5 ms·
High frequency traders will want the best connectivity humanely possible. It’s a zero sum game. If they can beat a competitor firm’s signals by the smallest fra
by dannyw 6y ago
High frequency traders will want the best connectivity humanely possible. It’s a zero sum game. If they can beat a competitor firm’s signals by the smallest fraction of a second, even a nanosecond, they reaps in the profits. So $100m for some better fibre? Go nuts!
- andrepd 6y agoEh, that's a really sad thing to optimise for, don't you think? Revealing of a broken incentive system. I'd prefer if HFT practices were banned.
- TomMarius 6y agoCan you be more specific about what is so broken about the need for efficient and fast trade?
- pms 6y agoVolatility is the main concern [1]. Making trading decisions based on microseconds is speculation. How can anyone or anything reach an understanding of something as complex as markets in this short time? Other concerns include unnecessary waste and unfairness [2]. The point is that this is a zero-sum game that makes us invest in infrastructure that will soon get obsolete once someone else builds a faster infrastructure. Fast trade isn't "efficient". Long-term stability is. [1] https://en.wikipedia.org/wiki/2010_flash_crash https://en.wikipedia.org/wiki/2010_flash_crash [2] https://en.wikipedia.org/wiki/Flash_Boys https://en.wikipedia.org/wiki/Flash_Boys
- xwdv 6y agoNo you wouldn’t. The development of HFT has created massive liquidity for markets and reduced bid ask spreads in a way that benefits retail investors more than past alternatives.
- nicoburns 6y agoDoes it benefit society though? I'd argue that increased liquidity in markets is a net negative. That means it's easier for investors to sell their assets quickly, which is exactly what you don't want if you want to encourage long-term investment.
- scottLobster 6y agoPeople buy and flip houses, despite the fact that it takes months. What timeline would you think would discourage short-term investment? Ultimately I'd rather place the responsibility for long/short term investing on the investor. The fact that most retail investors should be investing passively and long term is well known at this point, the issue is one of education. Don't dumb down the system when it's easy to make people stronger.
- nicoburns 6y ago> People buy and flip houses, despite the fact that it takes months. What timeline would you think would discourage short-term investment? Months would be pretty good. But I suspect that wouldn't fly in practice. Flipping houses is long-term investment, no? Even if the asset isn't held for a long time, long-term value is added. This is very different to speculative trading where no value is added. > The fact that most retail investors should be investing passively and long term is well known at this point Nope! The primary purpose of the stock markets isn't to generate income for the investors. The purpose is to effectively distribute capital to those business that can use if effectively. Passive investment does not help to achieve this. It reduces the "wisdom of the crowd" effect, because anyone investing passively is not contributing their wisdom. > Don't dumb down the system when it's easy to make people stronger. I have completely the opposite view on this. If there's an obviously correct way of doing something, then don't leave it to chance. Design the system so that people can't fail.
- Haga 6y agoEndstage capitalism.. Feudal Thiefdoms and recursive fleas.
- esmi 6y agoPersonally I waffle on this myself. My instinct tells me it’s bad but then I try to decide what trading cadence I would limit it to. Assume there is a CEO who wants to do something nefarious to manipulate the stock price. If they make a decision how fast can a company actually act on it? Would or could they make the company act on 1 hour boundaries? Even if we limited trades to once per week is that the long term we want a CEO to consider? At what time frame does a stock stop being a stock and convert to something else? Then I decide even by minute trading is not really something that management can even consider, even if they want to, so there is really no point to artificial limits imposed by law and take a more laissez faire attitude.
- ncmncm 6y agoDon't limit cadence. Just tax transactions, and cadence will limit itself.
- esmi 6y agoI waffle on this one too. Institutions that do HFT have a huge capital cost and pay large fees before they even make their first trade. They also employ programmers, traders, etc. Basically, these taxes are going to need to sum to something really big to dissuade them. Too small, they ignore it. Medium, the market will just consolidate. Large, you'll stop HFT, but at what consequence? The probability of unintended consequences is high. This is when I decide I don't really understand the market as well as I think, and I should stop solving the world's problems, and go back to designing circuits.
- ncmncm 6y agoA tiny tax on each order (bid, ask) entry would cut the total volume of orders by a huge factor, cost them only a small fraction of their takings, and probably fund free college for everyone in the US.
- gruturo 6y agoHigh frequency traders do NOT want to be limited by the speed of light in a fiber - due to its refractive index (1.5 for ZBLAN according to Wikipedia) this is only 66% of the speed of a radio signal. Over 10km this amounts to an 11 microseconds difference - already significant for HFT. This is why they have been leveling hills to carve a (mostly, with minimal amount of repeater towers) line of sight path between NYC and Chicago, gaining 3ms (1), or set up a line of sight laser link between NYSE and NASDAQ to eke out a handful of microseconds (2), or set up multiple links across the English channel to reduce latency between London, Amsterdam and Frankfurt (3). 1. https://www.nytimes.com/2014/04/14/opinion/krugman-three-expensive-milliseconds.html https://www.nytimes.com/2014/04/14/opinion/krugman-three-exp... 2. https://www.extremetech.com/extreme/176551-new-laser-network-between-nyse-and-nasdaq-will-allow-high-frequency-traders-to-make-even-more-money https://www.extremetech.com/extreme/176551-new-laser-network... 3. https://sniperinmahwah.wordpress.com/2016/01/26/hft-in-the-banana-land/ https://sniperinmahwah.wordpress.com/2016/01/26/hft-in-the-b...
- DennisP 6y agoI've seen speculation that low latency may be a big part of the business plan for Starlink.
- spullara 6y agoIt should be faster for long distances than cables for sure. Also less routing.
- LargoLasskhyfv 6y agoOne could use hollow core fiber, which has much higher limits. Up to 99% of light in a vacuum IIRC.
- tornato7 6y agoA wave guide?
- 6y ago