April 9 Update: Read my new article: How HFT firms access secure government briefings to get the jump on market-moving data
You’re here because you googled FCC license 1215095, right? Perhaps you’ve already discovered the registrant, Converge Towers LLC, and its corporate ties to Cantor Fitzgerald, a New York financial firm best known for the devastation it suffered in the 9/11 attacks.
Or maybe that information is incorrect. (See the comments.) In my work, I discover quite a lot of errors and outdated information in internet databases. I haven’t actually called the FCC to check so for now, I can’t actually confirm anything about 1215095.
I think I might, however, know the story — or at least part of the story — Michael Lewis hints at at the end of his fascinating new book, Flash Boys: A Wall Street Revolt.
That’s because a couple of months ago I wrote a post about how Vigilant Global — a Montreal-based HFT prop shop was building its own microwave networks.
I later learned this was something of a trend so I wasn’t surprised, but rather perplexed, by the final paragraph of Michael Lewis’ fascinating new book, Flash Boys, where he writes about a microwave tower he discovers in the wilds of Pennsylvania:
The application to use the tower to send a microwave signal had been filed in July 2012, and it had been filed by … well, it isn’t possible to keep any of this secret anymore. A day’s journey in cyberspace would lead anyone who wished to know it into another incredible but true Wall Street story of hypocrisy and secrecy and the endless quest by human beings to gain a certain edge in an uncertain world. All that one needed to discover the truth about the tower was the desire to know it.
Any inside information I have about this situation comes mostly from the anonymous correspondents who have gotten in touch with me over the years due to my coverage of HFT. I have written about it from a completely different perspective than Michael Lewis, and I can’t help but wonder if the hint at the end of Flash Boys is pointing to this other HFT story.
For the record then, here’s what one of my anonymous sources told me about HFT firms’ microwave networks:
Vigilant (which is now owned by DRW) had the fastest microwave line from Chicago to NJ … They shopped themselves around and Tower just, and now regrettably, missed out. DRW needed better technology which is why they bought (Vigilant) — not because it was making a lot of money, in which case they wouldn’t have been for sale.
According to this anonymous source, there are also microwave networks between Chicago and Washington, D.C., where economic news and indicators, are regularly announced by government agencies. To get their hands on this information as quickly as possible, several HFT shops even set up not just their own private microwave networks but also their own “news agencies.” Vigilant Global, for example, funded the now defunct Canadian Economic Press or CEP News. And the Wall Street Journal had a front page story last summer on the connections between Need to Know News, its owner Deutsche Börse, and various Chicago HFT traders.
According to my anonymous source, “The firm that has probably made more money from DC news services than anyone else is Virtu (formerly Madison Taylor and EWT),” which is now in the process of going public.
“I *think* Virtu was/is a NTKN client,” wrote my source in an e-mail. They were the early adopters on the EIA numbers (oil and nat gas inventories) and less so on the macro economic data. They also put these strategies on FPGA cards to bypass the OS which shaved microseconds of reaction time. The telco is only one part of the latency and if all HFT clients get it at the same time the one who processes it first wins.”
HFT-related microwave operations are also under way in Europe too where my source says that Jump is rumoured to have bought a de-commisioned NATO telco tower in Belgium to secure the fastest London to Frankfurt route. He also said:
London-Frankfurt has three commercial providers (that I know of): Colt, Perseus, Fixnetix. CIC was recently trying to sell various assets that they’d put together. In Europe, the regulatory structure for link licensing is byzantine, and you have to deal with UK/Ofcom, France/Belgium, and Germany. I suspect that several household-name HFTs are already running their own routes. (Jump, Virtu, and Final probably. Tower isn’t there yet but will be. Allston and Getco are doubtful) Not sure about the UK coast where the trans-atlantic cable terminates to London or London to Stockholm which all will be done if it isn’t already.
I can’t say that this is all clear to me, but for those of you that follow HFT and microwave networks maybe you’ll find something of interest here. If you do, I’d love to know about it. Please contact me at email@example.com.
Oh, and if you want more on the telco aspect of all this, this Chicago Tribune article is detailed and interesting. Finally, here’s something I wrote on HFT and lock-ups.
