Since I know (thanks to Bloomberg) that the country where I live (Belgium) is a very interesting place for High-Frequency Trading microwave towers (Belgium is in the middle of the London-Frankfurt network), I am working on a map of the different paths using radio towers to carry data between London and Frankfurt. I’ll publish a detailed map next week. In the meantime, here is one path I have reconstructed (here in purple), used by HF traders but owned by a technology firm leasing the network to them. There is one tower which is not included in the network. Is it what the industry calls "tower squatting"?

My map of the microwave towers used by HFT players between London and Frankfurt is almost finished. I have found a lot of them. Some places are a kind a "meet-me room" for those players. See for example Swingate in the UK :

I’ll publish the full paths and the full names early September, in an epic article. By then, it’s time for holidays!

Article protégé par mot de passe. Il faut visiter le site et entrer le mot de passe pour lire la suite.

Article protégé par mot de passe. Il faut visiter le site et entrer le mot de passe pour lire la suite.

Article protégé par mot de passe. Il faut visiter le site et entrer le mot de passe pour lire la suite.

In April 2013 I visited the Trade Tech Europe event in London – it was my first immersion in the technical world of high-frequency trading (HFT). I collected some copies of Best Execution and because I am interested by the space-time issues of the market microstructure – for example: how orders are routed between the different lit exchanges/dark pools/crossing networks located in different places – I have found this paper quite absorbing.

Then I have learnt one thing : exchanges are not time synchronized. That means the matching engine of  NYSE is not time synchronized with the BZX matching engine, the matching engine of EDGX is not time synchronized with the Nasdaq matching engine, and so on, and that is problematic: without synchronization – in the HFT world microseconds count – it’s impossible to exactly know how and when an order was filled (here or there), which kind of order type was used, etc. So both academics and investors don’t have precise data. Data is missing.

In this previous post I talked about a proposition McKay Brothers co-CEO Stéphane Tyč made in Paris, in December 2013. In short, the proposal was:

Exchanges produce timestamps at the most important times of the matching cycle. They also identify every matched trade with a unique number. The regulators could define standards for this information and impose the end of day publication of the raw data from the exchanges. Every matched trade should have attached :
- the timestamp of the order that triggered the trade at the time of entry in the matching engine.
- the timestamp of the publication of trade in the publicly available data stream
- the anonymous trade id number.
All timestamps should be accurate to 1µs of the atomic time, which is readily available with PTP (Precision Time Protocol). This way, all the issues of synchronicity could be studied very precisely.
This should be made available in the raw market format at the end of each day. A simple code to read this data should be provided under an open source license.
Based on this information, any client of a broker could receive the collection of trade id numbers that constituted his execution and could go and find for himself if the execution was to his standards.
Open data, open formats, open source code, those would help further the debate [my emphasizing] and enable research to shed light on this very challenging problem.

The amazing fact was after I published this post, I received a mail from the Autorité des Marchés Financiers (AMF, the French SEC) saying that they were working on the synchronization issues and they would have been happy to meet me in order to have my opinion on the topic. It was a little fun to see that a regulator was interested by my point of view, but I didn’t have a lot of things to say, even if I met the AMF people involved in these issues and told them to speak with Stéphane Tyč. They did and with the help of Stéphane they refined the proposition (about timestamps synchronization) they have sent to the European regulator ESMA. I was only a go-between, and I’m still a go-between here.

Since I know Stéphane quite well now (he helped me to understand the world of microwave networks used by some HFT players), he kindly asked me to read and comment a white paper he was working on about best execution, the time synchronization of exchanges, etc. I can’t define myself as a true specialist but I think what Stéphane Tyč proposes is a very interesting way to level the playing field. Here is the white paper:

INTRO

As soon as the summer sun returns to Belgium, there is a natural human migration from Francophone Wallonia to the west coast sea in the Flemish part of the country. Here, à la mer du Nord, people relax on white sand beaches and ride the famous quadracycles called cuistax. The Belgian west coast is the (maritime) backyard for many, and so I could not refuse accompanying the two local members of my family for a week à la mer du Nord. For once, I decided not to take dozens of books with me, just one debated paper entitled “Can Brokers Have It All?”, but the sun was so hot that I decided to drink mojitos on the beach.

The only tour I did was in Veurne, a small typical Flemish city where I tasted some local beers. A week later I was back in Brussels, and on July 16, 2014, Bloomberg released this story about a prop trading firm from Chicago purchasing an old tower for €5.000.000 in…Veurne. What a coincidence! I’ve researched and written about high-frequency trading (HFT) since 2012, and am quite aware of the microwave networks used by some fast players, but I did not know that as of January 2013 Jump Trading owned a tall tower in Belgium – in my backyard. Belgium being in the middle of the problem, geographically speaking, I decided to investigate the presence of microwave networks owned or used by the big HFT players here in Europe for my holiday homework. If Jump has a tower, competitors are all around – it’s just a matter of where . My question about the towers was “Can I Have Them All?” and thus my investigation, perhaps obsession, began. I discovered too many things to discuss in a single post, so I’ve split my story into three parts. This first part we’ll call “Mapping the HFT Microwave Networks” – even if “How I Became an Expert in Potato Fields” would be as accurate.

FROM CHIGAGO TO BELGIUM, A BRIEF HISTORY OF TOWERS

The first part of my book, 6, was mostly about the rise of the HFT machines in the pre- and post-REG NMS era. Charting the surge of Electronic Communication Networks such as Island, I realized the history of Chicago markets in conjunction with the ascent of these technologies was more interesting than Wall Street SOES bandit stories.Hence the second and last part of my book, 5, was focused on Chicago, particularly the transitions from the old human trading pits to the computerized ecosystem now (co-)located in huge data centers..

The history of exchanges in the Windy City is a true case study for anyone interested by the computerization of commodities and financial markets. I wrote about the arrival of the telegraph in Chicago; the first telegraphic “tweet” arrived in Chicago just three months before the first organized exchange, the Chicago Board of Trade (CBOT), was created in April of 1848. I even found the phrase “flash orders” in some 19th century Chicago newspapers, used to describe the practice of traders using fast private telegraphic lines to buy commodities they knew would rise in price in the next few seconds so that they could resell at a higher price what they had bought seconds before – sound familiar?

There are many other interesting predecessors of modern practices to be found. For instance, long before the existence of microwave towers like the one in Veurne, the CBOT aimed itself to tower. Its first location was the tallest house in the city, and when the exchange was obliged to erect its own building in 1885, it was both the tallest property in Illinois and the first building in Chicago to run on electricity. Frank Norris describes it in his novel The Pit: “The lighted office buildings, the murk of rain, the haze of light in the heavens, and raised against it the pile of the Board of Trade Building, black, grave, monolithic, crouching on its foundations, like a monstrous sphinx with blind eyes, silent, grave,—crouching there without a sound, without sign of life under the night and the drifting veil of rain”.

In the 1920s, members of the board voted to erect a new building to become “the cathedral of capitalism”, and initiated discussions with architects, traders and some of the major telegraph companies in order to find the best way to improve the social space of the pits. The traders weren’t happy with the excess noise, and telegraph companies like Western Union and RCA were lobbying to install a multitude of fast lines. When the 1930 CBOT building was erected on Jackson Boulevard near Lasalle Street,it had more than 4300 kilometers of telephone lines and 240,000 kilometers of telegraphic lines – first step of the rise of the machines. It towered over Chicago until 1965, a faceless statue of goddess of agriculture Ceres adorning the top.

The heart of exchange was the pit, and it correspondingly saw advances in technology. The first pit was patented by Ruben S. Jennings in 1878, and soon the CBOT build different pits without paying fees to Jennings. The inventor tried and failed to sue CBOT, and his patent was invalidated by the judges. Jennings designed the pit so that a trader could see and hear another trader under the best possible conditions. Pits had different steps for that reason; a trader sitting on the tallest steps had a better “skyline”, and an advantage in seeing or hearing a counterpart quickly, and trade faster than the traders in the center/bottom of the pit. It was all about speed.

Height, then, played an important part in facilitating that speed. A trader’s physical height became an advantage, which is part of the reason some traders were former basketball or football players –“taller traders were easier to see”.In the 1990s, some traders wore high heels in the pits to trade faster, and inevitably experienced injuries due to lack of balance. This prompted the Chicago Mercantile Exchange (CME) to impose a ruling making the maximum size of platform heels two inches in November 2000.Standardization of heel height was a way to level the playing field in terms of speed on the floor. Recently, the last European open-outcry pit the London Metal Exchange, or Ring, fined traders for breaking the rule to remain seated while dealing: “Dealers that stand create an unfair advantage and might obstruct the view of other dealers and LME pricing committee members,” said a spokesperson for the UK exchange.

Computers tower above humans in speed no matter how high a person’s heels, so by the 1980s and 1990s, computerization of exchanges was rampant. A story documented by Donald MacKenzie in “Mechanizing the Merc: The Chicago Mercantile Exchange and the Rise of High-Frequency Trading”describes what I believe to be a crucial moment in the transition between human and electronic transactions. After the CME launched E-Mini contracts in 1997, the exchange built a semicircular platform above the pits where the first users of Globex terminals were installed. “So was born ‘the bigs and the littles’,” MacKenzie writes. “The arbitrage between the pit- traded S&P 500 future and the E-Mini, a fifth its size. Pairs of traders would collaborate, one in the pit and one sitting above it at a Globex terminal, communicating by hand signals or radio headsets.” The old hand signals used by human traders since the 1930s – known as “arbs”, they were faster than yelling – now pointed to the sky, as if intended to draw the attention of Ceres. The new god was not a statue; it was a faceless computer, and Ceres Trading was now a HFT firm.

An ex-trader CME S&P 500 pit describes the physical platform to MacKenzie: “When you went to the floor, did you see the almost towers, kind of towering by the S&P pit? Almost gets to the ceiling, and you get a bunch of guys sitting there with terminals? That’s the guys that trade the E-Minis … some of these guys are doing very, very, very well, extremely well.”Note that the ex-trader used the word “tower” here, as many HFT firms started on this platform. Getco’s founders Daniel Tierney and Stephen Schuler were there, but rereading MacKenzie’s article, I realized two other traders were working with Globex terminals on the tower: Paul Gurinas and Bill Disomma. The firm they founded in 1999 was none other than Jump Trading. Even before the advent of microwave networks, it was all about towers.

