SERVING THE QUANTITATIVE FINANCE COMMUNITY

 
User avatar
dalecooper
Topic Author
Posts: 7
Joined: July 15th, 2009, 8:14 am

why is it impossible to manage market risk

November 8th, 2009, 7:56 pm

HiI'm thinking about writing my dissertation as an answer to question: "why is it impossible to manage market risk (in commercial bank)?". It's not the final choice - I'm still wondering if it's worth exploring. I'm at the very beginning of my work and have these ideas:A) when valuing the same instrument with different valuation models you get different values (examples maybe on CDS, fx options, CDO?)B) when valuing the same instrument (and calculating risk measures) using different contributors for market data you get different values (examples: - discount curves may differ when using quotations from different brokers or markets – IRSes/fras/depos; - CDS spreads may differ depending on source: Markit/CBGL)C) when calculating delta for fx options you get different numbers depending on delta type: analytical VS numerical sticky-strike VS numerical sticky-deltaD) when calculating VaR using different distributions of returns you get different values (also you can't find any 'right' distribution for market data)E) when calculating IR VaR (parametric) using interest rates in different conventions (CONT vs ANNU) you get different valuesF) when market is 'shocked' and your instrument becomes illiquid then your mark-to-model valuation is worth less than nothing (some examples)Having these calculations I want to show that you can't really rely on your prices (models) or risk measures. Hence you don't really know for sure how big your exposure is, or how much money you can get (lose) when you'll be forced to close your positions. What you need to manage risk is something more than these numbers.Now to the point: I don't want to explore something that someone else already did. That's why I want to ask you:1. are there any papers/books concerning generally market risk management or something mentioned above?2. do you think my ideas are reasonable and empirical data will confirm it? (in some cases I'm sure the data will prove it's right)I will appreciate a lot any comments or hints.Marcin
 
User avatar
Martinghoul
Posts: 3256
Joined: July 18th, 2006, 5:49 am

why is it impossible to manage market risk

November 8th, 2009, 8:51 pm

I might be going out on a limb here, but, IMHO, exploration of somewhat self-evident points A through E is senseless navel-gazing that will provide no useful insight whatsoever. Point F alone, on the other hand, contains enough juice for a few dissertations.That's just my Z$2c...
 
User avatar
daveangel
Posts: 17031
Joined: October 20th, 2003, 4:05 pm

why is it impossible to manage market risk

November 8th, 2009, 9:12 pm

i think Martinghoul is correct. I think F is the most interesting. The Achilles heel of financial risk management is essentially any assumption that you make about the distribution of returns and how this impacts on your ability to manage tail risk. Perhaps, you should focus your dissertation on this ?
Last edited by daveangel on November 7th, 2009, 11:00 pm, edited 1 time in total.
knowledge comes, wisdom lingers
 
User avatar
Alan
Posts: 10211
Joined: December 19th, 2001, 4:01 am
Location: California
Contact:

why is it impossible to manage market risk

November 8th, 2009, 10:20 pm

Marcin,I will chime in on the thought that these points are (at this point) well-appreciated, including F).Here is how the regulators see it: Volcker committee recommendationsThe question is, can you significantly advance the argument beyond what they have done,or make a case for some fatal flaws, fundamental omissions, etc? Another topic critique is that nobody forces a commercial bank to buy securities beyond liquid Treasury stuff,AFAIK. They can always take in deposits, park the money in short-term Tsys, and then try to really loan out the money to homebuyers, businesses, etc.. If they do make such a loan, nobody forces them to sell it -- theycan keep it in their portfolio. Let's say they make loans to buy houses. Now you canmake the same laundry list a)-f) of why it is difficult to truly value that in-house portfolio of loans: a) different appraisals, b) different data sources, ..., f) illiquid markets, etc. Well, so what? That is the nature of banking. The question is: how do you deal with all those issues in a reasonable way? One last point. In the US, there about 8100 insured banks -- so far during this financial crisis,140 have failed -- 25 in 2008 and 115 in 2009 to date. While the failures/forced sales have certainly included a few big players and other very big guys have needed needed lifelines, the survival figures also suggest that there are *many* conservatively run (midsize and smaller) banks.So, perhaps you may want to research how the silent majority successfully managed their risks.
Last edited by Alan on November 8th, 2009, 11:00 pm, edited 1 time in total.
 