46 thoughts on “The mystery ending of Michael Lewis’ Flash Boys: FCC License No. 1215095”
nice and timely post. seems like a lot of energy has been brought out from Michael’s book. Hope your HFT story is equally compelling.
I hope so too 🙂
Well done. Only comment I find on this. Guess Lewis is hinting at exactly the story you are mentioning.
Thank you. I’m really not sure what he’s hinting at.
Umm…he uses the word “hypocrisy” and he had been detailing Goldman Sachs apparent experience on the road to Damascus…isn’t he saying the owner of FCC License No. 1215095 is on the public record?
The owner of FCC License No. 1215095 is Converge Towers LLC, which is tied to Cantor Fitzgerald. That’s very easy to find.
I just don’t get where it’s all supposed to go from there.
And Lewis does say, it would take a day in cyberspace as opposed to five minutes Googling.
Only it’s not tied to Cantor Fitzgerald. It’s a shell company affiliated with Epsilon Networks.
Epsilon Networks is a partnership between BCG Partners and Thesys Technologies. Note the address listed for Converge Towers LLC matches that of Thesys.
Thesys a technology infrastructure provider partnered with Bank of America Merrilll Lynch, is an offshoot of HFT firm Tradeworx. The contact listed in the FCC database, Arzhang Kamarei, is listed as the President at Tradeworx on Linkedin.
Tradeworx won the contract to develop data analysis tools for the SEC (MIDAS, or Market Information Data Analytics System).
[The next entry listed on the above captioned website, Wireless Internetwork, LLC is another finance based microwave network shell company. It’s related to Strike Holdings, an offshoot of HFT firm GTS. There’s a bank related with that network as well…]
See, this is where things get murky. Can you explain why Converge Towers’ web domain is registered by Cantor? Simple error — perhaps based on the fact that BCG used to be part of Cantor?
BTW, thanks for your comment and info. 🙂
Also, Converge Towers has a listed address in Carteret, NJ. The same address as the NASDAQ exchange.
“used to be part of Cantor”? Still is really: Howard Lutnick is the Chairman and CEO of both Cantor and BGC. Cantor owns 23% of BGC stock. Another 34% of BGC stock is owned by “Employees, Executives and Directors” which I’d guess is a lot of current and or former Cantor employees, executives and directors.
Interesting. These things are almost always far more complicated than they appear at first. I guess that is why Michael Lewis says you’ll need a day in Cyberspace as opposed to five-minute Google search. Thx for your info.
Maybe this helps http://wireless2.fcc.gov/UlsApp/AsrSearch/asrRegistration.jsp?regKey=2612600
Before I actually put in a call the FCC, does anyone know if 1215095 is the licence for the tower itself or for one of the microwave devices that has been placed on it? My understanding is that each towers has several microwave devices that all belong to different companies/people.
It is pretty easy to figure out from public source data. According to the FCC website, the application for the microwave device was filed by Converge Towers LLC. The entity is located at 770 Broadway, Second Floor. It is not registered with the NY DOS, but the contact person listed on the application is firstname.lastname@example.org. If you google “Elizabeth Kim 770 Broadway” you will find that she works at Thesys Technologies (address, you guessed it, 770 Broadway, 2d Floor). According tot the Thesys website, they are the “infrastructure affiliate” or Tradeworks, the Red Bank, New Jersey HFT operation that was engaged in 2012 to work on Midas by the SEC. The NY Times noted in its October 2012 article that Tradeworks was involved in “building towers that can beam trading data to Chicago via microwave, a faster method of transmission than fiber optic cables.” Problem solved. http://www.nytimes.com/2012/10/08/business/sec-regulators-turn-to-high-speed-trading-firm.html?pagewanted=all&_r=0
I don’t wish to appear ungrateful, but I’m not seeing how the problem is solved.
At the most basic level, I still don’t know if FCC 1215095 is the licence for the tower or a microwave device that has been placed on the tower.
And bigger picture, what is the “story” here? All we seem to have is a collection of facts that don’t fit together. Or is it just me who can’t see how they fit together?
