Looks quite ambitious. (And cool)
The numerical python ecosystem is definitely improving, but it's getting pretty complicated, particularly when heterogeneous architectures factor into your work flow.
111 karma · joined November 9, 2011
The numerical python ecosystem is definitely improving, but it's getting pretty complicated, particularly when heterogeneous architectures factor into your work flow.
I don't know of a single firm that's successful in the time horizon that a lot of pseudo-HFT systems operate in. (The 10ms-3s range) And that's assuming an ideal fee situation...
Most of the people ignoring the LL arms race target RV opportunities on the 30s+ range. This involves taking a pretty big step back from microsctructure/toxicity models.
Typically, the types of projects I'd outsource to a consultant are non mission critical and the pacing is such that a 1/2 time consultant is totally fine.
If you want to heavily incentivize a cultural shift towards more vacation time for US employees, I don't think a federal mandate is as attractive as just an economic incentive. (Tax break to companies where X% of employees take more than Y days per year etc)
Otherwise -- it's definitely something we'd consider using; I end up doing 99% of my work on my laptop because AWS setup fixed cost time is 30minutes plus every time.
I've never really spent much time looking at pandas' to_datetime but I believe it has to handle a lot of variety in what you pass to it, which probably cause a bit of a perf hit. (Lists, arrays, Series)
http://dl.dropboxusercontent.com/u/14988785/ciso8601_compari...
I try to meet up with young people interested in HFT for coffee or beer if/when they come through Chicago and I've really found the experience profoundly useful to me thus far.
Generally, this has just been facilitated by random LinkedIn messages by bold students from one of my almas, but I imagine if I was less of a slacker, I could more proactively be useful to people with an interest.
You and others on voice chat were in a 'virtual' room and the sound was stereo/volume adjusted accordingly.
I remember that when people misbehaved on group voice chat we used to "put them in the corner"
It didn't map to the game environment but the execution was pretty nice; I fear that, if you use the game's physics, you'll quickly realize why radio-comm became such a thing
Early data access has been around for a decade and generally has NOTHING to do with HFT. Events desks typically have very different architecture than other groups in HFT, and, well, are a very very small cog.
Honestly, the three things I'd most love to have are: - more wrappers, support, documentation on PyMC -- a LOT could be done here and this could replace HUGE amounts of code in all sorts of place - libraries/models on top of cvxopt - more polished ipython notebook -> pdf or d3.js html -- I do this a fair amount for auto eod reports etc
And maybe a fourth...a good open source alternative to kx...preferably one that can accept .h5 files.
Anyone know of a good short C implementation of TCP that includes SACK?
(Yes there's no point as it's better in hardware blah blah)
Add some of the exchange volumes? Num transactions? Number mined?
I remember in school a million years ago we discussed an algorithm of the following type for sending short covert messages.
1. Negotiate cipher/mapping for where to look for hidden information 2. A wants to send B message "Let's get drinks @ 9 @ Bill's" -- instead of inserting this into some random file, he instead maps to the cipher/mapping area and then iteratively searches for images/texts that are closest possible matches in those bits to his message. 3. Ideally, given access to enough cover files and a short enough message, he has an EXACT match. A sends B picture of puppies with NO bit twiddling. B knows to meet at the pub.
Other things:
- Prices around the world across all exchanges are kept in line. If you need to hedge your Yen exposure and can only trade in Europe, you will get the closest possible price as someone in Asia or the US. This is true across all asset classes. This lack of massive arbitrage may get a lot of investors/hedgers/speculators significantly better prices.
- Derivatives on almost any asset are similarly tight as their underlying assets. No longer are you crossing wide/illiquid markets to hedge back month products or options. In a lot of markets, you can trade tight/liquid markets at almost any tenure; this is potentially a LARGE cost savings particularly compared to what some of the markets looked like a decade ago.
- Stat Arb + Relative value spreads are kept in line with their variance structures. Stocks move in line with their eigen components, options move smoothly with their volatility structures, futures at various tenures are priced in line with their synthetic interest rate exposure. Without being able to accurately price something, most participants are able to get a 'good' price on almost anything.
While a lot of the HFT is indeed zero sum between firms, the net economic effect of their activity is certainly not. The amount of $ that most firms make from the HFT business is really not that large; most of the big trading firms have made their fortunes from large speculative/fortunate events, not from being the best at constantly collecting a 1/10th of a tick of edge over and over.
It's a rather diverse industry in terms of the type of developers being sought. People with skills ranged from low level hardware design all the way up to Haskell/Scala are pretty constantly in demand.
That said, working with the GPU requires a very specialized set of high level constructions as well; I imagine that most workloads could have significant parts of their computational path abstracted out to pretty common easy to port fpga processes.
If you ever launch a beta...let me know!
Have there been any performance improvements to command line tshark recently?
As for fully remote, we don't have any full-time employees currently working remotely, but we do have some people that do projects off-site, so depending on the role, it's a possibility. Generally, for developers, I think remote is pretty hard to make work in this industry, but I'm open to being wrong. On the quant side, it's a bit easier assuming you all work on the same stack.
We're one of the many, many HFT startups in Chicago; on the developer side, we're looking for C/C++ people with significant networking expertise or those with hardware experience.
We're also in the market for individuals with general HPC/networking expertise on the IT side of things.
Drop me an email if you're interested: green at glt-llc.com
It's like more about the derivative (delta/improvement) of quality, the velocity of the game relative to previous efforts not its actual position in space.
We're one of the many, many HFT startups in Chicago; we're looking for C/C++ people who love sockets (TCP/UDP not wrenches) Hardware design skills are a plus.
If you instead feel more comfortable talking about multivariate time series, subspaces, manifolds, volatility surfaces, or just general data analytics, we'd also love talking to you.
Financial industry and/or low latency experience are not musts. Drop me an email if you're interested: green at glt-llc.com