The $300m cable that will save traders milliseconds

The $300m cable that will save traders milliseconds → … conds.html

How long before they realise that a great circle is NOT the shortest point between two locations on earth. That would be a tunnel straight through. For this particular route, the tunnel would be 5,794km, would reach a maximum depth of 707km, and would shave 227km off the great circle distance, and nearly three quarters of a millisecond off the light-speed journey. Must be worth it to somebody eventually. 8DD

Or they could just ban this high frequency trading nonsense. :smiling_imp:

Mathematicians have been wrestling with this problem for a looonnnng time.
Some people might find the lecture by Gregory Chaitin from the year 2000 interesting.

(start 4 mins in)

pt 3:

pt 4:

pt 5:

pt 6:

pt 7:

pt 8:

Indeed - and by digging the tunnel really straight, sealing it really well and sending the signal in a vacuum rather than by fibre, you’d drop the lag another 30% (almost 10ms).

On second thoughts, keeping a tunnel straight at 700km depth might be a challenge. How about neutrino-based communication? You’d need some sort of free neutron generator at one end, and a billion gallons of carpet cleaner at the other. No tunnel required.

Just one of many reasons why investing based off ‘fundamentals’ has been a very frustrating pursuit for years.
Its all about finding a ‘process’ that works and will pay-off over a very short time frame…again and again.
Buying stock X because it is ‘cheap’ relative to Y is a lost cause, because there are 10s of 1,000s of people who can see the ‘obvious’ anyway
Its a sad reflection of where we are today, but its also reality. For now anyway!

Fermilab already doing something along those lines.

  1. The feckin fibre would melt at 10km depth never mind 700km down. :nin . You’d need DCF every couple of feet.
  2. A fibre through the earths crust was proposed in 2010 between New York and Chicago which would ‘only be 30km deep’ .
  3. Hibernia will take a bath on that cable if they really gave the guarantee averred to by their salesman. “If it ever gets beaten for speed we end up giving our customers their money back”. That is because it already HAS been beaten for latency by THIS other cable which is on a more circular great circle than Hibernias weak circling effort is.

There is a rumour abroad of a third cable in the offing which will bury Hibernia whose proposed cable route along the seabed south of Cork is much too far south of optimal for a genuine great circle play.

Why not use HF radio!? It’s a bit cheaper… :wink:

Bandwidth and excessive jitter caused by these lads. I could see a move to an ulta low latency coax architecture for the next cable. :slight_smile:

So assuming this is not just PR hype to bamboozle the investors in the cable when is this suppose to go live? 2013, I see. With a bit of luck HTF/LLT will be illegal by then. By this stage LLT is just a cover for front-running. So little different from the big insider scams of the late 1920’s.

The only reason for a low latency cable for trading is so that the current in-exchange co-location front running scam can be extended to inter exchange transactions. As the big boys have all gone dark pool or else pulled out completely its just the clueless and the small timers left to be fleeced.

Low latency is all about trading intermarket across correlated assets, ie spreading - if the FESX tanks sell ES/ buy Nestle sell Kraft etc. Getting your order off or even just getting to the front of the queue before the other guy is not the same as front running (how can it be when it’s often based on traded prices)

If you’re getting “fleeced” you need to change your broker or just take off your tinfoil hat

They’re not spending millions to take a few cent from you when you place your thousand lot order on Sony stock.

Not from what I hear out of NYC.

Remember, I not only know how to build the models, I know how to write the software…

With every post you just reinforce the impression that you are a STT. And you obviously dont know enough to know who the mark is…which means…

Why does the phrase picking up pennies in front of a steamroller comes to mind.

Just because I know something about STT doesn’t mean I am one. [edit] (I’d love to hear your rational for drawing that conclusion, for a man who likes to talk about making evidence based decisions you do talk through your hole a fair bit)

How can you claim to know how to build the models? On the Apple thread you claimed that HFT was driving up the share price: it’s like confusing the frequency of a signal with the power of a signal!

As for picking up pennies/being a mark blah blah leave the trite cliches somewhere else. I’ll read the Irish Times letters page if I want ill-informed hand wringing

looks like ethernet is determining the next standard instead of transmission standards :angry:

What, are we going ‘back’ to SDH or summat ? That has nothing to do with STT/HFT or whatever.

That’s dashed clever of them. Don’t neutrinos have a half-thickness of something like a light year of lead? I was joking about the billion gallons of carpet cleaner. It would be too little by many orders of magnitude, even assuming the only challenge was neutrino detection.

So they’re all jumping on board this thing, and no-one will benefit except the cable owners?

That neutrino cable replacement idea will have new legs if this FTL neutrino experiment at CERN turns our to be valid. They’ll be able to watch a stock movement and then make a trade before it happens. Now that’s worth a lot more than a piffling $300m. :laughing: