A T1 to the internet will shave 50ms per side off your latency and should improve the reliability of the connections. Your T1 provider should be able to get you on the internet at a major backbone. In that case, it'll be just another 15 or 20 ms to the IB servers. Once you are on the main connections, its very fast, takes few hops and the routes are stable (they don't change that much). Bottom line, I don't think a dedicated T1 connection to IB will be that much faster than just putting a T1 to the internet. They both have to cover the same routes. But a T1 to the net is alot cheaper.
That's all very nice if you're Victor Niederhoffer, doing one trade a day. If you're actually trading actively, like a scalper who reacts to the very moment of a bid/ask shrinking and needs to get an order in like, ideally yesterday, things look very, VERY different, and if you think you can identify the kinds of "bottoms" I trade 3 seconds in advance, then we're clearly doing different timeframes and R:R scales. To give you an idea, I did 173 RT yesterday. So we aren't on the same wavelength here. Cheers for the words, though.
500ms? Are you sure? How many times did you try? Maybe it's latency in your computer or sth? I just tried the same and averaged 167ms in 30 tries. While the odd men out were anything between 90-240ms, about 90% were 160-170, so I assume that's my "brain latency". Now, I've thought about this, thought "why is it that stock777 takes longer than me? Maybe because he isn't a martial artist? Oh, most probably he's just older than me. Isn't that a major reason why prop firms prefer hiring very young people, because they react very fast? Myself, I'm just in my early twens, so my reaction should be pretty fast? I wanted to verify this. I got my mum over yesterday (closest relative), to do the test. Oh man, my mum embarrassed me. She hit the button at just over 120ms very often, and 60-100ms a few times, too (I'm confident she was jumping the gun, but then the trigger distribution for the light going on is random, so it's unfair to say that unless you believe in clairvoyancy). Well, the typical range (80% of 30 tries) for my mum was 120-150ms, the average was about 143ms, a whopping 24ms faster than me! Why does my mum beat the crap out of me? She hasn't done much martial arts (she only was in my class for a couple of years or so) or anything else to train reaction, plus she definitely is 25 years older than me. That said, she is pretty fast at catching balls or catching falling dishes / glasses "in-flight". But wait, this brings to my mind my Karate grandmaster Takahashi Sensei of Japan (9th Dan, President of JKA Australasia). He is about in his 70's, and I know of nobody (including myself) who can see his fast moves coming, not to mention block. He is extremely fast. So there goes the age argument, completely swept under the carpet. Now I'm a bit confused. Does age matter at all? Well, only marginally I guess. Does training matter? Perhaps. Most likely. Maybe it's genetic? Maybe it's nutrition-related? Quite likely it's even got to do with the amount of hydration in your brain. Water drives and accelerates your brain, as in synaptic functionality. Neurotransmission can be greatly inhibited by dehydration. Most people don't drink enough, are more or less dehydrated most of the time. I drink 3L+ every day. The consumption of caffeine or nicotine has different effects on concentration and synaptic activity. I drink caffeine, but don't smoke. Caffeine puts your synapses on "full rush-through" mode and accelerates your hearbeat. On the other hand it is also a powerful dehydrant, and the more coffee/tee you drink, the more water you must drink on top of that. Nicotine temporarily enhances concentration & focus (apart from being a neurotoxin in many other ways). My mum smokes. Perhaps nicotine had an unfair influence? Then there's a huge quantitude of vitamins, such as pantothenic acid (B12?), which have an influence on neural function. Let's not mention all the things you can find in energy drinks, ginkgo or ginseng. Anyway, I'm not exactly in the faculty to go deeper into the neuroscientific realms of the topic. Let's hope somebody with more knowledge in this field gets lost on this thread and makes a few useful comments on this topic, answering questions like these: - What influences your functional brain speed / reaction times? - Does age matter? In how much? - Does training matter? If so, how much? - Does reaction speed, once acquired, need to be maintained? - What is the fastest possible "reaction time" a human could score on a test like this?
Good ole Europe, ey? Jeez, it's been a few years now but I still miss the technology standards over there...
Gums, This is not 100% accurate. Most people assume that latency = RTT / 2 (RTT = Round Trip Time.) Remember, there has to be sig/ack. It is entirely possible that one direction is 10 ms, and the return is 40 ms. What is worse, it is really hard to be able to measure this without expensive equipment at the POP (Point of Presence.) I know, friends of mine that need the absolutely fastest connection have tried... In other words, do not confuse, or pehaps a better word is do not use interchangeably, RTT with latency, although I think that 99.9999% of the people on the planet use them interchangeably, even people in the business. nitro
I think you are missing the point entirely: 1) These very low latency connetions are meant for high speed computers that need to respond to situations and put an order on the wire with as little latency as possible. You measure your reaction time in 250 ms, or 250/1000 of second. The computer will have gotten the quote, analysed it, routed the order and put it on the wire in 1/500,000 of a second. And if all you measured was it's reaction time to the actual quote, the benchmark woud be closer to 1/10,000,000 of a second. 2) _EVEN_ for humans, believe me, the difference between a connection that is 50 ms slower than another traders', _all_other_things_being_equal_, is _huge_. I know, I live it everyday. nitro