User avatar
Traden4Alpha
Posts: 23951
Joined: September 20th, 2002, 8:30 pm

why is it impossible to manage market risk

November 8th, 2009, 10:24 pm

F is the most interesting by far. I'm wondering if there's a more general issue with the choice of time horizon on the VaR relative to the bank's position size and market liquidity. As liquidity degrades, the least feasible time horizon of exposure to the market increases (toward infinity if the market becomes completely illiquid). This type of market risk resolves as a risk that the time horizon on the VaR is too short and thus the estimate of risk is too low.
 
User avatar
dalecooper
Topic Author
Posts: 7
Joined: July 15th, 2009, 8:14 am

why is it impossible to manage market risk

November 9th, 2009, 6:36 pm

Thank you all for comments.First - I agree that A-E are more or less obvious and because of this, they're not worth investigating. But not JUST that was my point. The point were to show that you can loose a lot of money if you rely on your numbers (pricing and risk measures). Anyway - maybe you're right that it isn't that exciting as I thought it is Alan - of course nobody forces banks to trade risky/illiquid assets. But they do. They ( we ) still want higher and higher yields, better EPS, ROA, ROE than others. Some say that banks want to provide superior returns to their shareholders... Even if they are not forced to take much risk - they will, and I want to write about it.Your idea about banks-survivors may be pretty cool, but I work in Poland and it will be hard for me to gather necessary data.Far more interesting seems to be 'Volcker committee recommendations'. Earlier I had this idea: the higher you look at management levels in banks, the lower is awareness of risk. Managers often just don't understand their business or how much risk they generate. The best way to prove it would be interview various managers (also higher level) - and that's impossible to me at this point of my career Nobody would ever want to talk to me.Traden4Alpha - the idea for longer time horizon on the VaR on less liquid markets sounds good - I will investigate it Thanks.
 
User avatar
bojan
Posts: 347
Joined: August 8th, 2008, 5:35 am

why is it impossible to manage market risk

November 9th, 2009, 8:45 pm

QuoteOriginally posted by: dalecooperThank you all for comments.Alan - of course nobody forces banks to trade risky/illiquid assets. But they do. They ( we ) still want higher and higher yields, better EPS, ROA, ROE than others. I think you should definitely include this sentence in your thesis! People literally risked everything for a few basis points more, believing the AAA rating from some guy clutching a copy of the first amendment... This behavior is a far bigger risk in most institutions than using the wrong discount curve
 
User avatar
sunra
Posts: 16
Joined: October 31st, 2009, 4:11 am

why is it impossible to manage market risk

November 10th, 2009, 6:35 am

OP i would respond that it is possible for banks to manage risk, but they are heavily incentivized not to. for some discussion of this topic that may stir your thoughts i refer you here ... http://www.econtalk.org/archives/2009/1 ... on_th.html. furthermore, quant models can't fully manage risk since they are built on history, which cannot predict the future.however, i doubt whether these are good topics for a phd as they go against the orthodoxy ... you're probably better off writing some pseudo-science about liquidity vs correls.
 
