This page has been archived and commenting is disabled.
Growing Concerns Over Tomorrow’s Leap Second
Tomorrow at 8pm, we’re adding an extra second to the day. Over the past 200 years, the length of a day has increased by two milliseconds, which is all well and good, but the insane accuracy of the atomic clocks we’ve been using since 1967 doesn’t account for that, so we’ve had to add leap seconds 25 times since 1972. Tomorrow, however, is the first time a leap second will be added during trading hours since markets went electronic.
Naturally, this has some experts concerned with estimates that around 10 percent of large scale computer networks will experience issues that stem from the added second. It happened to Reddit, Mozilla, and Qantas Airways back in 2012. To account for any issues that could occur tomorrow, evening trading will close five minutes early – Nasdaq will stop trading at 7:48 EST, and shut down at 7:55. Any data transmitted five minutes before and after the second will be held for another five minutes. Trading in Asia and Australia will start one second after the leap, but they also plan to smear the leap over the course of two hours by splitting it up into 7,200 pieces.
According to the Financial Information Forum and World Federation of Exchanges, in one second, 1.4 million order messages are sent to U.S. equity-trading venues, $4.6 in stocks are traded around the world, and $3.7 billion of stocks change hands at the bell in Korea, Japan and Australia. Couple these staggering numbers with the expiration of Greece’s bailout, and tomorrow could prove to be a very interesting day.
[Original]
»
- EquityNet's blog
- 26281 reads
- Printer-friendly version
- Send to friend
- advertisements -
We define a day as the time the earth takes to rotate 360 degrees about its axis of rotation. But due to tidal friction (mostly caused by the gravitational pull of the moon but also the sun) energy is dissipated slowing down the spin a little each year. By adding time back into our clocks with can maintain this 24 hour daily period illusion. Otherwise, we'd have to have say a 24.001 hour day tomorrow, 24.002 the next day etc. Better to just incorrectly define it as 24 hours and then add in the leap second when it finally accumulates enough tidal drag to slow the rotation down enough. The orbit of the earth around the sun could also be affected by other planets and their gravitational pull, but there's no dissipation mechanism to lose orbital angular momentum (if the earth weren't a waterworld, the dissipation of its spin angular momentum would also be reduced vastly (see Mars for example)). Conservation of angular momentum is always true in the absence of an energy dissipation mechanism, as is linear momentum.
This mechanism wasn't mentioned in the article or any of the comments, which is rather unfortunate.
I could give a flying leap fuck if they all crash and burn in their extra second of thievery.
When time was invented.......all men became slaves.
Imagine the man who invented the clock. The next morning when he arrived at work his boss said, "You're late!"
Quite the accomplishment, being able to be derailed by a single second. The Mayans must be laughing their asses off in Xilbalba.
Did you really think that a leap day every four years was exact? They have been doing leap seconds for many years.
I really, really, really hope this f?cks with the HFT's
don't do anything that relies on GPS during that time
Are they going to just ignore this in Arizona?
we turn clocks back and forward for an hour in daylight saving
So - tomorrow I'll become not just a day older but a day and a leap second older. Just freaking great.
I don't believe for a second that any of this matters.
I second that.
Heeeey, wait a se..........., oh, never mind.
"but the insane accuracy of the atomic clocks we’ve been using since 1967 doesn’t account for that"
That sentance needs some work. If they were so accurate, why do they need to be adjusted?
The clocks are accurate in keeping precise time in the form of equal sized intervals. The problem is that the planet Earth is not consistent in the time it takes to orbit the sun.
In the past the systems for measuring time (or more accurately, accounting for time) were NOT as accurate - when the Julian calendar was replaced by the Gregorian one, there was a ten day difference (the Julian calendar had added too many days for leap years since the actual orbit of the Earth takes slightly less than 365.25 days). Clocks for measuring time over the course of a day were often quite inaccurate - a serious matter when clocks were needed for navigation in calculating longitude by comparing star or sun positions to charts for Greenwich England. This was the driving force behind the development of accurate clocks.
My point was to highlight a flaw in the logic. Yes, Time in today's society is defined by a fixed interval. The logic flaw is why does the clock need adjustment? Is it because it does not agree with the time it takes to revolve around the sun? If this is true, then time is defined as the time to revolve around the sun, not a fixed interval. So which is it? IIRC, the Juilian calander was based upon epochs (events) and not fixed intervals and there no time adjustments were needed.
A more accurate statement would have been "insanely consistent clocks."
Here's an outstanding movie on the development of accurate clocks:
http://www.imdb.com/title/tt0192263/
"In the 18th century, the only way to navigate accurately at sea was to follow a coastline all the way, which would not get you from Europe to the West Indies or the Americas. Observing the sun or stars would give you the latitude, but not the longitude unless done in conjunction with a clock that would keep time accurately at sea, and no such clock existed. After one too many maritime disasters due to navigational errors, the British Parliament set up a substantial prize for a way to find the longitude at sea. The film's main story is that of craftsman John Harrison: he built a clock that would do the job, what we would now call a marine chronometer. But the Board of Longitude was biased against this approach and claiming the prize was no simple matter. Told in parallel is the 20th century story of Rupert Gould, for whom the restoration of Harrison's clocks to working order became first a hobby, then an obsession that threatened to wreck his life."