It’s for the tower. The microwave device number can be found by searching the tower location in the application database on the FCC website. Lewis was possibly suggesting that it is hypocritical for a company assisting the SEC by making its watchdog program (Midas) while at the same time installing a Chicago to Jersey microwave link to game the system via HFT methods. I don’t know, but as for who owns the microwave amplifier (and many others straight to Illinois), the problem’s solved. (They also do have links to Cantor, but I do not think that is what Lewis was getting at).
Manoj Narang (Tradeworx) has featured largely in the Flash Boys “debate”, stone-cold challenging Lewis on everything and strongly praising the SEC’s surveillance capability without revealing his role with MIDAS. Hunsader claims that the SEC does not/can not delve into sub-second trades…and Manoj trades sub microsecond and maintains it’s all good, the SEC is on it. At the same time Manoj repeatedly acknowledges the whole HFT industry trades off direct feeds against the consolidated SIP price feed which, according to Hunsader, is a violation of Reg NMS. It’s all hiding in plain sight. It’s weird.
The righteous indignation of these HFT guys is ridiculous. Imagine that eBay had a buyers and sellers posting all day long on two servers – a “buyer” server and a “seller” server. Rather than post buyers against sellers simultaneously, eBay’s business model was to pipe the buy and sell side to a third party (“ScalpCo”) for a split second first (ScalpCo paid for this special access). Whenever ScalpCo’s computer algo saw a buyer for antique doylies at 5 dollars and a seller for 4 dolllars, it would buy from seller at 4 and sell to buyer at 5. Seller would get rid of his doylie, and buyer would get his doylie, but they would never know each other existed. ScalpCo, of course, provided no service to either – eBay could have matched their bids – and it certainly provided no liquidity in doylies (there was only one doylie – the one seller was trying to sell). ScalpCo profits by a buck for doing nothing, pays a commission to eBay, and both seller and buyer pay for it.
HFT, in this form, makes money only because it is allowed a speed advantage by the system and it’s participants. It makes money even though buyer and seller would match naturally at a price that both would prefer. Such a system should not exist. It is little more than a tax on capital, it is economically inefficient, and it rewards economically meaningless activity.
Ps, a smart (and unscrupulous) HFT creating a surveilance system for the SEC like Midas would measure trades in nanoseconds,then create and sell systems that trade just one millisecond shy of a nanosecond. Who better to fly under the radar than the guy who made it blind at 100 feet!
In my post above, I meant “measure Midas in milliseconds, trade a nanosecond faster.”
Is the story that the SEC contract for MIDAS is the funding to build an HFT/prop shop’s own infrastructure? that just so happens to be state of the art. perhaps not illegal, but it would be an odd arrangement.
I’m with Doug although I’m a complete layman when it comes to this stuff. One of the many interviews I watched this week was with Lewis, Katsayama (sp?), and some guy named Manoj (from Tradeworx) on Bloomberg. They got into a heated discussion with Manoj taking issue with the claims in the book. At one point Brad says to Manoj – then why are putting up microwave towers?
Hmm. That’s interesting. Some of the critiques of Lewis are that he’s late to the story, that HFT players have lost their big advantage, that chasing speed has become a war of diminishing returns, and that this whole thing is just not an issue any more. My response to that — since I first heard about the microwave networks — has always been, then why are these HFTs all building microwave networks? Sounds like that’s what Katsuyama was asking too.
Wait until they have an operative quantum computer (zero latency over any distance using quantum “action at a distance”). The investing world will be divided into those with and those without. It will make microwave amplification look like grandpa driving down the 405 on Sunday. People should make money in the market by picking good companies, not picking off prices. There is no public good in what’s being done. Don’t be fooled.
I think a lot of the anti-Lewis pro HFT “debate” is deliberately mis-informed, when engaged very few have read the book, they “know it”. This link to a Wapo interview with Eric Hunsader is a good antidote for those interested:
The ending of Flash Boys is one of the saddest bits of the Lewis book. It was sad for journalism, unless you support deceptive journalism.
Mr Lewis goes to great pains to weave a narrative around Spread Networks and then leaves with the New New Thing of microwave at the conclusion of the book with clear sinister overtones. Good people, like those on this page have been hunting down who could it be doing this New New Thing that was the evolution of speed after Spread Networks?