The semicircular structure was probably demolished when the CBOT and the CME finally merged. On June 12, 2007, the two oldest markets of Chicago became one of the most powerful exchange in the world, now located in the  largest US financial data center, in Aurora. One month after the merger, on July 9, 2007, Reg NMS was fully implemented and the face of market microstructure changed radically. One hundred and fifty-nine years after the CBOT was created,among the most profitable trading companies are HFT firms and some “are doing very, very, very well, extremely well”. According to another Bloomberg article, Jump Trading are one of these; they are certainly doing well enough to buy a tower in Belgium for €5.000.000. The Houtem tower attracted them not just for its location, but for it’s height as one of the ten tallest buildings in Belgium. Just as it was in the old pits: the taller you are, the faster you trade.

MICROWAVE & HFT 101

Simply state, traders are algorithms, and exchanges are data centers.Most, if not all traders co-locate at the same distance from an exchange’s matching engine, which matches buyers and sellers. Matching may occur in a few dozen microseconds. In this new HFT ecosystem, information – between two exchanges for example – needs to travel very fast. This has led some HFT competitors to use faster technologies than the usual optic fibers, the most recent of which is microwaves. Microwaves are actually an old technology with what would appear to be significant drawbacks: a dislike for rain and fog, and limited bandwidth of 10% that of optic fibers which necessitates the rewrite of existing algorithms. However, microwaves offer both easier start up and a more direct route. There’s no need to tunnel through mountains as optic fiber operator Spread Networks did to cross the Allegheny Mountains, documented by Donald MacKenzie in this paper in his book  and later Michael Lewis’ Flash Boys.

Just install dishes on towers and find the shortest path between point A and point B, period. Market orders travel faster through the air than they do through optic fibers beneath the ground. The following map displays microwave networks between New Jersey data centers and Chicago (if you want to know more about the New Jersey-Chicago networks, read this paper):

Some New Jersey-Chicago microwave networks. From McKay Brothers/Quincy Data Trading Show 2014 presentation

 But have in mind that Earth is a sphere. You can go from a point A (a New Jersey exchange) to a point B (a Chicago exchange) but you have to take into account the curvature of the Earth. Even with customized dishes, you have to deal with nature (that’s why as an anthropologist I spent a lot of time to work on these networks: it’s all about dominating nature.) If you want to go far, you need tall towers to bypass the curvature of the Earth. But tall towers are rare – so they are expensive. What’s more, if there is only one tree (or one leaf) between two points/dishes, the network is dead. Paths need to be free of obstacles.

Let’s recap. Microwaves need dishes. Dishes need towers. Towers need to be as close as possible to the straight line between the two points/exchanges. The various competitors have to find the towers they need to build the shortest path between the two points. The fact is: sometimes you can’t install your dishes because someone else is already there (some towers can’t handle a lot of dishes and towers are squatted by other operators – mobile operators, national radio operators, etc. So you need to be smart when you think a HFT microwave network – competitors may have squatted a tower you needed. HFT is hard.

THE MAP

Early July, after the Bloomberg article about Jump and the Houtem tower was published, I realized that my home is very close (four kilometers) to the straight line between two european majors points: London and Frankfurt (I am a transparent guy, so I put my home in the map, you’ll be able to see where I live). In a recent presentation in Chicago, McKay Brothers co-CEO Stéphane Tyč showed this map of the microwave networks owned or used by some HFT players in the UK:

Microwave networks in the UK. @McKay Brothers

Problem is: you don’t see anything, and there is no name. So I tried to have all the towers, to find names. Finding the towers was easy; drawing the paths is complex because most of the HFT players try different possibilities – Jump had a path between Belgium and England before purchasing the Houtem tower, but now they can go directly to Ramsgate, and Basildon. My map shows not only some real paths, but also the various attempts to beat space:

I’ll detail in the Part II how I designed the map (and my crazy visit to the Jump tower in Houtem) but in short: 99% of what I have found is public. Just try Big Brother, aka Google (just click on the pins, you’ll find links with all the public files). I also talked with informants working in the industry, but they have been fair, helping me to understand this very, very, very small world. This map is a only network of possibilities, shortest paths being the Holy Grail – besides, some operators move from a tower to another… In this small world all the operators know where competitors are. There is no secret. I just found where some firms are, or could be, or will be.

You can’t really watch this world with photographies. So I made what all the HFT players do: a Google Earth file, a map with all the towers I could find. Here is the map. Open it, and zoom… zoom… zoom… zoom… and enjoy. Captions will be in the Part II – Part I is just an introduction. In the meantime, here is the tower purchased by Jump in my backyard:

This is insane: more than 28,000 visitors came to read Part I of “HFT in my backyard”. Some of them are inevitably bots, but my logs reveal the majority to be human readers from bank desks, technology providers and so forth. Who would have imagined a simple map would garner such interest? Microwaves have been used for HFT since 2010 after all. Since so many readers want to learn about these networks, allow me now to continue my story. In order to give you a clear picture of what to find on the map and how to locate all these towers, let’s return to Houtem.

FROM FRANKFURT TO LONDON… AND CHICAGO

In the US, the New York Stock Exchange (NYSE) is currently located in Mahwah, in New Jersey. For the Lenape people who originally lived there, the word mahwah meant “Place Where Paths Meet” – a perfect description of a present day exchange. At NYSE, co-located traders meet in a data center, and microwave paths converge on the roof. In Europe, there are two main data centers in England near London: one to the east in Basildon, the NYSE facility housing the Euronext/Liffe exchanges plus Goldman Sachs’s dark pool Sigma-X; and one to the west in Slough, the LD4 data center containing the BATS exchanges. Frankfurt, Germany is home to the Equinix FR2 data center which hosts Deutsche Börse and Eurex.

There are two types of competitors in the very small world of microwaves. First are HFT prop trading firms: Chicago-based Jump Trading (aka World Class Wireless), Dutch companies Optiver and Flow Traders (aka Global Connect) and  DRW (aka Vigilant Global). Some of these firms sell part of their bandwidth to other customers. Second are actual providers: McKay Brothers, Custom Connect and, more recently in EU, NeXXCom or Latent Networks. Their customers are banks, hedge funds, even other HFT firms. Some like McKay are only interested in the Frankfurt-London path, while others such as Optiver, Flow Traders, Jump and Vigilant also join the  Atlantic Crossing 1 cable landing station in Whitesand Band, Cornwall, England, to allow data to cross over the Atlantic and go straight to Chicago using US microwave networks. Let’s start with Jump Trading.

AN EPIC AUCTION SALE

Contrary to the legend, the Houtem tower featured by Bloomberg was neither used nor owned by NATO; rather the tower was built by and for the US Army in the 1970s. The tower (and two others located in Belgium, one of which is also now used by Jump) was sold to Belgium in 2006. “Closure of these sites will result in an estimated annual savings of over $84,000 based on a comparison of the current annual operations and maintenance costs to an annual replacement commercial communications cost,” wrote the US Department of Defense. In 2012, the Belgium Ministry of Finance used the same money-saving rationale to auction the tower.

The sale took place in Veurne on December 18, 2012. While the auctioneer is now retired and my request for his phone number was refused by the Services Patrimoniaux office, I did speak with someone who attended the auction. Only eight people attended: the auctioneer, his assistant, a government official, and five potential buyers – two or three Americans including a Jump representative, one large Belgian law firm probably acting on behalf of a competitor and a single observer. Prior, the prospective buyers had all visited the Houtem tower in situ so they entered the auction knowing the tower was in poor shape and required renovations my informant estimated would cost the winner $1,000,000. The starting price was €255,000 and the Belgium government would have been very happy with €400,000. The first tick-size was a €5,000 increment, but after half an hour the price was at €700,000 and the tick-size was increased to €10,000 accordingly.

The auctioneer didn’t know who the buyers really were or why they’d have such interest in a lousy old tower; this unexpectedly high price left him perspiring nervously and calling for a break. Collecting himself in the bathroom, he exhaled quietly, bewildered, “What the hell…” The auction resumed, bids climbing to €1,000,000, €1,500,000, €2,000,000. At this point the auctioneer asked for another pause. Finally, after three and a half hours, the tower was sold for €5,000,000. One attendee quickly left, as his car was parked in front of a police station and the meter was long expired. Another losing bidder approached Jump to ask, “Can’t we arrange?” meaning their company would purchase bandwidth or rent some dish space on Houtem tower. I wish I had been a fly on the wall to confirm my suspicions about which competitor attempted to make arrangements with Jump at the auction deep in Flanders that day.

A CRAZY VISIT TO THE HOUTEM TOWER

Since I missed the tower over my early July holiday, I decided to go back to Houtem with my family at the beginning of September. I spoke with different people from the microwaves industry ranging from Jump’s competitors to technology providers, and they explained how I might conduct my field work. I learned to be discreet, parking my car far from a tower, taking binoculars and a camera, checking the GPS of my mobile and then walking – one is obliged to walk far, as most of the towers are in potato fields. Several expeditions to different towers gave me the practice to perfect my technique before I was confident to approach the object of my visit. At Houtem, I parked near a farm and walked along the Chemin des Limites road near the Belgium-France border. Nearing the site, I noted a van and a workman in the tower’s basement. “Not good news,” I told myself, thinking an intelligence officer should always remain unseen. I began to surreptitiously photograph the Houtem tower:

When the worker came to burn things in the field, I moved closer:

There was a “No Trespassing” sign on the railing so of course I entered only to discover an unexpected windfall – the workman left all the doors open! My 63 year old mum stood guard, eyeing down the worker as I visited each room, taking pictures of the equipment until there, right in front of me, was a big red button functioning as the emergency stop. It was amazing to realize I could have cut the Jump microwaves network just by pushing it. But I didn’t want to bother Jump nor Perseus’s customers, and besides, microwave networks have fiber optics backup so any sabotage would have been useless. A message to the jump lawyers: don’t sue me. I know I was on “private property” so I won’t publish the photographs I took, save this one:

It’s fascinating to see the extremely small basement of a 243 meter tower. These kind of towers are engineering marvels. Here the diameter of the foundation is only 25 centimeters and the structure stands erect only thanks to 48 guy wires. Okay, I have to publish one more photo, as the viewpoint is amazing:

I was actually more interested in trying to jump and climb the tower than cutting the power, but the worker came back so I disappeared quietly.

The tower has been fully refurbished from top to bottom with all new guy wires, but my goal was to check the dishes. When the Bloomberg story came out, there were two dishes at the top of the tower:

© Bloomberg

My photograph shows three dishes, so Jump added one between early July and early September:

This brings me to the question that so flummoxed the auctioneer: why? Why would Jump pay so much for the Houtem tower? The first reason is height; HFT players need tall towers for speed, and the Houtem tower is the fourth tallest tower in Belgium. The list below expands Wikipedia‘s list of the highest structures in Belgium to include my HFT firm entries:

Since the Houtem tower was owned by the US army between 1974 and 2006, I dug up this document published by the US Department of Commerce in 1979 and titled “Signal Level Distributions and Fade Event Analyses for a 5 GHz Microwave Link Across the English Channel”. The article contains these two charts:

(Incidentally, Jump uses a frequency between 7.448400 and 7.48401 GHz for this 87.8 kilometer path..)