User avatar
Paul
Posts: 10789
Joined: July 20th, 2001, 3:28 pm

why is it impossible to manage market risk

November 10th, 2009, 7:56 am

The problem with F though is that in a couple of years' time there are going to be hundreds of theses on this topic. A-E are topics that you'd think would have been fully addressed by now...au contraire, the majority of quants still seem to think there ought to be a single, correct value for everything. Some even believe that a certain well known hedge fund has found the Holy Grail of the perfect model. The answer to the question is almost entirely dependent on your personality. Do you want an easy life, get the dissertation over with and get that high-flying banking job? Or live in misery, abused by your peers, and hope to be appreciated after your death?P
 
User avatar
daveangel
Posts: 17031
Joined: October 20th, 2003, 4:05 pm

why is it impossible to manage market risk

November 10th, 2009, 8:27 am

Quote Do you want an easy life, get the dissertation over with and get that high-flying banking job? Or live in misery, abused by your peers, and hope to be appreciated after your death? Does it have to be so extreme ? Can't we have a muddle through scenario ?
knowledge comes, wisdom lingers
 
User avatar
dalecooper
Topic Author
Posts: 7
Joined: July 15th, 2009, 8:14 am

why is it impossible to manage market risk

November 10th, 2009, 9:53 am

QuoteOriginally posted by: daveangelQuote Do you want an easy life, get the dissertation over with and get that high-flying banking job? Or live in misery, abused by your peers, and hope to be appreciated after your death? Does it have to be so extreme ? Can't we have a muddle through scenario ?Exactly - I still hope not to live in misery and write dissertation that will not be useless and pseudo-science. I even hope it will be (most of all) innovative :)Anyway, thanks for advice. I'll be back soon with more detailed ideas and questions (I hope).Marcin
 
User avatar
list
Posts: 2041
Joined: October 26th, 2005, 2:08 pm

why is it impossible to manage market risk

November 10th, 2009, 3:06 pm

Still subjectively, there are 2 primary drawbacks or errors in the modern pricing. The first one the PV concept. For example, in 2 steps stochastic economy it is easy to construct 2 not equal realizations of an asset price dynamics which have equal PV but admits arbitrage with probability as close to 1 as we wish. The second drawback is the illusion that the asset price is always the expected value of smthg. The EPV could be used sometimes as an estimate of the price but it could not be used as a definition of the price. Otherwise we will have what we have had with market risk. Excepting EPV as definition leads to non random pricing that in turn leads to comparing non random risk free bond return with non random derivatives return. It is not clear why professors did not cover by this theory stock pricing itself regardless derivatives just state that at any t the S ( t ) = EPV of the nuetralized stock having expected return risk free r.
 
User avatar
Anthis
Posts: 4313
Joined: October 22nd, 2001, 10:06 am

why is it impossible to manage market risk

November 10th, 2009, 3:29 pm

QuoteOriginally posted by: Traden4AlphaF is the most interesting by far. I'm wondering if there's a more general issue with the choice of time horizon on the VaR relative to the bank's position size and market liquidity. As liquidity degrades, the least feasible time horizon of exposure to the market increases (toward infinity if the market becomes completely illiquid). This type of market risk resolves as a risk that the time horizon on the VaR is too short and thus the estimate of risk is too low.On this issue i am afraid accountants have done much more meaningful work than quants.
 
User avatar
Paul
Posts: 10789
Joined: July 20th, 2001, 3:28 pm

why is it impossible to manage market risk

November 10th, 2009, 3:51 pm

list, you are unfortunately correct that most quants assume the answer to most problems will be an expectation of something, and this leads to a great deal of putting carts before horses. There are many models that do not lead to this, and they also cannot be easily solved by MC...the universally popular sledgehammer!P
 
User avatar
list
Posts: 2041
Joined: October 26th, 2005, 2:08 pm

why is it impossible to manage market risk

November 10th, 2009, 4:47 pm

One question that i do not know an answer. What the type of the risk is when the government decides to pump money to support market while people are loosing jobs?
ABOUT WILMOTT

PW by JB

Wilmott.com has been "Serving the Quantitative Finance Community" since 2001. Continued...


Twitter LinkedIn Instagram

JOBS BOARD

JOBS BOARD

Looking for a quant job, risk, algo trading,...? Browse jobs here...


GZIP: On