What you more need is perhaps some of the types that formed the MIT gang that raided the BJ tables in Vegas.
Rebels and misfitseven before they have graduated.
Look more for someone who is not likely to graduate and thus not be drawn inexorably into the City perhaps ?

And in case you don't know Magician, subversion is quite modestly not pointing out that he is a retired ex-quant.
So he does know a few things about all this.

first City job? that seems such a long time ago now. my background was engineering and tech, so had the maths/tech background they were looking for. started out in non-business roles, studied the financial maths on the job. worked my way closer and closer to the trading floor, until i was as close to the chaos as i wanted to get.

pretty common story really... an engineer disheartened by the lack of good engineering jobs in the UK, lured by the City.

i probably agree with most of what has been said about how (perceived!) better offers will be available to a lot your targets

have you thought about some alternatives?
1 - employ students part time / in holidays?
2 - find gambling enthusiast student / new grad?  the one thing that you offer to this kind of candidate is access to capital

Buzzer

no not really... assume a few of them are still betting here.

Micky Biggs, called me perhaps a year ago for a catch-up (but it might have been longer ago time flies)

city probably use c++ because its fast and they are looking for ultra low latency. titting about with the kernal etc

might be limited extra benefit for betting depends what your apps use. Engineers probably are more familiar with c or c++ these days than comp sci students who woyld imagine would be mostly OO

Magician - I think the advice you've been given is a little oversimplified.

just to give you a bit more about my background (so you can better judge whether I'm full of sh!t or not), towards the end of my career (as I got more senior) I spent quite a lot of time in 'strategic graduate recruitment' roles for the whole 'analytics' based skillset - ie quants, quant developers, and front-office tech, and finding pools of graduate talent to convert into these roles. by 'strategic' it basically meant scouring the various business regions, and has led me to take part in recruiting trips from Oxbridge, to top European Unis, to Indian IITs and mainland Chinese Unis, and even west-coast USA institutions such as CalTech, UC Berkeley and Stanford (just try convincing wannabe future google-founder type Comp Sci grads that they'd rather work in finance!)

its true that quants and HFT in the city still heavily use C++, and in fact there is a lot of C++ used in the city fullstop (although an awful lot of this is now legacy, ie new projects are using C++ less and less).

but this is the slightly perverse thing about your ad. areas such as graduate 'quant' courses will often teach C++, and low-level programming areas like electronic engineering may still teach C/C++ since its 'closer to the metal'. but C++ teaching in Comp Sci has been in decline for years, and your ad is specifically asking for a Comp Sci-esque degree.

because of this reduction in C++ teaching at Uni level, someone with good C++ definitely can demand a 'premium' in the jobs market compared to other languages (i often ended up looking for raw intellect in the graduate candidates to work for me, and training them in C++ myself)

in addition, although C++ is still the 'language of choice' for certain tasks, in others it is overkill.

in short be careful of taking techie advice from techies. in my experience, many of them are too keen to relate problems to their own skillsets (if you have a hammer, every problem looks like a nail), and haven't had to actually figure out how to solve problems at a business level, such as how to actually build and retain teams with certain skillsets etc

anyway, maybe we should continue this discussion offline? I don't really want to clutter up the forum with too much techie jargon.

In reading this thread, I fear for the world of gambling ( or my world to be more specific) if, to be successful, you need anything remotely like the skillsets, in part or whole, being described/discussed on here.
I suppose one can still hang onto the encouraging fact that apparently although Warren Buffet still has trouble using his office answer phone equipment, he can still apparently manage an operation that makes money head to head against all the quants and traders and hedgies on Wall St.
But maybe even his days are numbered ?

To sub, my last post.

In reading this thread, I fear for the world of gambling ( or my world to be more specific) if, to be successful, you need anything remotely like the skillsets, in part or whole, being described/discussed on here.

This highlights betfair's stupidity FAFH. The information available through betfair's API encourages parasitic behaviour and has created a glut of "professional gamblers" who generally supply nothing in the way of real liquidity and know nothing about the sport they're "betting" on. It took betfair 8 or 9 years to realise this but their solution to the problem is as imbecilic as their creation of it.

FAFH - tbh, I've never been involved with hiring traders directly (mainly wannabe traders, ie dangling the carrot of a possible future trading job after they've put in the hard time in non-trading roles)... but more generally, nearly all revenue generation in financial markets, online gambling, and 'traditional gambling', involve the ability to extract value from systems or games in various ways. so yes the skillsets overlap and are used in recruiting.