All the HFT competitors have to cross the English Channel from France or Belgium to England. They mainly go to Swingate in the north of Dover, where there are two old towers housing Optiver, McKay and Vigilant, or to Hougham in the south of Dover where Latent and Custom Connect have dishes:

The Swingate towers are located on the famed white cliffs of Dover, at about 141 meters above sea level. However, most HFT competitors link to Swingate from Dunkirk, a city which is at sea level elevation:

In Dunkirk, dishes have been installed on the Tour du Reuze:

For the logic, let’s go back to amend the US Department of Commerce chart:

Without getting too deep into the physics, microwaves don’t travel in straight lines. Waves curve looking for the earth, which is why height becomes so important. If you have a tall tower, the waves your dishes transmit are less attracted by the earth. Jump can bypass problems because the Houtem tower is of such great height. It doesn’t necessarily mean they’re faster (because HFT is nothing if not prohibitively complex) but it does give Jump a distinct advantage.

I don’t know if Jump was in Dunkirk before purchasing the tower, or if they used another tower in Calais (Mollien, 50°57’23.31"N | 1°52’19.86"E) to link to Swingate. I do know that their current microwave route in Belgium is divided in two parts from Houtem: the Houtem-Swingate path enabling a way to Slough west of London, and the Houtem-Ramsgate path allowing access to Basildon east of London.

THE WAVRE TOWER

Jump uses other towers in Belgium as well. There is a distinct lack of transparency in Belgium radio regulators, but I have found dossier after dossier after dossier after dossier allowing me to piece together some paths:

In Flobecq, another old US army tower, they co-locate with Flow Traders and Optiver:

In Hannut, they co-locate only with Optiver. Comparing this two documents reveals Jump added a third dish in Hannut between February 2012 and February 2013, perhaps after they bought the Houtem tower. In Liège, Jump is alone on a tower but Vigilant and McKay are within 1.07 kilometers:

I only visited one other tower housing Jump, in Wavre near Brussels. The Wavre tower is the third tallest structure in Belgium. Once again, I parked my car from the tower and walked through the fields. Some techs were working at the top of the 250 meter tower, but they looked tiny from my vantage point on the ground. Unlike the Houtem tower, which needs guy wires to remain erect, the Wavre tower is a beautiful “standing structure”:

Wavre is also used by practitioners of the extreme sport “base jumping”. Here you can see two crazy people jumping from the tower – hope they didn’t damage the Jump dishes during their fall:

My reason for trekking to the Wavre tower was that it made big news in Belgium when it was sabotaged last spring. On the early morning of May 24, 2014, a few people started a fire which caused severe damage to the tower – this video coverage in French actually shows Jump dishes in a shot. The tower is property of the Belgium national TV/radio operator RTBF but also supports many dishes used by mobile operators. The fire received such media attention because May 24 was an important election day in Belgium, and the police wondered whether the fire was related to the election. RTBF quickly erected a temporary tower to install mobile, TV and radio dishes, but the Jump dishes stayed on the tall tower because the 50 meter temporary one wasn’t high enough.

During my visit to the Wavre tower, there were still some burnt parts on the ground:

It is certain that all the cables burned, rendering the Jump network ineffective for some time. I wonder if police interrogated any of Jump’s competitors, or anyone who knew a trading firm was on the Wavre tower. As of early September all the cables were knew so the Jump network is likely in good shape. Following are the three big dishes owned by Jump, two pointing to Hannut and the other facing toward Flobecq or Oostvleteren:

 HEIGHT AND SPEEED

I would love to know if Jump, Optiver and, above all, Vigilant really use these paths between Newhaven and the Isle of Wight:

The Jump competitor who asked, “Can we arrange?” at the auction for the Houtem tower may not have been asking to install dishes on the tower; rather his proposal might have been to share bandwidth. Despite the small world of financial microwave networks being as silent as the potato fields in which stand their towers, I have discovered that Jump is sharing its network with Perseus and another data provider I’m not at liberty to name. The facts are as follows:

• A microwave link can split bandwidth up into 10Mbps increments.
• The microwave link has an aggregate of 100Mbps. Every 64 byte packet send over these links takes 6 microseconds and change to serialize.
• Each participant sends 64 byte frames of trade data to the service.
• At 10Mbps clients can only send 1 packet every 60 or so microseconds.

Now, imagine every customer in a service sends data at about the same time, a somewhat likely scenario due to the likelihood of one customer sending data nano or microseconds before the next customer and so on. In this thought experiment, Customer 1 is first, taking 6 microseconds to serialize into the 100Mbps microwave link. Customer 2 sends their trade data 2 microseconds later, so they wait 4 microseconds to serialize and another 6 to actually serialize. Customer 3 sends trade data 1 microsecond after Customer 2, waiting 3 microseconds for Customer 1, then 6 microseconds for Customer 2, then 6 more microseconds to actually serialize. Customer 4 is quicker; they send their data 1 nanosecond after Customer 3. And wait 2.999 microseconds for Customer 1, then another 6 for Customer 2, then another for Customer 3, then another 6 microseconds to actually serialize.

This example illustrates the supremacy of the owner of network in terms of speed in a shared bandwidth agreement. Jump shares its bandwidth with Perseus. While Perseus has additional customers, none will be as fast as Jump. This doesn’t mean Jump is the fastest operator in the HFT world, however. Despite purchasing a tall tower in a strategic place, they still aren’t as quick as contenders like….well, I’ll detail the fastest of the HFT competitors in Part III.

© Bloomberg

OUTRO

Thank you Katie.

Before posting the Part III of my story, a few words about the microwaves networks map I have published in Part I. More than 2,196 people have download it. This is huge. I didn’t think that this map would would garner such interest. I thought a few hundreds of people from the industry would have been interested. I’ll talk in detail about the “making-of“ in Part VI, but since a lot of people are interested, I must clarify one or two points.

I want like to make it clear that the map is a web of possibilities. Some paths (between two towers) are proven and used; others are possibilities, or old paths, or future paths, or tested paths. Sometimes it’s easy to find the real paths, but I talked enough with some competitors to learn that it’s virtually impossible to draw the perfect map – even if you visit all the towers, check the dishes, note the azimuths, etc. This is a job for an intelligence officer, not for an anthropologist. In the Google Earth file, paths have different opacities: bold paths are 100% sure – check the links, they draw you into the silent world of microwaves. Just one example:

What’s more, some microwaves networks are still in progress; and others may modify soon a path, jumping from a tower to another. Jump Trading had dishes in Belgium (in Oostvleteren) before purchasing the Houtem tower in January 2013. So, this is a map still in progress – I improve it as soon as I get new information –, and I update the file (almost) daily. If you want to be up-to-date, check the more recent file. Last update: two hours ago.

I’ll talk later about the mess regarding the data needed to draw such a map. I (and HFT competitors as well) have to deal with three national regulators (England, France, Belgium). For example, take Jump Trading, aka World Class Wireless (WCW). Like all the operators, they have to ask to the French regulator Arcep for using some frequencies in France. Even if they don’t use dishes in France, they need authorizations if a path goes over France (and the Houtem-Swingate path runs over France). So you can trace WCW activities by looking for the public documents released by the regulator.

Yesterday, a few microseconds after I posted the Part II, I realized a new document came out last week on the Arcep website, on September 16, 2014. About Jump Trading. Here are two screenshots:

I’m now quite familiar with these legal documents. It tells us that WCW is authorized to use two frequency ranges – 12,75-13,25 ghz and 17,7-19,7 ghz – between Belgium and France. That is a shame but Arcep don’t publish the annexes (where all the locations are specified) but if you are familiar with microwaves, you know that with a 12-19 ghz frequency range, you can’t do more than a 60-kilometer path. That’s why this document is interesting. Why Jump Trading would build a short path between France and Belgium? They have the beautiful tall tower in Houtem, so they have the perfect paths to cross the Channel an go deep in Belgium. So why they ask for new frequencies/paths in Belgium and France? One answer would be: Jump still use some dishes in Oostvleteren, and the new authorizations are valid only for this tower. That would mean: Jump may build two different networks – and that’s not impossible. Another answer would be: they will install new dishes on the Houtem tower – a sub-network? Who knows.

I was finishing this short post when I received a tip: Jump Trading may work on the Houtem tower tomorrow in the morning. I am not sure to be there, but I have a message for the readers who may live near Houtem: go there, and take a camera. Parti III will be online next Monday. Thank you for reading me.

I was in the downtown area of Brussels two days ago, and I passed a building which appears to be the office of the Belgian “Service Public Fédéral”, the same which sold the Houtem tower to Jump Trading. That was funny to see these offices are in a tower. A “Finance Tower”. I am surrounded by towers.

I was about to release the Part II of HFT in my backyard (an overview of the different HFT competitors using microwaves networks in Europe) but over the past days I have classified and read again my data, and I have found some bizarre details about one competitor. I need to check the data. I will post the Part III tomorrow, or Wednesday.

INTRO

Last Friday evening, I received a tip via telephone, “You should go to the Houtem tower tomorrow morning. Jump may work on the network.” Unfortunately I couldn’t make it back to Houtem. But it would’ve been amazing to see contractors adding or realigning dishes, or working on equipment. Who knows what they got up to on Saturday! While it appears the €5,000,000 Houtem tower was the best deal the Belgium government made in 2013, the question of whether the purchase was a good deal for Jump still remains. The military history of these towers prior to their use for HFT is fascinating.

To learn more about Houtem tower, take a look at these photos, check out this Google Group, and read an account of the removal of all equipment, antennas and associated hardware by the 1st Communications Maintenance Squadron: “The site in Houtem Veurne, Belgium, has a 799 foot tall communications tower. Not only did Airmen have to climb this enormous tower, but if the support lines gave way and it fell in the right direction, a portion of the tower would have landed in France. While many of the technicians won’t miss the painstaking climb to the top of that tower, most are sad to see it go since it was such a challenge”. A “Cable and Antenna Maintenance Craftsman” who worked for the United States Air Force based in Ramstein, Germany, inadvertently reveals in his LinkedIn profile that long before HFT, Houtem tower was all about high frequency: “ Team project lead of microwave antenna shroud replacement at the 800ft level, Houtem, Belgium radio relay site” and “[I] installed and maintained microwave, Ultra High Frequency, Very High Frequency, and High Frequency antenna systems, cabling, connectors and their structures.” The “U.S. Air Force Wire and Cable Dawgs” Facebook page also includes anecdotes and the only photograph I found which was taken from the top of the tower:

From the HFT microwaves industry I have also heard many stories. Two competitors visiting the same tower at the same time studiously avoided eye contact. A contractor working for various competitors complained about “all of these American guys asking for high towers… they just don’t exist!” Of course, some exist and are now used by high-frequency traders. There is this anecdote about Jump: “They were at a structure in Belgium waiting on an authority who controls this structure to give them permission. It was taking awhile. So they put a tower on wheels at the site. That pissed some people off. But then the tower was moving a lot. So they started pouring concrete around the base of the vehicle. The police were called out and there were forced to stop.” I don’t know if it’s true, but it’s fun – se non è vero è bene trovato. I was also told about a contractor who was supposed to install a dish for a firm on a tower already home to another HFT firm’s dishes. The workers made a mistake and pointed the new dish toward a tower used by a competitor!

