Landing a Job in the "P" World

[goldmanta]

I want to tailor my self-studies (and perhaps part-time graduate education) for the next two years in order to prepare myself for a buy-side job. I am not as interested in Derivatives Pricing (Q) because, generally, doing so seems to entail (1) more math, (2) more C++ programming, with (3) shrinking number of jobs.

Can someone please help me assess the amount of math/statistics that an entry-level job (analyst/strategist/researcher) such as the one below may require?

Responsibilities:
-Development and maintenance of financial models, using various programming and other tools
-Working on new investment strategies and frameworks
-Doing customized research to address investment policy issues for clients
-Build robust software tools for internal users

Qualifications
-Strong mathematical and analytical skills
-Programming expertise and ability to transform concepts and ideas into robust software
-Familiarity with and interest in financial instruments and methods
-A self-starter, should have ability to work independently as well as thrive in a team environment
-Knowledge of firm's proprietary platform is an advantage

 

[IlyaKEightSix]

What particular job are you looking at? The high-frequency trader I normally communicate with says if you want to do strategy research, write strategies.

 

[goldmanta]

I'm not particularly picky about what type of job I get, so long as I land myself on the path to portfolio management. Of course, the closer I get to PM, the better.

What would you say is the level of math/statistics that is required to write strategies that are sufficiently sophisticated to deserve merit?

 

[IlyaKEightSix]

Can you replicate academic trading papers? You need enough math to understand those =O

 

[petercaya]

Hi Ilya,

Is there a good place to find these academic trading papers?

 

[IlyaKEightSix]

Google for them.

 

[goldmanta]

Given that all I know is Multivariate Calculus at this point, what other mathematics and statistics courses do you say that I should take in order to possess sufficient breadth for understanding buy side academic papers?

I am not looking for the bare minimum - neither am I trying to go above and beyond. I am simply trying to optimize my preparations for the buy side so that I do not study things that I wont be needing in the short term (ex. Delay adv. Derivatives Pricing for the "long-term" to-study category).

 

[IlyaKEightSix]

Have you ever read those papers? Just a general comfort of being able to digest lots and lots of derivations, along with absolute fluency in linear algebra.

As someone trying to do this myself, what I find most difficult isn't any of the mathematical techniques in particular, but just keeping track of which variables represent what, and what common matrix expressions I should look out for and internalize. There's no differential geometry, algebraic topology, and other sort of tongue-twisting insanity. Multivariable calculus, linear algebra, and just the ability to read mathematical notation.

I wish I got more practice with that in school >_>

 

[Andy Nguyen]

Hi Ilya,
Is there a good place to find these academic trading papers?

http://symmys.com/resources
That site belongs to Attilio Meucci who wrote an article on the subject of P versus Q that the OP referred to in the title of this thread.
https://www.quantnet.com/threads/p-...n-the-two-areas-of-quantitative-finance.6657/

 

[IlyaKEightSix]

Thank you, Andy. Odds are, the sort of trading quants fall under the "P". And what's interesting is that so many of the top programs are all focused on creating the Q type of quants. Where does one get tutorials on creating trading systems?

 

[Bill Sun]

because the so-called P-world mostly calculates the technical indicators to trade, like every under high-school level educated amatuer day-trader. Thats's no science, just rubbish, why should top progamms teach that. If you are serious in finance, go sell-side Q-world.

 

[alain]

because the so-called P-world mostly calculates the technical indicators to trade, like every under high-school level educated amatuer day-trader. Thats's no science, just rubbish, why should top progamms teach that. If you are serious in finance, go sell-side Q-world.

Actually no, read Attilio's papers to convince yourself.

 

[goldmanta]

because the so-called P-world mostly calculates the technical indicators to trade, like every under high-school level educated amatuer day-trader. Thats's no science, just rubbish, why should top progamms teach that. If you are serious in finance, go sell-side Q-world.

I am not very familiar with the topic yet but what you're saying sounds very ignorant and reminds me of why I don't post questions on Wall Street Oasis. I personally found my personal investment management quite enjoyable and am "serious" about doing that line of work in finance. But can you please elaborate on what you mean by the majority of the P World calculating technical indicators? Are you talking about stuff like MACDs and Bollinger Bands used by day traders?

 

[JulianT]

The following blog post might be of some interest to you
http://quantivity.wordpress.com/2011/09/21/p-q-convergence/
Sorry I don't know how to hyperlink from my iPad.

On a more related note, the idea that the P world (commonly used for the real world probability measure) mostly used technical indicators to trade is not entirely false, since there is a lot of retail trader using technicals. But there is also a lot of funds using a pure fundamnental approach which is not quantitative whatsoever yet they tend to do quite well. If you consider the P world to be the buy side, it is indeed moving into a more quantitative world where they are using factor models for asset allocation and or building portfolios. If anything, serious people in finance go to the P world since it is where you actually manage money. For sophistication, in the P world look no further than the proliferation of ETFs and the low vol craze we are seeing.

Also, I would argue that large swaths of the Q world are largely using the same high school maths technical indicators, as soon as you step off the derivatives desk. The simple fact is that if I am market making a bond market, I'm more interested in limiting my inventory risk and the real world than the theorical arb. Markets can stay irrational longer than you can stay solvent, or on the desk, because the tap on the shoulder will come as your losses mount.

In short, there are quants on both sides, the main differences depends on which products you trade. Buy side quants that trade var swaps are likely to work in the q measure (risk neutral) just as much as sell side quants working in risk management will do their stats with real data, this working in the p measure.