Surely it would have been responsible journalism for Lewis, especially when writing on the scale of a book, to “google” the topic so he could have found out that his narrative was wrong and RF microwave had been used on the Illinois – New Jersey link since 2009, BEFORE spread networks.
It took me less than thirty seconds to get a reference. Article from the Chicago Tribune from 2012:
Shoddy journalism from Mr Lewis but I guess he didn’t want any facts to get in the way of a good story.
Barksdale is the force behind Spread Networks and an investor in IEX and an acquaintance of Lewis from the New New Thing days.
Also, why is Tradeworx copping flak for this? It was been public for sometime. Wired.com even had a big coloured box with the Tradeworx name on it in 2012 which even included the price:
In the words of the great Lleyton Hewitt, “Come on!”
There is no news here, just scuttlebutt and harmful innuendo.
Here is my review of Flash Boys: http://meanderful.blogspot.com.au/2014/04/flash-boys-misleading-information.html
Making money every day means the market is rigged? It’s not rigged! It’s just math!
Should mom and pop investors really be sold this dangerous Lewis misinformation? What harm is he doing to confidence in the US equity National Market System? The book should be reclassified as fiction or withdrawn from sale?
As far as I can tell, microwave became a big thing in the summer of 2012. That Chicago Tribune article, which I linked in my post and quoted from in a previous post (last year) on microwave, states that earlier microwave networks were unreliable and could be affected by bad weather.
What’s more, Lewis discusses the problems with earlier microwave transmissions on page 267 of Flash Boys so I don’t really see any “shoddy journalism” there.
The ending of the book is indeed unusual.
Maybe you’re right and I’m being a bit harsh on Lewis.
I believe the spread networks versus RF story was much more complex than his simple, convenient narrative. The good work from people like yourself has helped uncover that.
I feel similarly about his review of the May 2010 flash crash FWIW and many other parts of the book, such as dates for PFOF, maker taker, etc.
Perhaps ask some of the buyers of Spread Networks how they felt when they realised they’d locked in a second best? I don’t think it undermines Lewis’ thesis of special tech for traders at all. It just makes his story a bit shaky considering his emphasis which casts doubt elsewhere.
I was in and around that space at the time. RF was a big deal. It was an especially big deal for those who were not necessarily the happiest to find out their Spread Network contracts gave them second best. It would be interesting to hear some of their comments.
I did enjoy reading his book. I just found it quite misrepresentative. Imagine if the book had started with the RF solution which mainly worked in 2009 and Spread Networks decided to build a slower network that was more suitable to some investors. That would not be such a great story, but it didn’t happen that way either.
I disagree that the book misrepresented things. Just because it’s not the way you would tell a story doesn’t make it wrong.
Opening Flash BOys with the building of the Spread Networks line set the scene very clearly. Readers leave that chapter with an understanding of how important speed has become to traders and just how secretive they can be.
Honestly, I have tried for years to explain to others how trading is no longer about minutes and seconds but milliseconds and failed dismally. Lewis’s narrative device did the trick.
Now, although I am a big fan of Lewis’s work, I don’t by any means think it’s perfect and I have some major issues with The Big Short. That said, I think most of your criticisms are misplaced and it takes major chutzpah to tell a master storyteller he started his book the wrong way.
It is a tax on capital. HFT provides no service. By definition, it matches buyers and sellers at a price that is worse for both, when the exchange would take care of it at a better price a millisecond later. Risk less profits. This aspect of HFT (time based arbitrage) should be banned. They can do whatever else they want after that.
It is not a tax. Taxes go to the government from which the taxpayer get something in return.
HFT is pure larceny.
The important takeaway from the book is that brokers and infrastructure providers (exchanges etc) are actively supporting increased trading volume (they get paid by volume) even though the providers of some of that volume are able to front-run the brokers’ clients. Any other points or questions are a distraction. What must be asked next is whether such activity is a) illegal, b) legal but not permissible, C) fine and dandy.
I would ask whether releasing data at the same time, knowing that it will reach one person faster, is the same as giving the data to one person before the other.