Most invaluable to researching these microwave networks was the book The Death of a Pirate. British Radio and the Making of the Information Age (available in French here) by one of the greatest historians I know, Adrian Johns. This must-read tome focuses mainly on the famous UK pirate radios located on boats in the 1960s, but also cover the radios installed on the incredible Shivering Sands Army Forts, which were part of the Maunsell army Forts used for anti-aircraft defence during WWII. Thanks to these standing structures, the dishes were both more stable and taller than those on boats. By squatting in the old forts, broadcasters including Screaming Lord Sutch’s Radio City could use their height to reach more listeners and thus more revenue via advertising. It was all about height.

Here are the precise locations of the Shivering Sands and Red Sands forts:

See how the Shivering Sands forts are exactly on the straight line between NYSE’s facility in Basildon and Frankfurt’s Deutsche Börse. Optiver and McKay Brothers may each have a path very close to Red Sands as well (but I’m doubtful). I wonder if any HFT players thought to use these old forts. It would be amazing to replace the old pirate radio dishes with the big Andrew dishes Jump Trading favors. Torrent provider The Pirate Bay has already tried to install servers on another fort further north, the famous  Principality of Sealand to convert it to a “data haven”).

The most interesting discussion in The Death of a Pirate, however, is the history of the first radio pirates. They were not pop broadcasters earning cash using waves in international waters, but rather people who fiddled with the British Broadcasting Corporation (BBC) receivers to listen to non-BBC programmes like Radio Normandy. Waves are fruits of nature, to me a common good. When the first pirates emerged, states quickly monopolized these fruits and then were obliged to sort them. The International Telecommunication Union (ITU), originally founded in 1865 as the International Telegraph Convention, was charged with allocating waves. In the present, high-frequency traders have to pay state regulators for permission to use microwaves. If a firm wants to cross the channel from France to England, it has to ask state regulators for frequencies. Since most frequencies in the area were already booked or squatted, crossing the channel wasn’t an easy task – one competitor had some difficulties to cross the Channel. HFT is hard, and pirates may not be dead.

CUSTOM CONNECT

According to Bloomberg, “‘Custom Connect’s network was the first microwave link connecting the main financial centers in Europe that multiple trading firms could pay to use […] About 25 companies use Custom Connect’s network, including trading units at investment banks and high-frequency trading houses,’ said CC co-owner Jan Willem Meijer. ‘The network, which cost about 5 million euros ($6.8 million) to build, relies on 13 towers across Europe’”. Jump purchased a single tower at Houtem for the price of a whole network. My investigation of Custom Connect (CC) commenced in early July. I was visiting the CC website and at one point a chat window popped up and someone asked if he could help me. It was Jan Willem Meijer. When I told him I want to know where the CC towers are, he answered, “I can’t share. Confidential”. What a pity. On the CC website, however, there’s an  article from De Volkstrand (English version here), containing a map.

© De Volkskrant | April 30, 2014

While not a truly accurate map, it’s interesting. Belgium is the least transparent country regarding data, but I was able to use various dossiers to find some towers housing CC dishes. Others I located thanks to Cartoradio, the official map of the French Agence National des Fréquences (ANFR), one of the French state regulators with Arcep. If you click on the right tower, the names “Custom Connect”, “Global Connect”, “Latent Networks”, etc., will appear and there you have all the details about the dishes. Azimuths are of particular interest, as they tell us where the signal goes. This allowed me to draw my first paths on Google Earth:

Comparing the map from the Volkstrand article with mine revealed a problem with the Genappe tower (depicted as a white rectangle in the image above). Custom Connect should have two dishes on this tower, but the tower is north of the paths I drew. I have no data to link this tower to another, so my bet is on CC setting up a new network using the Genappe tower to replace the one in the south of Belgium. This is an educated guess, of course, but it is within the realm of possibility. Now let’s take a look at how the CC network crosses the channel:

Custom Connect has dishes on a 107.2 meter guyed tower at Mont Lambert, near Boulogne-sur-mer (where Vigilant also likely have dishes), and the Cartoradio data tells us the two dishes have a 327.2 degree azimuth. This means the CC signal goes to another guyed tower to the south of Dover. CC also has dishes in Wrotham and Dunkirk, England, but reviewing my papers before writing this post, something disturbing jumped out at me.

When I started my map by drawing CC paths in July, I was unaware of the precious data one could find on the Ofcom website. Ofcom is the English regulator which allocates radio waves to customers such as radio and television broadcasters, mobile operators… and high-frequency traders. The Ofcom map allowed me to check where operators have asked to use frequencies, which in turn allowed me to find towers in England that may be used by these operators. One competitor gave me the tip to use Ofcom’s map to search for licenses used to cross the channel. Click here for a list of the English-licensed operators using the 2-26 gigahertz range (which is used by high-frequency traders) in the north of France and Belgium:

You find Vigilant Global (aka Smartable LLC), McKay Brothers (aka Decyben), Jump Trading (aka World Class Wireless), Flow Traders (aka Global Connect), Optiver and Latent Networks but not Custom Connect, and that’s strange. I decided to compare the different data coming from the different regulators. I downloaded all the available public data from the French ANFR. The files list all the towers and all the dishes used (or not) by all operators. I dug a little and realized that all (without exception) the Custom Connect dishes in France are listed with this tag: “Dish authorized but not activated”. Very strange.

I dug more. I realized that the planning application for the CC dishes in Wrotham was very recent as the firm received an approval in July 2014. But the network is said to be operational since March 2013; Wrotham being in the middle of the Dover-Slough pathway, does that mean Custom Connect didn’t have a Dover-Slough link prior to July or August 2014? That would mean CC was only providing a network between Frankfurt and Basildon. There is more: the planning permit for the CC dishes in Dunkirk was granted in January 2014 – that’s problematic because Dunkirk is in the middle of the Dover-Basildon way, and this network is supposed to be operational since March 2013. What’s more, there is no Ofcom license with the name “Custom Connect” in Dunkirk. How they send a signal to Basildon? The most interesting is the planning application for the Dover tower:

© Dover District Council

First of all, the permit was submitted on July 25, 2014, and it’s not approved yet – files say CC wants to remove the dishes pointing to Boulogne, add a dish to go to Wrotham and keep the dish which points to Dunkirk. Secondly, the files also say two new dishes will have a 109° azimuth, so they won’t point to Boulogne but somewhere further north. Last but not least: the Ofcom database tells us CC has two licenses in Dover – 0996747 and 0956492. The license number 0996747 is needed to send the signal from Dover to Wrotham, and you can locate the license 0956492 in… Basildon. All of this sounds strange. I doubt it’s possible to build a path from Dover directly up to Basildon – if it’s possible, why all the competitors need at least one more tower between Dover and Basildon? What’s more, if CC has a license for a Dover-Basildon path, what about the planning application for the Dunkirk tower? That makes no sense (I can’t go more into technical details, but those who know will understand: it’s impossible to get a license on Dunkirk using the frequency assignment CC have at Dover and Basildon as the Dunkirk tower is in the same parity as Basildon and Dover.) So the situation is:

I talked with people from the microwaves industry. I have been told it would be possible to cross the channel without a license but that would have been tough – and not very legal. I also heard rumors about a competitor having installed dishes without permission on the Dunkirk tower (some say that could be Custom Connect). But rumors are rumors, and I prefer facts. Having said that, it would not be the first time a HFT player would sneakily install dishes. Once Jump installed illegal dishes on the famous Swingate tower and they humbly apologized:  “The two largest dishes proposed for World Class Wireless were actually installed in advance of the submission of the planning applications. Both WCW and Arqiva [the owner of the tower] wish to offer their sincere apologies for this oversight and no disrespect was meant towards the Council”. LOL. Both the Dunkirk and Swingate towers were a part of the famous Chain Home system built in 1937 (I’ll detail on Part V); they have been very important for the Royal Air Force during WWII and now they are a part of UK’s history. Installing dishes without permission on these historical monuments is a “felony”. These monuments are the crown jewels of the United Kingdom – be respectful, guys!

I don’t know about the illegal dishes in Dunkirk, but I know for sure CC doesn’t have radio licenses there. As for the channel, I have found all the licenses needed by the different competitors, except the Custom Connect one – perhaps they are hiding very well, using other names, etc. I visited again the Custom Connect website, hoping the chat window would open but nobody was there. But I discovered a very amazing detail. The motto of the firm seems to be a statement once made by Steve Jobs. Check the LinkedIn profile of Olav van Doorn (the other co-owner of Custom Connect). The description of the company starts with these two sentences: “‘It’s better to be a pirate than to join the Navy’ (Steve Jobs). Nothing describes better what Custom Connect is all about”!

What? Custom Connect runs a piracy business? No way… I thought radio pirates no longer existed! During my three-month investigation I have heard about some guys trying to pirate frequencies but these tales may by myths. I can’t believe one of the fastest state-of-the-art network in the world would pirate radio frequencies. That can’t be. But that would be amazing: the old Maunsell army Forts, built in 1937, have been used by the pirate radios in the 1960s; so I would not be surprised to find pirates in Dunkirk and Swingate, as the tower was built in 1937 too, for the same military purpose. There is much mystery about Custom Connect – data is probably missing – but what cannot be disputed is the fact Jump Trading once installed illegal dishes on the Swingate tower. Pirates are not dead.

OPTIVER

I didn’t know who I would find when I was drawing my first paths on the map. I was pretty sure to meet McKay Brothers (I knew they were working on a network) but first I met firms I had never heard of before. One is Communication Infrastructure Ldt. It’s probably a network provider à la Custom Connect, not a prop trading firm. I have found them everywhere, from Cornwall to France. One of the anonymous e-mails I have received when investigating (yeah, I received anonymous emails!) said: “Communications Infrastructure is a shell with no customers and only licenses to sell to a sucker like…” – stop. Wait a minute. I won’t publish the name of the “sucker” yet, as the firm will be a guest star of Part IV. But I confirm: Communication Infrastructure asked for a lot of licenses in England but I have found no planning application. I even doubt they have installed dishes (in France, some of them are “not activated”.) An empty shell – for now.