Feck/cat - re. API-based trading etc... as usual the truth lies somewhere in between.

up to a point, increased liquidity in financial markets via electronic trading has indeed led to smaller spreads, and thus better prices as well as much faster execution for investors.

however, beyond a certain point... things change. the speed of trading (millisecond speeds and faster) drove a lot of traditional market makers out of the markets, and now the vast majority is driven by algorithms. we now have a new phenomenon - the 'flash crash'. a spectacular one happened on May 2010 in the Dow Jones. what good is liquidity if it can all disappear in milliseconds?

i can't be bothered to write too much on this subject, but there are other problems too. google '2010 flash crash', 'quote stuffing', 'subpennying' for starters. curiously enough, these phenomenon are quite similar to the things Feck often complains about.

Just a few points c++ for api programming is not really required as the speed benefits are effectively lost due to the speed of your internet connection which will always be inferior.  Therefore using say c#/vb.net is probably easier and quicker to develop.  Though server side calculations etc will be quicker in c++ if there is a need for database connection then going directly through the dbms i.e., tsql sprocs in sqlserver, would be quicker still (except memory intensive processes where it is quicker to "download" data into an app and process).  I try to avoid c++ where possible as its easier to create errors and slower to build solid applications... I know my comments are slightly off topic as I know you would like somebody who can dip into previous code but wanted to make this point if it'd help you widen your search?

Sandown, I agree there are probably few opinion players amongst us (I am one).  The problem is automation is becoming a necessity due to the platform betfair provides...bets have to be drip fed in to hide large bets from being copied.  The opinion players are being unfairly treated (imo - obviously biased lol) as they don't require massive resources from betfair (a check of bets status every minute or so is sufficient) and provide most "real" liquidity but this is another topic and has been discussed elsewhere...  The issue of the levy could be easily solved imo if the bha were to develop their own exclusive exchange but I would fear they would get it seriously wrong - I haven't given this any thought as may be obvious?...oops I keep digressing :S

The difficulty of the task of beating the market, was nicely illustrated in the BBC prog "The Code" in which 160 people were asked to estimate the number of jelly beans in a jar. Nobody got the anser right but the average of all the widely differing estimates came to only 5 more than the actual number which was something like 4510 if I recall proving how accurate the market can be.


Sandown,

If you haven't already read it, I highly recommend Surowiecki's book "The Wisdom of Crowds":

http://www.amazon.co.uk/Wisdom-Crowds-Many-Smarter-Than/dp/0349116059/ref=sr_1_1?s=books&ie=UTF8&qid=1313150457&sr=1-1

about precisely this subject, and its relation to markets etc.

Yes, great book.

subversion
Date Joined: 07 Aug 03
Add contact | Send message
When: 12 Aug 11 10:50
Joined:
Date Joined: 07 Aug 03
| Topic/replies: 3,685 | Blogger: subversion's blog
Feck/cat - re. API-based trading etc... as usual the truth lies somewhere in between.

up to a point, increased liquidity in financial markets via electronic trading has indeed led to smaller spreads, and thus better prices as well as much faster execution for investors.

however, beyond a certain point... things change. the speed of trading (millisecond speeds and faster) drove a lot of traditional market makers out of the markets, and now the vast majority is driven by algorithms. we now have a new phenomenon - the 'flash crash'. a spectacular one happened on May 2010 in the Dow Jones. what good is liquidity if it can all disappear in milliseconds?

i can't be bothered to write too much on this subject, but there are other problems too. google '2010 flash crash', 'quote stuffing', 'subpennying' for starters. curiously enough, these phenomenon are quite similar to the things Feck often complains about.


I guess none of those risks exist in betting exchange trading though. Also uncorrelated markets, really are uncorrelated. They don't all start moving in lockstep in times of a 'crisis'.

You could say to paraphrase:

up to a point, increased liquidity in betting markets via automated trading has indeed led to smaller spreads, with no negative side effects for leisure punters (but negative side effects for winners displaced by bots).

subversion,
The good market makers still make the markets although they now do it after the bots have loaded the money in, how else does the market form accurately?  Probably the poor market makers can't compete but that's tough.

I'd say the bots help to create tighter spreads, not just because they happen to be bots, but because the more winners there are competing with each other the tighter the spreads will be.

So leisure punters still lose, but they suffer no negative side effects as they were already losing anyway, and will probably lose less with tighter spreads.

Take a look at the "tight spreads" on horse racing on last5minutes.com and try pinning the tail on the donkey. I suppose these pendulum bots are just doing it out of their love of leisure punters?

I suppose these pendulum bots are just doing it out of their love of leisure punters?

Obviously, virtually all price offerers operating on here are acting entirely out of self-interest (even you, Feck!). It is a feature of a free market, though, that such collective pursuance of self-interest on the part of the 'suppliers' tends to result in a fairer deal for consumers.

such collective pursuance of self-interest on the part of the 'suppliers' tends to result in a fairer deal for consumers.

WALOFS