Why not use GPS?
A Millennial
I got's me onna dem swuare waches dat tells me every thin fo' bout three hours.
And what percent of 1.4 million order messages are sent to U.S. equity-trading venues are then ILLEGALLY cancelled!
How will this extra second affect gay marriage, the confederate flag, and Bruce Jenner's cock? That's what I worry about.
I think Bruce Jenner's cock is out of our misery, once and for all. The other two most-important issues in the world you mentioned? We'll all have one more second in our lives to think about them. Spend it wisely.
"but the insane accuracy"
So we can measure time more accurately than the earth spinning around the sun which varies. So what makes it insane? Are you anti-science?
Insane is a perfectly acceptable descriptive, along with "awesome" and "totally" which both entered the vernacular around the same time.
Incidentally, since the topic here is accuracy:
The earth spins on its axis. It revolves around the sun.
http://tycho.usno.navy.mil/leapsec.html
don't you know that every scientist on the planet has been bought and corrupted???
That's right; all scientists have indeed been bought and corrupted, by the envirotard Hippies who have overwhelmingly outspent all the known corporate industrial interests combined. Because, it turns out, dirty smelly pot-smoking tree-hugging Kumbaya-singing Hippies are not just unimaginably wealthy, but also masters of organization and coordination, and are utterly ruthless into the bargain.
I know; it sounds crazy; but that appears to be the going explanation for why science keeps telling us we can't just do whatever we want without consequences. Which can't possibly be true, so we're back to the only explanation that fits.
I personally think High-Frequency "Trading" is one of the most goofiest notions ever perpetrated. If one must out-compete machines that always make money (why else would they exist?). That means I am playing cards with the deck stacked against me. What next?
Negative-Frequency "Trading" (NFT) of course. You can lose money before you ever even heard of Wall Street. Oh wait...
HFT is easy to defeat.
1. a fraction of a cent transaction charge on every posted bid/offer.
2. a rule that each quote remain open for two seconds before it can be withdrawn.
That would put an end to HFT. The fact that this hasn't been done tells you everything you need to know about this so-called market.
Yup, look what happens when someone can click a second faster than their algos!
A miniscule transaction tax would raise billions of dollars - and go a long way towards paying off national debt (bwahaaaaaa......) or providing more funds for gov to piss away....... But the main benefit would be limiting high frequency trading and all the games played.
The one thing I am still missing on ZH is a BS article about the correlation of chem-trails and stock prices ...
Would it have to be a BS article correlating chem-trails and stock prices? A non-BS one wouldn't do it for you?
You might enjoy this website, if you haven't already run across it:
http://www.tylervigen.com/spurious-correlations
Some very funny stuff there. Tradeable too, if you've got such a bent of mind.
For instance, over the past 3 years, precipitation in Hennepin County, MN (where I live) correlates extremely closely with online revenue on Black Friday.
You must have moved, Hennepin county isn't in SW MN.
-EastCentralMNguy
In years a Palin announces a bastard, the markets are down.
i wouldn't mind trying to make one with mom. don't want to succeed or anything but i think the practice would be nice.
The executive, legislative and judicial branches of government and every mother fucker associated should all be trapped in that "leap second" and claw at each other as they struggle to get out , for all eternity.
Nice article, and good point.
I wrote some GPS-interfacing software recently - gotta account for dem leap seconds, especially when your time is GPS-synced...
The proper thing to do when adjusting time is to adjust it using a monotonically increasing method - the Linux function adjtime takes care of this. Problem is, it can only adjust at 500 parts per million, i.e. 500 microseconds per second. To make a full second of adjustment takes ~33 minutes.
Now imagine a shitload of computers around the world, all of which run programs relying on microsecond-precision timestamping, jumping a whole second due to lack of planning for such jumps in software.
Just as long as none of those computers are connected to anyone's ICBMs.
Heartbeat pulse is 4 microseconds off? Must have been a nuclear strike. LAUNCH.. LAUNCH... LAUNCH...
NTP.
"Now imagine a shitload of computers around the world, all of which run programs relying on microsecond-precision timestamping, jumping a whole second due to lack of planning for such jumps in software."
Sh't, they wouldn't plan a false flag event at that 7 minutes (7:48-7:55) around "the second", would they?
Hey this is important stuff ..if you are a HFT..time is money...
In our new global world there is always a major market open, so there will never be a good time. Nice to know we ain't worried about the minor shit, like the EU collapsing......
Last I checked there was a stock market in Khartoum. Couldn't we do it then.
LeapSWAN!!!
Is this like Y2K, again?
Hopefully not. Last time that scare happened, they outsourced a bunch of shitty jobs to India...