I also got to know NexxCom, which built a microwave network between New Jersey and Chicago. I discovered them in Cornwall, at the very end of the microwaves paths. There are two kind of competitors in Europe: some are only interested by linking London and Frankfurt (Custom Connect, McKay Brothers); others decided to spread their networks to Chicago (Vigilant, Optiver, Jump Trading, Flow Traders, Comm. Infra and NexxCom), so they have to go to South Cornwall where they can meet the Atlantic Crossing 1 (AC1) landing station – from there, data is sent in the US thanks to optic fibers. But in two years Cornwall will no longer be important, so I don’t really understand why NexxCom submitted a planning permit in Four Lanes last February. (I have heard NexxCom may be supported by a HFT firm but I have no name.)

One big competitor – far more serious than Communication Infrastructure or NexxCom – is Dutch prop firm Optiver. Optiver has some elegant paths and a good reputation in the industry. Some say they may be the fastest – being the fastest is cool, but the most fundamental task is to remain the fastest, and that’s not easy. Optiver has two licenses to cross the channel and they co-locate with Custom Connect and McKay Brothers in Dunkirk. I visited the Hannut tower in Belgium and shot their cabinet:

Optiver has interesting paths in Cornwall (they have offices in Chicago):

FLOW TRADERS

You won’t find the name “Flow Traders” (another Dutch prop firm) in any legal document, you’ll only read the name “Global Connect”. I had never heard of “Global Connect” before, but the postal address is Jacob Bontiusplaats 9, 1018 LL Amsterdam, where Flow Traders is located. I have nothing special to tell about Flow. Some say they were the first European firm to build a microwave network between London and Frankfurt (and they have licenses to cross the channel). Like Optiver, Flow Traders has paths until Cornwall. There is a third big Dutch firm in Amsterdam, IMC, but I couldn’t find evidence IMC has a microwaves network, even if in this LinkedIn page published on January 31, 2014, we learn the firm is “seeking a technical Project Manager or Lead Engineer to improve our efforts in building out the fastest network paths used for transmitting financial market data between major financial locations around the globe”. A new competitor may appear on my map soon.

VIGILANT GLOBAL

Here is Vigilant Global, a Montreal-based prop trading firm. Vigilant is now a subsidiary of DRW.  One people from the industry once confessed: “Vigilant is my best competitor.” Another said: “I actually admire Vigilant/DRW. They were first. They seem to be pretty class acts about it all.” Yes, Vigilant were the first to build a network in Europe. They probably initiated their network 2011, perhaps even in 2010 (there is a funny anecdote about an old dish on the Equinix data center in Frankfurt). By working on the map I understood why some people in the industry admire Vigilant: they are really good.

Ann Brocklehurst wrote about the difficulties Vigilant encountered when they tried to install dishes in Benfleet, but I don’t think this was a big problem. Why Vigilant is good? First, they don’t ask for a lot of licenses before installing their dishes (Jump ask for a lot of frequencies but often they don’t use them at all, even if they have to pay the UK regulator). I think the Vigilant engineers checked all the possibilities very carefully before asking for licenses. Secondly, they have some very elegant paths. I would be very curious to know if Vigilant really built these paths from Boulogne to Cornwall:

These are long and complex paths (microwaves don’t like dampness nor fog). But their best move is probably this path:

To understand it you must have in mind that a new optic fiber cable is in progress between England and the US: the notorious Hibernia cable (the plan has a long and fun history I will detail in Part V). It has been said the Hibernia cable should be faster than the AC1 cable by saving six milliseconds – that’s why the title of my book is 6. But the landing station of Hibernia (in Brean, Somerset) is far from the AC1 landing station (Whitesand Bay, Cornwall):

The Hibernia cable is due to open in two years. If the high-frequency traders can save six milliseconds with the help of new optic fibers, they won’t use the AC1 cable anymore. That would be stupid to have these fast microwaves networks without exploiting a new faster cable. That means allllllll the paths going to Cornwall will be dead in two years. And the map will look like that:

So Jump Trading, Vigilant and Flow Traders booked already licenses/frequencies to build new paths between Slough and Brean (Optiver may be around too), and Vigilant has anticipated the Hibernia cable with this long and classy path:

I would love to have a chat with the Vigilant engineers who designed the paths. These people definitively have good neural networks.

MCKAY BROTHERS

Let’s finish with McKay Brothers. Even if the first Chicago-New Jersey microwaves network was built by Alex Pilosov’s Windy Apple Technologies, everyone thinks McKay are the fastest in the US. In this document released by Aviat Networks (Aviat provides technology to most of – if not all – the HFT players, both in the US and in EU) you can read this: “The McKay Brothers LLC low latency network between Chicago and New York—the world’s busiest route for low latency financial networks…”. McKay is now a fierce competitor in Europe. Their network is still in progress but I have found towers where they have put dishes in place. One of their best path is between Liège and Sint Pieters Leeuw (in Belgium). Here is the only Belgian document I have found with all the details about the dishes put by a private operator (I parsed more than 18,000 documents to find it):

That’s how I could draw this 96,3 kilometers path between two of the tallest towers in Belgium:

The Sint Pieters Leeuw tower is the tallest standing structure in Belgium (300 meters). McKay having the fastest microwaves network in the US, I am not surprised to find their dishes on this tallest structure – a new “cathedral of capitalism”. I went to visit the building and shoot McKay’s dishes. Again, the tower is surrounded by potato fields, and right at the basement of the tower there is a small pen where a calf was quietly watching me shooting the tower. This beautiful building is owned by Norkring. Norkring is a is a network provider managing 24 transmitter masts in Flanders and Brussels. They even offer colocation: “On eight locations, Norkring België offers all facilities for hosting telecommunication operators, local broadcasters, emergency services, television operators, the police, or any other party that wishes to set up a wireless network. Several heights are possible for installing antennas on our transmission towers, ranging up to a level of 300 meters.” I don’t know where are these eight towers (you can’t find this information on the Norkring website) but I have been told Vigilant could use one (or more) Norkring tower. Other HFT players may also use the Norkring network.

© Norkring

But the most interesting part of the Sint Pieters Leeuw tower are these incredible stairs. I have been asked to appear in a French documentary about HFT and market microstructure, and I think the best place to shoot is there:

One last thing about McKay Brothers (and the parent company Quincy Data): they are transparent. They are the only providers to publish both prices and latencies “client port to client port”. “Client port to client port” means that latencies (e.g. 2.195 milliseconds between Basildon and Frankufrt, but McKay will be faster soon) are calculated for a trip between a co-located server (inside NYSE’s facility in Basildon) to another colocated server (inside Equinix data center in Frankfurt). I think they are the only ones to publish these latencies. What’s more, McKay Brothers co-CEO Stéphane Tyc advocates for more transparency in the market industry (check the last white paper he released two weeks ago about better execution). I think McKay Brothers goal is to "level the playing field" in the small world of microwaves networks. They are on the tallest structure in Belgium, but will they be the fastest? All bets are off.

One network is missing: Latent Networks. I had never heard of them it I found some of “their” towers. I tried to get information about them but the people behind the company did everything possible to hide. So I decided to dig. Then the first anonymous email arrived in my mailbox. Then the name of a big HFT firm from Chicago showed up – not Jump Trading. Then I started a strange relationship with a kangaroo. I’ll devote Part IV to Latent Networks – the kangaro deserves it, really. Stay tuned.

No luck: the Ofcom website/database is down at the moment (they are working on the servers), and I can’t post the Part IV of “HFT in my backyard” without checking data (it’s not about towers, it’s all about how then operators can buy and re-sell frequencies). I should have take screenshots before the website was down (how stupid I am). Anyway, if the Ofcom updates the database, it may be possible to get some new data about the HFT microwaves operators.

BACK TO DUNKIRK

In the meantime, let’s return to Dunkirk (the English Dunkirk, not the French Dunkirk – the French Dunkirk will be featured in Part IV). Dunkirk is on the Dover-Basildon path:

In Part III I talked about the fact Custom Connect doesn’t seem to have an Ofcom license there, and about the rumors saying they may have installed illegal dishes on this tower (illegal means: without planning permit, or before the permit was granted). I decided to dig more about the planning applications concerning this tower. Sometimes the devil is in the detail. There are four HFT players on this tower (Custom Connect, Optiver, McKay Brothers and Latent Networks), and they all have submitted permits.

The first was Optiver (submission in October 2013, authorization granted in December 2013). By checking all the documents, I found this comment (the eighth document), made by a neighbor: “There are already 2 Dishes installed on the Mast and have been there for awhile now so I’m presuming that this is what the application is now for!” How interesting! Before the first HFT player arrived in Dunkirk, there were two dishes on the tower (put before October 2013). Those of Custom Connect? I don’t think Optiver would have installed their dishes before authorization, they seem to be very very careful.

The second planning application was the McKay one (submission and authorization in November 2013) – no comments were made. The third one was by Custom Connect (submission in December 2013, authorization granted on January 2014). The fourth was the Latent Networks one (submission in January 2014, authorization granted in March 2014). The same neighbor who talked about the two dishes “already installed” made this comment: “Is there a limit on the number of dishes and small buildings on this site as we have I believe we currently 4 applications some already passed. Just concerned that if there is not a plan the site will be a mess.” This neighbor is well informed: there are four applications, all submitted by HFT players. That’s interesting because you can read the same statements in other applications, everywhere in the UK – people asking why, suddenly, some old towers are invaded by new dishes. It’s HFT, stupid! (Sorry… I don’t want to insult this lady.)

There is another comment (concerning the fourth and last application submitted by Latent): “Dunkirk Parish Council object to this application. The Council are concerned about the number of applications regularly being received for dishes on the mast and the cumulative affect that this may have. The health and safety risks of installing such equipment is not referred to in the supporting documents. The Council are also concerned that there is no certainty to the limit of future similar applications which may be received. The applications are each made by different telecommunication companies via different agents and there doesn’t appear to be any central control by the owner of the site as to what is installed on the mast and in the site buildings. Despite requests in the past the owner has not provided details of the programme of work, maintenance of the mast and site, and safety certificates in place. Again, there doesn’t appear to be anyone managing these aspects at the site. Dunkirk Parish Council feel that the committee should be aware of these concerns when making their decision. Louise Blackshaw Clerk to the Council” Bold is mine, as I already read the same statements in other planning applications. Sometimes authorities don’t understand who really is behind the permits. I don’t know why…

Let’s have a new look at the Custom Connect application. Words are important. Very important here, as the heading of the planning application is: “Retrospective consent for two 1.2m microwave dishes installed at a height of 40m on the existing 110m tower, associated cabling and the installation of a small equipment cabin at ground level”. Yes, you read correctly. “Retrospective consent” means: the dishes were installed before the authorization was granted. Rumors were facts: Custom Connect pirated the tower by installing their dishes prior to any authorization. Bad boys… (but we don’t know if the two dishes the neighbor talks about were the Custom Connect ones, or others).

