https://www.radioworld.com/tech-and-gear/market-makers-want-to-expand-their-use-of-shortwave
The FCC seeks comments on a proposal to use HF spectrum for financial data
From a technology stand point this is intriguing. But that much said, I am anti-capitalist. The one downside to HF radio is that the available bandwidth is tiny. You'd have to create a compression algorith capable of compressing enough data to complete a trade. Also HF is very susceptible to changing atmospheric conditions. I am a licensed ham radio operator.
They aren’t using it to complete a trade across it. You use it to send information about what is happening in, say, NY to trade locally in Tokyo or São Paulo before anyone else knows what happened in NYC and adjusts their own trading strategy.
You basically are arbitraging people’s access to information.
They’ve been doing this from NY to Chicago (using higher throughput) radio links for a while. See flash boys.
But if you have sophisticated algorithms moving fast enough in NY to only send the information that is worth trading on, this sounds compelling.
I tend to have a dim view of HFTs - I think they’re more or less just stealing pennies from lots of folks while pretending that’s liquidity, but I’m not an expert.
I am also a licensed ham, and you are right. HF digital modes are very low data rate. A low bits/sec data rate because of the limited bandwidth will result in high latency for any nontrivial message length. FT8 is only 6b/sec for comparison. I'm curious to learn more about how they would use the spectrum.
Also a licensed ham here (though no real HF experience). @lchapman has a good point about the latency being only a fifth, but HF is pretty low data rate still. The ionosphere is not ideal, so error correction would need to be put in place, further lowering the datarate. The only thing I can think of is that they would do something with more bandwidth maybe? But why? So some fat cat can trade at just slightly faster than the competition? This seems irresponsible.
I've yet to play with FT8. I've had to sell my radio equipment when I became disabled. One day, I'll get back into it again.
As a capitalist, this is problematic. Spectrum is a limited and common resource. It would drive the price of spectrum way above what the average organization could afford while providing no benefit over a fiber connection. Meanwhile those organizations that could use the spectrum are unable to.
It's not even particularly innovative because networking over HF already exists. Sure, there is some interesting profiling you could do but it's not particularly new.
1800 symbols per second is the benchmark for shortwave data transmission.
https://en.m.wikipedia.org/wiki/PACTOR
Not sure exactly how useful that would be, but the latency is low: 0.01 seconds to cross the width of the USA.
A round trip packet from NY to SF takes 0.05 seconds, a fifth of the speed. Fibre is quick, but not as quick as radio.
There’s about 1400 bytes usable in a TCP packet, versus the 1800 symbols per second over shortwave. Lots of TCP packets can be exchanged per second.
I don’t see the value proposition.
That’s plenty of data to trade on. If you believe in efficient markets the priced should be more or less the same globally, you can make a lot of money if you know what the market price in Tokyo should be because you know the price in NY first, or vice versa.
Pfffttt, they aren't thinking grand enough. Maybe if the use UHF they could then trade at a UHF frequency of trading.