buy them ahead of you and raise the price by a fraction, say 0.5 cents. 

2) Is this a US only problem?

The US model requires trades to be routed across all exchanges to find the best price, under Regulation NMS (Reg NMS). As an order is routed from one venue to another in a consistent route, at a consistent speed across a given network, from any given broker, arbitrageurs have a known window of opportunity to intercept a trade. Any market in which many trading venues exist could support the same model. 

3) Why are brokers and infrastructure providers under attack?

They have all offered facilities that have supported latency arbitrageurs in their attempts to take money from the trades of fund managers. The book alleges that brokers have provided high-speed connections for their clients to move between markets faster than the brokers’ other clients. Trading venues have all offered ways for latency arbitrageurs (and other HFT firms) to get market data faster than other traders. The book notes that the cost of this high-speed data connection (US$10,000+ per month) requires an economic model that produces profit beyond the reach of a typical retail investor. The current round of investigations by US authorities is focussed on the similarity between sending data to one person faster than another, and giving data to one person before another. The latter is illegal. The New York Stock Exchange was fined US$5 million in 2012 on the basis that they are the same thing.

4) How do brokers win from HFT?

The book estimates that brokers take 15% of what HFT firms make, but also by routing order to lots of venues (brokers own 44 out of 45 alternative venues in the US, according to Ronan Ryan, chief strategy officer of exchange IEX) they avoid having to pay the main exchanges fees, capture order data themselves (which is very valuable) and do not lose control of liquidity. In the 2000s, brokers began to hold on to liquidity as much as possible, as the ongoing market fragmentation made execution harder to achieve.

5) The book is very much in favour of IEX; why?

IEX is an exchange based on the principal of confounding HFT strategies, simply by routing orders to other venues so that they all arrive at the same time, preventing any gaming. It has enjoyed enormous success since the book was published, according to chief strategy officer Ronan Ryan. However that says more about some buy-side firms than it does about IEX. It is far from being the only venue that is able to counter HFT.

6) Is the book biased in favour of Goldman Sachs?

Lewis writes that GS co-operated with him, but the firm is said to be every bit as involved as other brokers. While Lewis does note is that GS decides to stop engaging in HFT earlier than some brokers, because it lacked the technological flexibility to support it, he is critical of the prosecution of Sergey Aleynikov, a programmer who was wrongly convicted for stealing HFT system code from Goldman Sachs.

7) Should I read the book?

Yes! And don't read any reviews until you have.

TagsTrade executionRisk & regulation

Comments: (8)

Jorge Yui - BankArchitects - Geneva | 16 April, 2014, 21:10

 

Good Summary. Michel Lewis books on Finance have been often missunderstood. I like "Flash Boys" so far

Ketharaman Swaminathan - GTM360 Marketing Solutions - Pune | 17 April, 2014, 19:34

Michael Lewis calls the use of technology to acquire a competitive advantage - admittedly an unfair one - rigging. I didn't accept this contention when I read the reviews. I still don't accept it after reading the book.

Dan Barnes - Information Corporation - London | 22 April, 2014, 13:21

@ Mr Swaminathan - Is it the technology that rigs the market or the provision of information ahead of other traders? I think he is asserting that the latter is the case, and that releasing information at two speeds is the same as releasing information at different times.

Ketharaman Swaminathan - GTM360 Marketing Solutions - Pune | 22 April, 2014, 17:38

@Mr.Barnes: The way I understand it, there's no "provision of information ahead of other traders". Every trader gets the price info at BAT - situated closest to downtown Manhattan - at the same time. A subset of traders are able to use the same info earlier than the others ("competitive advantage") at the exchanges located farther away because they have invested in Spread Networks' line ("superior technology"). If that's rigging, so would be many other situations e.g. the winning relay team that used (say) superior shoes to reach the baton to the finishing line earlier than the other teams. 

A Finextra member | 23 April, 2014, 09:19

Or, alternatively, the relay team used steroids...

It seems the main problem is that I place an order, someone knows about it and is able to buy ahead of me and slightly raise the price on me. That is unfair. And is more akin to the rival relay team knowing my team's shoes beforehand and placing tacks inside them to slow me down...

Dan Barnes - Information Corporation - London | 23 April, 2014, 15:08

@ Mr Swaminathan - BATS' CEO was made, by the New York District Attorney, to recant his statement in this interview - http://www.zerohedge.com/news/2014-04-01/bats-ceo-shame-you-letting-everyone-it - saying that BATS' Direct Edge market fixes its prices using direct market feeds.

The reason was that latency arb firms get their data from direct feeds and Direct Edge uses the slower public SIPs feeds, which some might argue creates a rigged game, as latency arb guys can game people using the slower feed.

Ketharaman Swaminathan - GTM360 Marketing Solutions - Pune | 23 April, 2014, 16:18

@FinextraM: Even Michael Lewis admits that all methods used are legal, unlike steroids. No one is slowing anyone down - they're just running faster. This is not the only issue that illustrates how vested interests hold sway over fairness, at least in Wall Street.

@Mr.Barnes: My comments are restricted to the book but, nevertheless, TY for referring me to the interview. Supply of direct feeds is not restricted to any specific trading firm - just that, until IEX / Brad K pointed it out to them, the average investor didn’t know about such feeds or their use in this (legal) manner by HFT traders.

IMO, there is a real potential for rigging when it's legal for banks to trade on their own accounts under the (naive) stipulation that their prop shops shouldn’t access info regarding their clients' orders. Volcker Rule will change all that but that’s a story for another day. 

 

Dan Barnes - Information Corporation - London | 24 April, 2014, 09:11

@ Mr Swaminathan - You are correct of course regarding prop trading; the activity in broker dark pools, which is driven by HFT engines (albeit put to a different purpose) is no different from prop trading activity. There is no governance that determines when a broker must execute and all sorts of ways to show best execution (as there is no standard) after the fact. As a result they can execute or not in dark pools at an advantageous point in time to them, and claim it is market-making, when in fact they are optimising their own return. It was 'market-making' in dark pools that the SEC and CFTC say was withdrawn during the 2010 flash crash.

Comment on this story (membership required)
Log in to receive notifications when someone posts a comment

Latest posts from Dan

Google search: What’s my credit score?

01 July 2014  |  1870 views  |  0  |  Recommends 0 TagsRisk & regulationInnovationGroupFuture Finance

Trade finance creates a 10 billion dollar risk

11 June 2014  |  1682 views  |  0  |  Recommends 0 TagsRisk & regulationWholesale bankingGroupFuture Finance News Analysis

Bad as gold

29 May 2014  |  1652 views  |  0  |  Recommends 1 TagsRisk & regulationWholesale bankingGroupFuture Finance News Analysis

Is Bitcoin mo' money or no money?

29 May 2014  |  1257 views  |  0  |  Recommends 1 TagsVirtual currencyRisk & regulationGroupFuture Finance
name

Dan Barnes

job title

Writer

company name

Information Corporation

member since

2013

location

London

Summary profile See full profile »
Award-winning, freelance financial journalist. Specialist in many areas, including; sell-side exe...

Dan's expertise

What Dan reads
Dan writes about

Who is commenting on Dan's posts

Ketharaman Swaminathan
Dave Kershaw
Jorge Yui
Ponnusamy Selvaganapathy
Christopher Mc Carthy