Now read this other statement sent to the Dunkirk Parish Council in December 19, 2013. The same neighbor who complained about the two dishes “already installed” also wrote: “As we are having so many applications for dishes to be put on the mast which also entails building various cabins can it be confirmed that there is a limit on how many the mast can accommodate. Also what size are these buildings going to be ? I would also like to add that one of the applicants has already cut down some of the trees on the Public Road opposite my property NO 26 Courtenay Road.” What? Someone cut down trees? It’s bad. No respect for the environment. Microwaves require line of sight from one tower to the next but we can’t really know who cut down the trees – some HFT player, or someone else? Optiver was the first firm to have authorizations to put dishes on the tower, and I doubt they cut them between December 9 (when they received authorizations) and December 19 (when the neighbor did this comment). When the trees were cut down? The neighbor thinks it’s “one of the applicants” and the only recent applicant names you can find on the Swale Borough Council for the Dunkirk tower are some HFT operators. But we’ll never know if the chainsaw was the one of the pirates…

The Ofcom spectrum website being down, it’s nearly impossible to continue the “HFT in my back yard” series seriously. I have most of the data needed to finish Part IV, but I don’t want to write an incomplete account of Latent Networks – on which I just found the last “smart move”. What’s more, I discovered a new microwaves network here in my backyard. In fact, it’s an old one – to be honest, he never existed but the story if verrrrrrry interesting (in the end, I think I could write a little book about all this stuff). I still have to review some US legal documents (this if my new gig) and have to check if I can report publicly what I have found – I already entered a private property by visiting the Jump tower in Houtem, some FBI and US Defense people came to read my blog, so I don’t want do make things worse for myself – jail is not cool I think. Last but not least,  I have to finish a 4,000 words article for a French academic journal of anthropology about my first hypothesis regarding the “nature of market microstructure”, and I’m quite late (very late, in fact). So I’ll post Part IV early next week, hoping that the Ofcom website will be available again in the next few days.

An interesting tower I’ll talk about in Part IV

Article protégé par mot de passe. Il faut visiter le site et entrer le mot de passe pour lire la suite.

“Like so many things, our technological Culture must « fail » if it is to succeed, for its very failures constitute the thing that it is trying to measure, harness, or predict. If the formulae and predictions of science were completely effective and exhaustive, if the operations of technology were completely efficient, then nature would become science or technology itself. (This is actually how we talk about things in our modern world of contextual relativity: nature is « system, » it is « biology » or « ecology, » whereas Culture is « natural, » an « evolutionary adaptation. ») Science and technology « produce » our Cultural distinction between the innate and the artificial to the extent that they fail to be completely exact or efficient, precipitating an image of « the unknown » and of uncontrolled natural force. Thus it is that science and technology are aligned on the side of conservatism in modern America. But it should be emphasized that even technologically our Culture « works » in terms of objectification, and only incidentally in terms of energy and efficiency.”

Nothing to do with HFT nor microwave, but this excerpt comes from a classic (and a must-read) book called The Invention of Culture, written 40 years ago by anthropologist Roy Wagner. Compared to this complex but very important book, my posts about HFT are as simple as Mickey Mouse Magazine. As my book company will publish the long-awaited French translation of the book in November, I have to send it to the printer this week, so my neural network is quite busy. What’s more, since the Ofcom website opened again, since I re-checked some data, I got some new (and very sensitive) information from the industry, which forces me to rewrite a section of the “HFT in my backyard” Part IV. This part is probably the most complex to manage. I’ll post it next Monday – I need to take time and be very careful. Have a nice week by then.

There’s a thin grey line
Between the black and white
It’s evidently hard to find at night
Depeche Mode, Alone

Times have been intense since Part III (a lot of work to do not related to HFT). I heard some people were “jumpy” because I put public data on a map, but others were quite friendly. I thank those who pointed out the mistakes I made. I am not a journalist, nor am I a whistleblower. I just wanted to see the EU microwave networks as I am trying to work on the “nature of market”, the topic of a Ph. D. for which I am trying to find a grant. French King François Hollande once said “finance has no face” (as the Ceres statue at the top of the CBOT in Chicago) but by mapping the microwave networks it’s possible to give high-frequency trading a face. A blog is not an academic space, and even if I am trying to be as accurate as I can (that’s not so easy, trust me), I’m trying to maintain a sense of humor.

That being said, during my investigation (which was more a diverting hobby than a full-time job), I had never thought to come across some stories that could be turned into a plot for a TV series on HBO or Netflix. The small world of HFT microwave networks is full of surprises. One of them is the only network I didn’t talk about until now: Latent Networks. Thanks to Latent I have learnt a lot about how nature (i.e. waves) is a tradable product for the high-frequency competitors. This episode was quite hard to write (I am on a grey line here) and the devil is in the (technical) details – if you are not familiar with the microwave world, I advice you to read my previous “HFT in my backyard” posts. Attention must be paid.

FROM MALTA TO CYPRUS TO PANAMA

I came to know Latent Networks when I discovered Cartoradio, the website of the Agence national des fréquences. With Cartoradio I started to map the towers in North of France. There was nothing special about Latent – they received a first license in France on February 5, 2013. Since I had never heard about the network, I downloaded the official list with the contact details of all the radio companies active in France. All the HFT competitors are transparent in the sense they have a postal address and offices with people working there. The first intriguing thing about Latent is that the company is incorporated in the Republic of Malta, and I thought it was quite exotic to find a HFT network in a tax haven. I asked some people in the industry about Latent and answers were evasive (but I realized later that some knew more than they told me first, I think they were waiting to read my first posts before giving me more data).

At this time (around mid-July), I found the firm in the official Malta Registry of Companies but the only information available are the name of the firm and the company ID; however, if you spent a dozen of dollars you can buy the “certificate of registration”, where I thought I could find names… but instead of names, I suddenly jumped from Malta to Cyprus and Panama. In short: Latent Networks Limited was incorporated in Malta on September 4, 2012, and located in Sielma, Malta. The secretary of the firm is a man with a Polish name, Krzysztof Kubala, who works for a firm called Taxways Group. The Malta address of Taxways Group is the same as the Latent one, but the main office of Taxways is located in Warszawa. There is another company involved in the Latent certificate of registration, called Alter Domus (such a lovely name for a tax haven specialist).

The best are the two shareholders of Latent Networks Limited: the first is Invest Group Limited, also located in Sliema, Malta, along with Latent and Taxways; the second (and main) one is Batlaw Holdings Limited, a firm incorporated in… Cyprus. If you search on the Cyprus public Department of Registar of Companies, you quickly learn that one of the directors of Batlaw is Krzysztof Kubala, the man behind Taxways in Malta, who appears to also be the CEO of Invest Group; what’s more, the Cyprus Batlaw secretary is another Polish person, who has connections with Taxways, Invest Group and Latent in Malta. You are lost? Look forward to the next move: on November 23, 2012, a few weeks after Latent was incorporated, the Batlaw shares were transferred from Cyprus to a company named Amicalle Corp., in… Panama. This a new kind of network. A SuperMontage. If you register to the Registro Publico de Panama, you learn the shareholders of Amicalle Corp. are two different firms, Dubro Limited SA and Aliator SA, and both these firms have the same shareholders, two other firms named Cheswick Inc. and Eastshore Inc. If you try to know who is behind Cheswick and Eastshore, you find a lot of directors, secretaries, etc., who also are directors, secretaries, etc., of hundreds or thousands of other firms. Ladies and gentleman, welcome in Panama! All those firms are located in East 54th Street, City of Panama, which is a tower, a new kind of tower in the HFT microwave world:

No doubt that those who created Latent Networks really love tax havens. Panama is a dead-end. I suspect the myriad of firms in this tower to be managed by Mossack Fonseca & Co, a big law firm which “is one of seven that collectively represent more than half of the companies incorporated in Panama”, according to Wikipedia. Back to Malta, I realized that at the same address in Sliema there are other firms connected to Latent: NGN Europe Limited, incorporated a week after Latent in September 2012, owned by Latent; and two others companies named Latency Engineering Blue Limited and Latency Engineering Green Limited. I am not paranoid at all, but given the Malta-Cyprus-Panama-Warszawa network, one may wonder why a high-frequency microwave network needs a worldwide SuperMontage. I thought such sophisticated configurations were made for drug cartels, mafias or banks. The Latent SuperMontage would be perfect in a novel à la John Le Carré.

 THE NEWGIG “SCAM“

I was more interested in finding guyed towers in Belgium than investigating the rough areas of tax havens. One informant told me to look at a firm called Newgig, saying it may have connections with Latent, but I didn’t pay attention. It was a mistake: two weeks ago, when I was gathering up all the data I had, I finally checked out Newgig, only to understand why someone wrote me about the firm with these words: “they were pure scam artists – as in, criminal fraud”. Wow. The Newgig story is worth reading indeed– and a good opportunity to go deeper into some complex issues.

Newgig Networks LLC was a Las Vegas company created by Carl Petrosky and Jake Zoldan in 2009, according to Zoldan profile; early 2012, the founders “jumped into the microwave industry when they smelled money”, according to a person who crossed their path. Around 2011-2012, in the US, a lot of people were trying to set up microwave networks by connecting the different market venues, the most important route being the New Jersey-Chicago one. Alex Pilosov’s Windy Apple Technologies network was the first to be achieved mid-2010; Marty Snyder’s Communication Infrastructure Corporation (CIC) was working on different routes too, along with other firms (check this interesting 2012 article). Newgig decided to come on board and appointed Aviat Networks in the US and Nexxcom in EU to build two networks – New Jersey-Chicago, Frankfurt-London. But things went wrong mid-2012. The rest of the story can be found in the Superior Court of the State of California records and there – and it’s quite interesting.