There’s a lot about this whole thing in the paper http://arxiv.org/abs/1302.5966
Just finished the book…
1. left us hanging on microwave tower and – – without a link to GS the twist makes no sense…
2. left us hanging on Sergey Aleynikov but I guess that IS still hanging
(and why the new prosecution) (any relation to GS?)…
3. and left us hanging on GS feeding trades to IEX… GS said we’ll be big tomorrow – – and that got left unresolved, unanswered, unfinished…
often good / great writer but seems like he had a few drinks toward the end and then didn’t have the time or energy to finish…
high frequency book publishing leaves something to be desired too.
On page 258 of Lewis’s book it explains that the prosecutor Joanne Li left for a job at Citigroup. I am assuming that the arrest of Serge was to parlay experience with HFT to a higher paying job for the prosecutor at a Wall Street bank or HFT. It seems to have worked for Li. It works for a lot of lawyers that sue or investigate Wall Street banks. SEC employees do this all the time, also. They start investigating a company, the next thing you know they have a nice, high paying job at an HFT or Wall Street bank, and then the investigation gets shut down. Either the investigation get shut down by SEC, and the investigator gets frustrated, and leaves or leaves to shut the investigation. Either way the SEC and private industry has become a revolving door.
Microwave is apparently too slow now. Laser networks are being built by Aoptix/Anova.
Lasers and microwaves are essentially the same speed.
Right, so does fiber optic (all signals travel at the speed of light). Fiber has to bend with the land so it isn’t straight and is therefore slower. Laser and micro both deal with that issue, but I think the advantage of laser over microwave is higher bandwidth, less weather interference, and fewer fixed relays.
The speed of light is slower in glass than air due to its higher refractive index (think of it like swimming through jelly instead of water). The bends in a fibre then slow it down even more!
What I really think ? This story is an increasingly old story about who is f…ing whom ? After the banks screwing the retail investor for decades with costly and murky order management, GLOBAL INCOMPETENCE has brought us the ” f.ck the customer” business model ( hi Goldman) to the institutional level.
“Get rich quickly” at any cost is going to leave nobody alive.
Am I a dinosaur ? Sure I am, and happy to be, trader with a real social role, my word is my bound, easing the deal, making the real economic value and information available to the public, eye to eye, cards on the table, man to man, a deal is a deal. Proud of it. Sounds like what RBC business was about. Sounds like something Warren Buffett would understand.
If you want a livable world for your children, it has to turn around, that’s what Michael Lewis book is about.
Love that champagne, by the way……….
Putting my Electrical Engineering degree to use for the first time in a decade:
Lasers/light and microwaves are essentially the same thing. The only difference between microwave signals and light signals is the frequency/wavelength. Light is just EM (electromagnetic) radiation at a frequency/wavelength that our eyes have adapted to. Microwaves are EM radiation at another frequency. Ditto for UV/X-rays, TV/radio, etc. Cellular phone networks are microwave signals essentially.
EM radiation travels at the speed of light, no matter what “form” that radiation travels at. The only difference between wireless communications and wire-bound communications is the medium. Microwaves can travel on wires or through the air. So the distinction here isn’t light vs. microwave, but rather wires vs. wireless.
Wireless signals have the advantage of being able to travel the shortest possible distance, while wire bound signals must travel that path. I’ll argue that it is effectively impossible to drop a wire between two points as straight (shortest distance) as a direct wireless signal. (Especially if the two points are 1000 miles away.)
Not only that, a light/laser traveling in a wire doesn’t travel in the wire/fiber itself in a straight-line. It “bounces” or refracts along the fiber. A little trigonometry will tell you that a signal traveling 1000 miles that bounces off the “walls” of the fiber is going to travel more than 1000 miles total. It travels mostly diagonally. Not to mention, light signals degrade and need to be repeated/re-amplified periodically.
The problem with wireless can be determined by anyone with a cell phone. Wireless signals are much more susceptible to noise and interference. Weather yes, but mostly from rain and thunderstorms, but rain fade only really effects broadband signals at certain frequencies and lightning is fairly random and unperiodic. You can design to protect from weather. And the FCC’s primary purpose is to dole out spectrum to minimize interference.
