How fast is the TT platform? It is so fast we are posting internal and external latencies in the front end so you may easily compare our speed and monitor your performance.
Fastest commercially available futures platform
- Thread starter patrickrooney
- Start date
TT latency is the difference between the time TT receives the fill and the time TT releases the hedge order when using Autospreader.
Is it safe to assume that TT latency is comparable to a tick to trade measurement? Or shouldn't TT Latency be faster since it just gets a fill and sends an order
It gets very hazy when metrics like this are tossed around. The calculation methods used and reported can vary tremendously. For example, when you ask for "tick to trade" are you looking at the exchange reported time of the tick or the time the platform receives the tick update? On the trade side are you referring to the time the platform submits the trade? The time the exchange receives the trade? The time the trade occurs?
We are striving to make the TT platform as transparent as possible and will continue exposing more trade time metrics over time.
We are striving to make the TT platform as transparent as possible and will continue exposing more trade time metrics over time.
Understood. I think of it as everything internal to the application/platform. So in this case, measure once TT gets the fill (start time), TT does some stuff, TT sends hedge (stop time)
In the top row of numbers in the screenshot you have
TT latency of 46us and Exch latency of 198us. The Exch latency is the time between when data is sent from the exchange until TT receives it, then the TT latency is how long it takes TT to respond, then we should expect another exchange latency until the order hits the exchange.
In short, the time from when fill data leaves the exchange until the hedge order hits the exchange is 442us in that specific example and at that moment in time (198us + 46us + 198us)?
Just trying to wrap my head around what is included or not included in each measurement.
TT latency of 46us and Exch latency of 198us. The Exch latency is the time between when data is sent from the exchange until TT receives it, then the TT latency is how long it takes TT to respond, then we should expect another exchange latency until the order hits the exchange.
In short, the time from when fill data leaves the exchange until the hedge order hits the exchange is 442us in that specific example and at that moment in time (198us + 46us + 198us)?
Just trying to wrap my head around what is included or not included in each measurement.
@Lee- , thanks for bringing this up so I may explain.
Exch latency is the time difference between the hedge order leaving TT and the time the exchange acknowledges the hedge order.
In your example, TT released the hedge order 46us after receiving the quote fill and the exchange acknowledged the hedge order 198us later for a total of 244us.
Exch latency is the time difference between the hedge order leaving TT and the time the exchange acknowledges the hedge order.
In your example, TT released the hedge order 46us after receiving the quote fill and the exchange acknowledged the hedge order 198us later for a total of 244us.
Got it. Thanks.
Drew Shields
Sponsor
A couple other things to note in addition to Pat's comments:
1) Hedge-to-trade times are actually higher than tick-to-trade because deserializing an execution report often takes more time than processing a book update. We aren't yet able to provide tick-to-trade in the user interface but plan to do so eventually.
2) You get these latencies (some hedges in less than 40 mics) without having to purchase infrastructure. This is a fully hosted platform and the screen fees which can be found on our website.
3) We provide exchange latency because it helps traders understand the behavior of algos in a real market. When exchange latency is increasing, that has nothing to do with TT but it does often indicate increased load/traffic at the exchange which is useful in understanding how an automated strategy is working. Users shouldn't be conflating the two metrics. One is a measure of TT and the other is additional information that may be useful for interpreting behavior of a strategy.
1) Hedge-to-trade times are actually higher than tick-to-trade because deserializing an execution report often takes more time than processing a book update. We aren't yet able to provide tick-to-trade in the user interface but plan to do so eventually.
2) You get these latencies (some hedges in less than 40 mics) without having to purchase infrastructure. This is a fully hosted platform and the screen fees which can be found on our website.
3) We provide exchange latency because it helps traders understand the behavior of algos in a real market. When exchange latency is increasing, that has nothing to do with TT but it does often indicate increased load/traffic at the exchange which is useful in understanding how an automated strategy is working. Users shouldn't be conflating the two metrics. One is a measure of TT and the other is additional information that may be useful for interpreting behavior of a strategy.