How to be a rogue trader
As a City headhunter I’m repeatedly asked to explain how lone traders find themselves flushing billions of dollars down the toilet. Rogue traders can pop up just about anywhere, and so I’ll share this curriculum for you to follow, which is not specific to any bank: this is just the way it works.
Being a rogue trader is a superior gig for an aspiring criminal; instead of all that hassle of rioting and having to carry home your nicked 42in TV (they’re heavier than they look, or so I'm told), a trader on £450,000 will have it delivered and installed by that nice man from Richer Sounds, which is a mere two-minute walk from the office.
You'll also get to be played by Ewan McGregor in the inevitable film that dramatises your economic escapades, and millions of people will get the impression you bonked your way through the likes of Paris Hilton, which is better than a grainy CCTV image of you wandering around a trashed Dixons turning up on BBC London News.
An interesting question is how much of a coincidence it is that rogue traders don’t often come from top universities, however you define that. Is that an indication of motivations, or because if you’re just a bit smarter you’re much less likely to be caught?
In every rogue case I know of, the trader was smart enough to bend the IT systems to his will and went undetected by security until someone eventually noticed that enough cash to bankroll a small European nation had disappeared. When I first came into banking it was common for traders to have graduation from Sandhurst, the British military academy, as their top qualification. Routinely now, they are better programmers than many in the IT departments and several have written successful books on C++ and computational finance.
Why you need to go Rogue
No trading strategy is a sure win: you need balls to cope with the fact that between taking your position and the pay off, the price will go against you for a while. Bailing out early will cost real money and/or your job; the dictum that the first loss is the cheapest is for those without your insight.
The Back Office Monkeys (no, not a terrible pub band, the generic term for anyone not in trading) don’t understand your strategy; if they did, they’d be in front office. Just as bad, they stop you making as much money as you should by imposing limits when you know for a fact that twice as much money in would mean twice as much out because the market isn’t pricing the stock rationally. We all know that markets can remain irrational longer than you can remain solvent so much of your pain is caused by ...
You might think some experience in Risk IT is useful to the plan, but it’s the worst paid and lowest status in any given bank. The risk office’s job is to help stop bad things happening, not to make money, so their IT people are in effect minions of managers who themselves are badly paid with little genuine respect. Yes, you will have access to systems, but you won’t ever get to be a trader.
You want a mid-office role, the kind that involves handling tactical reports and liaising between traders and the risk department. It's the mid-office people who actually move the money and stocks. People outside the banking industry are often surprised that relatively junior people deal with telephone-number scale money, not unlike the armed forces where a £58,000-a-year officer has a key to launch nuclear weapons - albeit only one of a two-key set.
Banks follow the same pattern: traders decide where to move the money, but someone else does the moving.
Be a problem solver
Some of the smartest IT guys on the planet work in investment banks, but looking at the systems they use for risk and compliance you’d think their software had been developed in a joint venture between Capita and Accenture.
Their tools are being worked on constantly as new structured products are developed, pricing/risk models evolve, new issues are spotted, and as the regulators invent new rules that they pretend will make the markets safer.
That means practical skills in SQL and Excel VBA can make you look far more of a star than expertise in exotic technologies like FPGAs and GPUs or hardcore C++ because it gives you ...
Success in a bank is wholly based upon what your betters see you achieve and absolutely nothing else at all, not even slightly. Not one tiny bit.
I emphasise this because El Reg readers are mostly well-intentioned IT people who seem to genuinely believe that building a reliable system and making code elegant and bug-free is somehow useful.
You see, the path to rogue trading is not through worthy effort but through writing reports that need producing now and dealing with systems issues that stop traders trading. It is little bits of Excel that do useful calculations, or speeding up things that are too slow.
It’s not a coincidence that although Quants mostly use C++ for derivatives calculations, the three most important textbooks by Wilmott, Haug and Jackson & Staunton all use VB and so should you.
Getting something working quickly and on a screen seen by an important person is far more valuable than an industrial-strength solution a few weeks from now in C++ or CUDA that lives on a server somewhere.
Excel may not be hardcore but it can make you look good. Do not completely fix anything. An IT pro fixes things the best he can, with success defined as never having to fix it again. Take the model of government IT projects as the perfection of what you should be doing instead: you need to fix enough that you are seen to succeed, but not enough that it won’t need fixing again.
Data downloads from something like Bloomberg are a good example. It’s possible to code a generic parser to cope with the way Bloomberg randomly changes its fields, but that’s a kill-once solution and takes longer. What you want is to hardcode the format in VBA so that when it changes you can come and fix it quickly, looking good and since it’s a VBA module, the IT department will flatly refuse to get involved since there is no UML diagramming tool to produce the bulky faux documentation they so love.
Getting data in and out of trading systems is thus a key goal: I can get Oracle, SQL Server and Sybase tables directly into Excel. You’re not impressed are you? You know DTS, bcp et al, but to many people this is necromancy. That you 'get things done quickly' is a good phrase to have said about you; indeed, you should discipline yourself to always mention that you’re getting things working again in passing conversation. Anything that shows you're capable of using admin-level tools will lead you to gaining deep access.
There are information reporting systems you can master with a bit of effort; rarely are they properly documented so that low-grade skill actually looks good.
Traders are hostile to IT departments because they see them as black holes sucking money out of bonus pools and delivering little of any use any time soon. This is why front office want their own IT people.
As the person who deals with screwups and quick wins, you'll know that if your rogue trading leaves traces, you will not only be able to hide them, you'll also get an early warning that the game is up and it is time to run. Bear in mind that my work involves reassuringly expensive lawyers that you might need, so if the fan is hit you know where I am.
Quick wins are anything you can do in a day or two, or over a weekend; the hours of sin are not short, but following the visible productivity doctrine, make sure that you drop emails to decision makers that subtly let them know you were in Sunday afternoon. You can gradually add to your access permissions with some weekend work - either the permission holder has to come in and watch you all weekend or quietly adds you to the list. Maybe he takes you off after, maybe he doesn’t.
The developers of the systems leave debug manholes for quick and dirty fixes - you need to acquire these. There are two types of IT executive reading this: those that know and try to manage this, and those who genuinely believe there are none because the nice man from KPMG who did a system audit said so.