The speed of light varies according to the medium it travels in. Fastest transmission velocity is in free space (vacuum). Light traveling through glass has a measurably slower velocity than free space (though the change in velocity is not great). Having written the above – microwave repeaters also cause a delay. And you need more microwave repeaters than fiber glass repeaters. So the actual minimal time from Chicago to NY is probably close to a wash.
book is Gr8. uncovers theft is theft. I feel ripped off. opaqueness of the trading process involving “extra unnecessary trades” that add no value to seller or buyer but simply requires new extra rules to allow intermediaries to pick off the round-up errors is systemic theft. “Gordon Getco” lives!!! This is an example of the very first computer crime. Google it up. The risk to the market due to the new chaotic input of super-short latency inclusion of unnecessary friction has been documented. The SEC is one of the most ineffective regulatory agencies, full stop. Collusion and incompetency abound there. “let the buyer beware” even if there is not a damn thing u can do about it. Adam Smith spins in his grave over the corruption of the “Capitalist Ideal”.
Argue with the author’s opinion, ok. Argue with the facts, liar, liar pants on FIRE.
Look for the next “Bubble”/market plunge. I hear they are using brokered “securitized sub-prime loans” for used car salesmen to sell cars to dead people; nuthin new here! Greed, not profit seeking, will be our collective undoing.
Obviously coming late to the party. Just read the book the other day. Lewis curiously steered clear of capital gains taxes in his book “Flash Boys” and that’s likely where the solution to this entire lash-up lies. It doesn’t require a major rethink of the tax structure: just some additional lines to existing law. Basically, the law regarding long-term and short-term ownership. for capital gains needs to be extended.
WHAT IS NEEDED is a provision that says if you have a position in a stock for even a nano-second and make a profit from that position, you need to pay a capital gains tax. And this gets into dark pools which are basically cheating the living bejeezus out of existing tax law by not reporting instant positions: only end of day reporting. Furthermore, there needs to be a difference between short-term capital gains, and REALLY short-term capital gains (i.e. positions of less than a second or so). And with REALLY short-term capital gains you level the playing field.
It’s what money parasites understand the best: the cost/profit bottom line. If you price the financial friction (the REALLY short-term capital gains tax) correct, you essentially eliminate the latency arbitrage advantage.
FURTHERMORE, if HFT adjust their game to only make trades when an existing trade-partner is already locked-in (i.e. a lateral transfer so no positional ownership occurs), the HFT is essentially providing a true electronic liquidity (i.e. an exchange function), and they are required to operate under the laws of exchanges.
It’s not a wash, in that the microwave links from Chicago to NJ are a full 3ms faster than the best fibre. It’s very low bandwidth, and unreliable, but HFT is a winner-takes-all game and the weather affects everyone. When the microwave links are work, they are enormously faster, and so anyone using fibre loses outright.
But this is all old, old news.
Ya, I was surprised when I looked up the transmission speed of light in fiber was something like 30% slower than air. It’s countered somewhat by the need for more repeaters in the microwave system but still, microwave should be substantially faster than fiber over that distance. Thanks for quantification of 3 ms.
Two things I really started to take from the book–and I hope someone will correct me if I’m wrong (I’m pretty smart, but this stuff is hella complex):
(1) What Lewis has outlined is people doing something immoral, but legal. Isn’t this the same stealing the fraction of cents issue that was presented in both the movies Office Space, and one of the Superman movies?
(2) It is only a matter of time before some of the firms figure out ways to game one another. If and when this happens it could mean financial disaster for all the markets. The fact that it appears to not have happened yet seems like a bit of collusion.
Lewis outlines the revolving door between the HFT companies and the SEC. This mirrors stories in other regulatory agencies like the FDA, and some of Obama’s appointments to his cabinet. (Please don’t think that because I mention Obama’s name that I am bashing him any more than any politician.)
I was involved in the Lind-Waldock MF Global fiasco where my 5 figured cash account at MF Global disappeared overnight and I was forced to wait for some 3 or so years while it trickled back to me. Following the proceedings of that case uncovered an equally complex system that was also handled poorly by the government.
The bottom line to me is that people are looking far too often now to earning money by “taking it from someone else” rather than “providing a service for a profit or fee:.