The story: Newgig pre-sold the bandwidth of the two networks to trading firms; the money was supposed to pay the suppliers mandated to build the networks (Aviat and Nexxcom). On March 31, 2012, Newgig and Nexxcom ”entered into a agreement involving the construction, implementation and activation of a wireless network between United Kingdom and Germany“, according to Nexccom’s “Cross complaint for breach of contract” filed  at the Superior Court of California later this year. But around June 2012, Newgig failed to pay the first invoices sent by Nexxcom. Nexxcom « is informed and believes that in July 2012, Newgig was financially overcommitted to other projects, including that Newgig was 100% over-budget”. According to Nexxcom, at this time one of the two founders of Newgig admitted that they were “100% over-budget, many months behind schedule and incurring penalties from customers that hey had pre-sold services”; Newgig also admitted “that the project was running into trouble as the route passed through Pennsylvania and New Jersey”. Thanks to the Federal Communications Commission (FCC) website (FCC is the American telecommunication regulator) and with the data available in the Court records, I have been able to easily draw the Newgig path:

It seems Newgig got tripped up by the famous Allegheny Mountains, discussed by MacKenzie, Beunza, Millo & Pardo-Guerra in “Drilling Through the Allegheny Mountains”. The Nexxcom cross-complaint states that instead of using microwave, Newgig “elected to use a fiber optic network in addition to wireless networks to fix the problem” and admitted “that this solution created additional unforeseen costs”. What a network… The idea behind a microwave network is not to use fibers, as microwave is faster than fiber; by failing to build a true microwave network, Newgig could not offer anything to the trading companbies which paid deposits to use a faster network. A lot of people I talked with are convinced Newgig never intended to really produce the networks. As a microwave industry veteran told me: “They admit to ‘taking deposits’ [i.e. advances paid by the trading firms], which means that the scam was their normal mode of operation. My impression is that they collected a ton of money as deposits and just ran away”.

Whatever the case, Newgig went bankrupt in May 2014 (the firm even had difficulties to pay their lawyers). But the most interesting in the Court records and the bankruptcy filings is that you have all the details about the network they were supposed to build and sell in Europe. The total network construction was estimated a little more than $5,000,000 (that’s consistent with the price Custom Connect paid to set up their network); sites acquisitions were estimated $1,000,000; spectrum/radio licenses were estimated $180,000 (By the way: I learnt that in the US frequency licenses are nearly free; that means the US regulator doesn’t consider waves – a common good produced by nature – as a real product for sale; that’s quite amazing to see that in one of the most capitalist countries in the world there are things which are not tradable). Above all, there is the map of the network which-never-existed. I’m not sure it would be legal to reproduce it here; besides, the Court records are are in low-definition and you can’t barely read the map, but it’s possible to foresee that from Basildon the network would have crossed the channel from Dover to Calais and crossed over Belgium with paths further to the north than the 2014 competitors routes. 

Last but not least: if you spent a few dollars, you can access the names of the 49 creditors of the deceased firm. An impressive lost: tower operators, microwaves contractors, cable operators (Hibernia Atlantic), fiber operators, exchanges (NYSE Euronext), markets data centers (Equinix), etc., and… car companies: Jaguar Land Rover, BMW and Mercedes Benz – the Newgig people loved big cars? More interesting, there are two trading firms: Dutch company Flow Traders and New-York based Tower Research Capital (also know as Latour Trading). I assume the two firms paid deposits to use the Newgig networks. I was wondering why my posts about the microwave networks received a lot of visits from Tower Research Capital (apart from their explicit name): Tower Research Capital is the main creditor of Newgig, and I can understand that if you pay a $1,500,000 deposit to be faster than others and you get nothing, there is a reason to be on edge.

CHICAGO IN BELGIUM, AGAIN

The Newgig affair was interesting but I found no connection between Latent and Newgig. I think some people were assuming the name NGN Europe Limited looked like Newgig Networks but that’s very speculative. I was not obsessed by Latent when I was trying to map the networks (I only thought the tax havens SuperMontage was quite unusual) but by sheer coincidenceI I finally found something. At some point I decided to check all the French licenses having cross-border authorizations (France-England and France-Belgium) as I wanted to be sure I wasn’t missing an unknown network. That’s how I found a document you can’t bring to light by searching for the name ”Latent“ on the Arcep website. The document is here. This is the last French license granted to Latent Network on February 29, 2104, but the application was not submitted by Latent; the applicant name is ‘Global Colocation Services LLC’. In other words, this is a very well known firm from Chicago named… Getco.

Global Colocation Services (GCS) is a subsidiary of Getco. Nowadays Getco is known as KCG since the firm merged with Knight Capital last year (Knight collapsed on August 1, 2012 because of a little problem with one of their smart order routers). Getco has been one of the first Chicago high-frequency trading firms, founded in 1999 by Daniel Tierney, a former trader on the CBOE, and CME ex-broker Stephen Schuler. I was quite happy to find Getco in Europe because along with Jump Trading’s founders, Tierney and Schuler also were on the famous platform (the “tower”) erected above the CME pits I talked about in Part I (I was honored to learn my post about the Great Fed Robbery was “bookmarked” in the internal Getco wiki in 2013).

But what’s the connection between Latent and Getco? I didn’t think Latent was Getco. I can’t imagine KCG would need tax havens (all you need in the US is in Delaware, no?). I googled the name “Global Colocation Services” in France and Belgium and realized GCS was incorporated in Belgium (Brussels) on October 25, 2013, and in France (Paris) on November 25, 2013. No doubt these two firms were created to manage the Latent licenses. Here is proof: I found the name “Global Colocation” in the French telephone directories, and when I read the postal address I just laughed.

I knew very well this place, in Dunkerque, North of France, even if I never went there. Remember the chart I talked about in Part I which compares the tall Jump tower in Houtem and an other tower in Dunkerque. The “Global Colocation” address is located on this tower, known as Tour du Reuze, where most of the competitors (Vigilant, Optiver, McKay, Latent) have dishes to cross the channel:

The “Tower of Reuze” is very small compared to the Jump tower in Houtem, but there is an interesting story about it. The tower was designed by a Belgian architect and built around 1974 using reinforced concrete and smoked glass. It’s pretty heavy: 27,000 tons, equivalent to four times the Tour Eiffel (!), which means that it’s rather impossible to destroy. The Tour was supposed to be a 3-star hotel with 126 rooms and a luxurious panoramic restaurant. But as far as I know the restaurant never existed, the rooms were transformed into apartments and the building quickly became in bad shape. Some residents consider it as an “ugly architectural wart” (see there). The building and the apartments were so “rotten” that the real estate developer who acquired the tower decided to restore it a little bit and to sell the whole building for €1,700,000 in December 2013. By digging a little you learn that some apartments didn’t even have internet access until recently. Here is a photo of the Tour du Reuze, with a dish pointing to the UK (owned by a HFT competitor?):

And here is the panoramic space designed to be a restaurant – one competitor told me the space is “stinking of urine”:

The only luxurious things there are dishes. It is of interest to see some state-of-the-art microwave networks working with one microsecond accuracy are installed in such a poor place. Last year, a few months after I paid a visit to NYSE-Euronext’s facility in Basildon (the most spectacular there is the emptiness), I had the chance visit to the Euronext control room in Paris, where dozens of very big screens show market activity. When I left the office I walked a few meters and I came across a phone booth where a young woman and two very very young near-naked children were sleeping. The contrast was striking between a very rich technological world and a very poor human world. I had the same feelings when I was reading discussions about the future role of high-frequency trading in the Nigerian market, Nigeria being one of the poorest countries in the world. In Tour du Reuze, sophisticated trading firms use wireless to silently send flow of orders at the speed of light above the head of people who barely have access to internet. We live in interesting times. That being said  the rents for the HFT dishes surrounding the panoramic space are not cheap, and that’s good news: that proves high-frequency trading may have a social benefit;)

Let’s get back to Latent Networks. The fact Getco/Global Colocation Services is taking over the Latent frequencies can be confirmed by visiting the official UK radio regulator website. On the Ofcom website there is this map allowing you to find the radio operator licenses. But there is another space called “Spectrum trading”. Before high-frequency traders can use microwave networks to trade whatever-you-want, the network providers have to book and pay for frequency licenses. In the UK a license is granted for a path between two points/towers. This license is a tradable product in a sense that if a radio operator doesn’t need it anymore, the “Spectrum trading” space allow them to re-sell their licenses to who is willing to buy it. As a product of nature, waves are tradable – it’s the market after all.

A few days after I found the French license with both the name “Latent Networks” and “Global Colocation Services”, I realized that Latent sold two UK licenses to Getco on March 31, 2014. The two licenses are those Latent Networks acquired to cross the channel from Tour du Reuze in Dunkerque to Dover in the UK (that’s consistent with the French license). Two weeks ago the Ofcom website was down for server maintenance but when it came back… what a surprise it was to see that on September 23, 2014 Latent sold other licenses to Getco:

I assume that’s the reason Global Colocation Services LLC was incorporated in the UK too. If you compare the Latent licenses numbers with those purchased by Getco, you quickly understand that Latent has sold to Getco the whole path from Tour du Reuze to Basildon NYSE’s data center (for once I’m sure these definitive paths on my map are 100% accurate here;). Here is the Latent-Getco network from Germany to Basildon. I know some towers are missing in Belgium but there is no available public data in my country, so I don’t know where Latent may have installed dishes between Tour du Reuze in Dunkerque and Incourt in Belgium (I left the Latent tacks corresponding to some old towers not used anymore):

DIRTY BUSINESS AROUND THE CHANNEL

Once again all this doesn’t prove that Getco is behind Latent – Latent just sold UK licenses to Getco, period. But I thought it was quite strange as there are two kinds of competitors in the HFT microwave world: trading firms like Optiver of Flow Traders, which build their own network for themselves, and providers like McKay Brothers or Custom Connect. These sell bandwidth to customers (trading firms, hedge funds, banks, etc.) but they don’t re-sell frequencies, paths or licenses to trading firms. The Latent case is therefore quite unusual: the network is not a provider selling bandwidth to customers, and it’s not a trading firm. I didn’t find public data showing that Latent sold UK licenses to other firms than Getco. Anyway, the fact the first licenses Getco acquired from Latent in March 2014 was those needed to cross the channel was interesting to note. More than I expected.

During my investigation I came across an absorbing public document found on the Ofcom website. This a statement made on December 13, 2013, at an Ofcom stakeholders‘ meeting (here stakeholders are operators licensed by Ofcom which are invited to talk about any problems they face). The statement was made by Ian Marshall on behalf of Aviat Networks. I don’t know Mr. Marshall but as a “Regulatory Manager” of Aviat he wrote several articles on his blog about wireless technologies. The paper untitled “Rules of the Game: Low Latency Microwave in a Multi-Regulatory Environment” is of particular interest. Bear in mind that Aviat Networks is a technical provider for various radio operators, HFT competitors included. “When Aviat Networks helped build the McKay Brothers LLC low latency network between Chicago and New York”  wrote Marshall, “there were some technical testing challenges with which to deal. […] As complex as that undertaking was, it turned out to be simple in comparison to the level of regulatory complexity we experienced when we went to Europe to construct a low latency network linking the continent’s leading financial hubs of London and Frankfurt. […] When building a trans-national low latency network, the regulatory bottlenecks will always occur on the border-crossing link. These can be further exasperated if the border is also a geographic challenge such as an ocean.”

