Keeping your name out of the press and staying secretive is incredibly important and I'm shocked that they didn't just throw a fair amount of cash to this guy directly or offer it up as a donation to some Carl Sagan foundation (and offer to let him keep his e-mail address). $200k for a shop like that is literally the shells of peanuts and worth it to keep your name out of people's mouths.

If their market cap is above 3 dollars I wouldn't touch the stock.

Another alternative is low bandwidth flag based roll-backs (for instances such as this where the network is congested but not completely lost).

There is a PR to get TensorFlow as the backend for PyMC3 too.

The author has a doctorate from NYU and the article was published in a peer reviewed journal.

That's just a funny example, but there are tons of poorly researched peer reviewed papers which makes it exceedingly difficult for me to accept credentialis as being a useful signal. I find it's much better to have ideas stand on their own merit, rather than on the shoulders of those who are espousing them.

So shipping working code (even if it comes with a required pipette) might be a nice requirement for a peer reviewed publication to take on in order to keep their journal relevant. Shipping in Docker or similar guarantees reproducibility.

And yeah it makes sense to have an extra prompt pop up if it's an order over X contracts or Y ticks from the market. And I've seen a lot of systems set up like that.

A lot of times traders will just punch the "OK" box and do their trade though.

That's of course if traders are manually hedging their portfolio/trade. A lot of times they just set their portfolio to auto-hedge based on certain parameters (ie at Z deltas or we've moved C ticks in a time period).

The problem with disallowing your trader from ripping through a lot of the levels of an order book is that it can be a risk reducing move and what you intend to do sometimes. This trade is a clear fat finger but there are times when you will want to sweep the book to get hedged.

For instance, let's say your desk just got slammed with a ton of risk on an OTC (over the counter) option trade. You can immediately alleviate a lot of that risk (while paying through the nose) by selling 2000 contracts or 5 price levels of the ES (SP500 future). You can immediately place that order and get it filled and be hedged. If there were multiple points of human intervention required then you might lose a substantial amount of money. 2k contracts on the ES is $25,000 a tick. If word leaks that people are going to need to start hedging big then it could easily move 10 or 20 ticks away from you while waiting for your risk management team to approve your trade as not a fat finger.

Generally it's cheaper to just fire error prone traders. Heh, and anyone that is about to execute a 2k contract option trade generally has their hedge order queued up and ready to send to the market as soon as they hear the other side agree to their price.

You are saying that the purchasers have broken even on the deal so long as they have not lost money on paper in the valuation since they can probably either borrow against it or sell it off.

It comes down to are you paying for current value, or potential growth in value. Frankly, in my opinion, markets that rely upon a greater fool to bail you out to make your money rather than just getting it back the old fashioned way by providing a good or service are way too frothy.

Counter-argument is that you are betting on growth, not a greater fool bail out. In this instance the only way to get that growth while still handling less than 12 billion in transactions is by increasing their take from 5% to something higher or adding on ancillary sales. I'd bet on ancillaries, but this isn't my space.

By the way for those reading who may not know, I was referencing Greater Fool Theory, not making a pejorative statement (heh, beyond what the theory itself may make). https://en.wikipedia.org/wiki/Greater_fool_theory

Ideally I want a hiring "fun"-nel where a 6.2 can learn the skills necessary to become an 8.3 (or show that they've been an 8.3 all along) through exercises or reading assignments. Sort of an external training/hiring funnel. If nothing else it may also help as a way to build internal training processes and advertise the to the world some of the neat stuff you are working on. I think this is how Matasano recruits and it makes a lot of sense to me.

Pretty much no matter what there is a learning curve associate d with bringing on a new hire. Finding candidates that will attack those curves with gusto is key to building good teams, I think at least. Technical skills can be taught and refined, gumption is a bit harder to instill.

One thing that I disagree with is the idea that filtering processes can't make candidates better. A properly created process should leave both sides of the equation happy, an engineer can learn new skills and a hiring manager can have a shot at making an offer to an engineer if s/he has the requisite skills. If the candidate doesn't have the skills then give them the tools to learn the skills and see if they come back showing mastery. Work ethic and being ready/willing/able to learn new things is the number one signal that an employee is going to work out well, at least on the teams that I have run.