HFT using neutrino physics: Stats Jackassery | Hacker News Reader

HFT using neutrino physics: Stats Jackassery

scottlocklin.wordpress.com

34 pointsmikexstudios15 comments

At best you could gain a 24.5ms advantage (through the earth to other side @ 12,800km versus around it at pi*6,400km).Actually quite significant in HFT terms.

But: sady this is completely impossible. The weak force is not called the weak force for no reason. Neutrinos hardly interact with matter at all - that's why they can stream through 12,800 km of solid iron almost unimpeded. You would need a detector the size of the solar system in order to collect any meaningful data volume in real-time.

> You would need a detector the size of the solar system in order to collect any meaningful data volume in real-time.

That's absolutely untrue. Why would you need to detect all neutrinos? Even if you were capable of detecting only a quadrillionth of all neutrinos sent your way, I could send 10^20 neutrinos per second your way.

Although this project would be extremely expensive, it's absolutely possible with today's technology.

Impossible is a strong word.

Assuming the radius of the earth is 6378.1 Km (it varies), the distance traveled over the circumference of half the earth will be 20037.375179 Km, compared to 12756.2 Km traveling straight through the earth. The difference is 7280.60115 Km. So the best possible case is a time saving of 7280601.15 m / 299792458 m/s = 24.285 mS (what parent said).

Expanding on that a little:

Using more realistic values, the distance between New York City and London is 5567 Km, while the path through the earth is 5392 Km (see http://imgur.com/Th6jW). So this would save 175000 m / 299792458 m/s = 0.583 mS

On the other hand, Spain and New Zealand are pretty much directly opposite (http://www.freemaptools.com/tunnel-to-other-side-of-the-eart...) each other, and Europe to Japan is pretty close to being on directly opposite sides of the earth.

So London to New York might be cost effective at $25 million, assuming the same $300M/6mS (http://www.telegraph.co.uk/technology/news/8753784/The-300m-...) rate, while best case (Spain to New Zealand) might be worth $1.2 billion.

Had to brush up on my geometry: http://www.mathopenref.com/arclength.html http://en.wikipedia.org/wiki/Law_of_cosines

Already implemented by the Rochester guys (well the messaging part, not the HFT part)

http://news.ycombinator.com/item?id=3705953

Its an interesting application of the potential.

Have people forgottten satellites? You don't have to go along the Earth's crust, you know!

A low orbit satellite with height above surface <= r * (0.5 * theta * cosec (theta /2) - 1)

where r = radius of earth and theta = angle subtended by the arc between NY and London at the earth's center

will always transmit signals faster than a transatlantic cable. With actual values, the height comes out to be around 510 kms.

I remember a brief discussion of this on HN from about a year ago: http://news.ycombinator.com/item?id=2425228

That's a really interesting idea. Does anyone know how much money you could make with HFT if you were able to communicate through the earth instead of around it?

If the profit potential is high enough, maybe this is a good way to fund research?

More than $300 million.

http://www.telegraph.co.uk/technology/news/8753784/The-300m-...

I'd doubt whether there'd be much money in it from an HFT perspective; traditional trading might benefit from it, however.

HFT shops co-locate, so they're running their code right there next to the exchange.

Edit: In case I've missed a point in the article, it wasn't intentional, it's banned at work so wasn't able to read it.

This would be true if there were only one exchange. Instead, there are many around the world, and information provided by one can influence another.

The use case for this would be communication between different colocations, which is where the networking "race" really is for HFT. For example, to arb products on the CME vs. the equities exchanges in the New York area (really colocated in various places in NJ).

At any rate, as the author mentions this isn't really a feasible technology.

I was at the R/Finance 2012 conference in Chicago about a month ago and one of the speakers commented that about "1 millisecond was worth $10,000," but I forget whether that was an annual number as in "1 millisecond saved is $10,000 a year" or "We can make $10,000 every millisecond." Hopefully someone else can corroborate this.

it's not 10,000 a millisecond. Trading desks normally think in terms of daily PnL - he prob meant every 1ms latency reduction would make them an extra 10k a day.

Yes, now that you mention it, that sounds right.

Reply on news.ycombinator.com