Two months after writing this article, Mr. Marshall used the same words at the Ofcom stakeholders‘ meeting in December 2013. Here is the first paragraph of his statement : “Aviat Networks understands that at the present time Ofcom does not impose any build out requirements once a licence has been granted for a point to point link. However, we have become aware that there are a large (>200) number of licences recently granted to Low Latency (High Frequency Trading) organisations that we have very good reason to believe have not resulted in the building of the associated network infrastructure. It is our considered opinion that there is a large amount of speculation occurring here and this ultimately leads to ‘blocking’ on the key routes across the U.K.” Here is the next paragraph, with the Swingate towers near Dover where most of the HFT microwave competitors (Optiver, Vigilant, Jump, McKay Brothers) have put or will put dishes to cross the channel (the red underscore is mine):

All this is very interesting. In short, Mr. Marshall told the UK radio regulator that some HFT competitors have asked for licenses without installing dishes just in order to “squat” frequencies and block competitors. In other words: you ask for a large number of frequencies so that your competitors can’t get them anymore – in this case, you can’t cross the channel, and if you can’t cross the channel, you are not able to build a network between Frankfurt and London. Simple to understand: trading firms need waves; without waves, the firms won’t be the fastests. That’s why there was a “a large amount of speculation occurring” around the channel. Mr. Marshall’s statement is consistent with the “rumors” about “frequency squatting or tower squatting” mentioned by Bloomberg. But there, around the North Sea, rumors were facts. I know for sure that (at least) two competitors have experienced real difficulties to get frequencies to cross the channel, because of the frequency squatting or speculation. One was a microwave network provider; the other one was a proprietary trading firm. So I wondered: who were the squatters? Who tried to speculate on the waves? A few days later an answer popped-up in my mailbox. And it was just staggering.

On August 3 at 23:13, I was quietly going to bed when I received an email untitled “Latent and more”. The mail was sent through paranoici.org, an anonymous remailer used “to hide one’s identity when sending messages by e-mail.” Strange, no? I read the first lines of the mail. “What… the f…” I told myself, wide-eyed. But let’s start with the three last lines: “If you want to know more, send a sign / Happy Hunting / Kangaroo Pin”. The only way someone would have known what I was working on was through Twitter. I posted one or two tweets about my investigation, and I was sure none of my informants in the industry wouldn’t have sent me an anonymous mail. The “kangaroo” thing was odd. The signature being “Kangaroo pin”, I suspected my anonymous informant to be a Jump Trading employee. Why? Because what a kangaroo does? He jumps! So I posted a cryptic tweet to the kangaroo but it was stupid. Really stupid. The kangaroo was not a Jump employee at all. One HFT competitor read my tweet and wrote me that I had been mistaken, adding that one person behind Latent Networks wears a kangaroo pin on his jacket. Wow. The “kangaroo pin” was not a signature, it was a clue. The kangaroo was one of the Latent people I was searching for. Incredible.

But the most astonishing in the anonymous mail was the first paragraph. Take a deep breath: “Look up [the kangaroo]. [The kangaroo] used to work at [the Firm]. [The Firm] sells a lot of microwave gear into the HFT vertical. [The kangaroo] had access to a lot, if not ALL of the routes of every HFT company out there. [The kangaroo], while in the employ of [the Firm], started a company himself, called Latent Networks Limited. [The Kangaroo] stole the paths of these companies, and then he registered for frequencies in order to squat them. His main and probably only customer is Getco. The ‘rumor’ is that Getco bought some channel licenses from Latent Networks Limited.“

Wow wow wow. Let’s take a break. My first aim was to map the EU microwave networks. By working on these networks I just wanted to understand the way high-frequency traders deal with nature. I searched for public data and tried to understand this small world with the help of people from the industry who kindly and fairly talked to me. One of these networks was quite unusual so I decided to dig more, but I would have never imagined I would receive a mail accusing a firm incorporated in Malta-Cyprus-Panama to speculate on the North Sea waves. As anyone familiar with the HFT world I knew the stories about some employees accused to have stolen trading codes from a firm to take them away (the Sergey Aleynikov case being the most well-known), but I never thought it would be possible to misappropriate frequencies in order to speculate on them. These were harsh accusations.

The last two sentences of the mail were quite accurate: Getco is a customer of Latent, and the “rumor” about the channel licenses are facts confirmed by the Ofcom public data. But there is nothing illegal here: Latent may have acted like a high-frequency trader by quickly buying a rare product (waves) before reselling it to who was willing to but – it’s the market after all. But the words “ALL of the routes of every HFT company” are not accurate. The Firm the kangaroo worked for is not the only one providing technical services to the microwave networks (there are others), so I was very doubtful the kangaroo could have known all the paths of all the microwave competitors. Not impossible, although not likely. But what about the accusations of theft? My first thought was: it’s bullshit, someone is trying to confuse me – this is a joke.

Two weeks ago I was about to write “Latent may have speculated on the channel licenses but front-running is not illegal here.” But ten days ago, two months after I spoke about the kangaroo with my informants, one of them decided to talk and blowed my mind. This competitor told me that as a technical provider, the kangaroo “took away” a path from him in order to register it in the name of his new firm called Latent Networks. “Taking away » is not “stealing” though (I can’t go into these sensitive details), but what I learnt from this competitor was not unrelated to the allegations made in my anonymous email. I couldn’t believe it.

There is more but as Ludwig Wittgenstein once said, “what we cannot speak about we must pass over in silence”. The way I talked with people in the microwave industry during my investigation was very fair. I’m not a prosecutor but by digging as far as I could, I came to hear about details I can’t share here out of respect for the people I talked with. Just let’s assume I may know what happened around the channel before Getco acquired the Latent licenses. From what I know, considering that at least two HFT competitors had difficulties to get frequencies between France and the UK, and considering the speculation on the channel waves, I have one question in mind: which was the price Getco paid to buy the channel licenses previously purchased by Latent?

THE KANGAROO AND ME

I won’t publish the name of the kangaroo and I won’t publish the name the firm he worked for before moving to Latent Networks. I know the kangaroo knows that I know who he is, but the most interesting was to study the way some microwave networks deal with nature (waves, fog, water), not to hunt an obscure firm stashed in Panama – but I think I now have the stuff to write a full TV series. When the anonymous email came in my mailbox I was both amazed and furious. Furious because I would have preferred to find the name of the kangaroo myself, as the investigation was quite fun! (Three days after I conducted a close examination of the Ofcom public data concerning Latent Networks – the Polish connection, the kangaroo, the name of the Firm, all this was very easy to flush out.) I was amazed not only because of the allegations, but because of the name of the kangaroo. “No… wait… that can’t be true”, I told myself. Here is why I couldn’t believe my eyes.

In April 2013, two months after the release of my first book, I decided to work more on high-frequency trading and asked for an authorization to attend the Trade Tech Europe 2013 conference in London. The first day was fully devoted to high-frequency trading, with some interesting moments I related in this post (FR). At the end of the day I gave a short “speech” about anthropology and HFT because the Trade Tech organizers allowed me to attend the entire conference if and only if I would give a talk (I couldn‘t afford to pay the requested €3,000). The day after, while discovering the different booths, a manager from the Dutch regulator AFM came to me and had very kind words about my rubbish talk (we discussed about finance, Michel Foucault and the Panopticon), and a French physicist told me about his view on Nietzsche and Toynbee. But the first person I talked with that day was an appealing man wearing elegant glasses. We gently talked about epistemology and he told me about his job. The appealing man with a Polish name, whom I never met again, was working for a firm providing technical supports to the HFT microwave networks. I don’t remember the pin, but the man was the kangaroo.

Further to Part IV of the “HFT in my backyard” series where I tried to catch Latent Networks, I received a letter from Malta untitled “Latent Networks Response to “HFT in my backyard IV“. Latent kindly asked me to publish this answer here, and I think it’s fair to comply with the request. Some answers may lead to other questions, which may lead to other answers, etc., but I’m not good at ping pong so let’s leave it there. Here is the Latent answer (also available in PDF):

I have been invited for a talk at the International Workshop Series: Investigating High-Frequency Trading. Theoretical, social and anthropological perspectives conference. The first workshop will take place in Copenhagen next week, on November 24-25, and I’m glad to catch up with Donald Mackenzie, Juan Pardo-Guerra, Alex Preda, Marc Lenglet ant other fine researchers for my first speech about HFT in an academic conference.

The full program is now available (more about the workshop series there):

Workshop 1
Copenhagen Business School | 24-25 November 2014
« HFT sociality, crowd psychology and dynamic collectives »

Day 1 | Monday 24 November 9:30–10:00
Introduction
10:00–11:00
‘Rhythms of the Market Crowd: A Lefebvrean Rhythmanalysis of Financial Markets’ Kristian Bondo Hansen
11:00–12:00
‘Trading Machines and Communication in Finance: Do Markets Still Chat in the Era of Robots?’ Alex Preda
12:00–13:00: Lunch Break
13:00–14:00
‘A Material Sociology of High-Frequency Trading’ Donald Mackenzie
14:00–15:00
‘High Frequency Trading and Price Discovery in the UK Equity Market’ Evangelos Benos
15:00–15:30: Coffee Break
15:30–16:30
‘How to Succeed as a High-Frequency Trader: Problematizations of Trading Subjects in an Era of Automated Trading’ Christian Borch
16:30–17:30
‘The Varieties of Affective Relations in Socio-Technical Collectives: A Study of Automated Trading’ Robert Seyfert

Day 2 | Tuesday 25 November
9:00–10:00
‘Where do Electronic Markets Come From? Regulation and the Transformation of Financial Exchanges’ Yuval Millo
10:00–11:00
‘The Sociology of Algorithmic Trading and Market Structure Regulation’ Nathan Coombs
11:00–12:00
‘Taming markets? A Philosophical Perspective on The Regulation of High-Frequency Trading’ Marc Lenglet
12:00–13:30: Lunch Break
13:30–14:30
‘From Crowds to Queues: How Technology Transformed the Nature of Finance’ Juan Pablo Pado-Guerra
14:30–15:30
‘Crowding of Adaptive Strategies: Signals and Noise in High-Frequency Trading’ Ann-Christina Lange
15:30–16:00: Coffee Break
16:00–17:00
‘Crowded Networks: Microwave Technology and the Nature of Exchanges in the HFT world’ Alexandre Laumonier

The first Investigating High-Frequency Trading. Theoretical, social and anthropological perspectives workshop being finished, it’s time to taste a « Black Imperial Stout beer » by Mikkeller before going to the airport. The chocolate & coffee flavour is fantastic.