View Full Version : Unit Outline

I would like to raise some questions on the unit outline of the APM course.

I took a brief look at it and on section 1 of the outline, it mention 'theory of equity on exotic options'. Anyone know what is this about?

And also, on section 2 it says we have to know 'Relative VAR'. Is that equivalent to compound VAR?

Another note, if we value exotic options, can we apply the put-call parity on that? Say for example, if I have the price for Asian call option, can we use put-call parity to calculate Asian put option? Just some thoughts :-p

Thanks.

rekrap

04-28-2007, 06:37 AM

I would like to raise some questions on the unit outline of the APM course.

I took a brief look at it and on section 1 of the outline, it mention 'theory of equity on exotic options'. Anyone know what is this about?

The exact line is:

Demonstrate mastery of option pricing techniques and theory for equity (including exotic options) and interest rate derivatives.

And it means: know not only how to price options (techniques) but also why it works that way (theory). This includes standard options on equity, as well as exoctic options and interest rate derivatives.

And also, on section 2 it says we have to know 'Relative VAR'. Is that equivalent to compound VAR?

I don't have my study notes at home, but I recall Relative VaR being discussed explicitly in SN 105.

Another note, if we value exotic options, can we apply the put-call parity on that? Say for example, if I have the price for Asian call option, can we use put-call parity to calculate Asian put option? Just some thoughts :-p

This is an examle of option theory. If you can use put-call parity, or if you can't, why? It would be a great question to ask in the Hull thread, referencing Ch 22.

Consider that an Asian call compares the current price at maturity to the average price over the life of the option: the average price is simply a calculated K instead of a pre-determined K. Shouldn't put-call parity still hold? What about lookback options, where for calls the final price must exceed the minimum, but for puts the final price must be less than the maximum; should put-call parity hold?

Thanks.

You're welcome. Hope this helps.

Rekrap,

Thanks a lot. Think relative VAR is mean zero absolute VAR.

On part (b) of this section,

it syas incorporate paratmeters affecting client's needs in an ALM framework including funding objectives etc.

What is the regulatory and rating agency requirements for this?

And concerns abou solvency?

Anyone have ideas on which notes to refer to on this?

On part (h)

Choose between alternative strategies and justify such choices. What kind of alternative strategies they would like us to know? Where can I read more of this?

Thank you.

bagheera

04-29-2007, 01:45 AM

Rekrap,

Thanks a lot. Think relative VAR is mean zero absolute VAR.

No, it's not that. In fact, it is an alternative to mean absolute VAR when your trading horizon is long and rebalancing is relatively infrequent.

This should clear up everything.

http://www.risksrv.com/risks/market/relativevar.html

Bagheera,

Think I went to that website too just then. Ok let's summarize,

Relative VAR = (Portfolio mean - Portfolio volatility ) * confidence

Is that confidence the percentile?

It says something I don't really understand. It says that absolute VAR are used for traders who mark-to-market daily. Is that because they assume only volatility as the risk. And whereas in relative VAR, they uses the portfolio mean, thereby more suited for individual investors who balances position weekly and monthly. What is the advantage of doing this? So they can see the returns earn on the portfolio, thereby we have risk-adjusted return percentile?

Thanks

No, it's not that. In fact, it is an alternative to mean absolute VAR when your trading horizon is long and rebalancing is relatively infrequent.

This should clear up everything.

http://www.risksrv.com/risks/market/relativevar.html

bagheera

04-29-2007, 05:26 AM

See below..

bagheera

04-29-2007, 05:38 AM

Bagheera,

Think I went to that website too just then. Ok let's summarize,

Relative VAR = (Portfolio mean - Portfolio volatility ) * confidence

Is that confidence the percentile?

It says something I don't really understand. It says that absolute VAR are used for traders who mark-to-market daily. Is that because they assume only volatility as the risk. And whereas in relative VAR, they uses the portfolio mean, thereby more suited for individual investors who balances position weekly and monthly. What is the advantage of doing this? So they can see the returns earn on the portfolio, thereby we have risk-adjusted return percentile?

Thanks

Yes, confidence indicates percentile. Strictly speaking, the mean return will have to be considered in any VAR calculation if it is significantly high. Indeed, if there is a portfolio with 25% return, volatility cannot be the only variable affecting the fluctuation in value. In this case, mean return will also govern changes in value.

Now absolute VAR does assume that the "drift" or the return of the portfolio is zero but this assumption is valid only over the short run. This is so because, over the short run, the positive returns, more or less, cancel the negative returns. Therefore, the short run mean is much smaller compared to the volatility and can even be ignored in the VAR calculation. However, if you are rebalancing the portfolio every ten days, the overall return for ten days may not be zero and it would be misleading to concentrate solely on the volatility to determine the fluctuation in portfolio value.

Another way to think of it is to note that for the assumption of lognormal distribution for the portfolio value, mean return increases in proportion to the length of the time horizon whereas volatility increases in proportion to the square root of the length of the time horizon. Therefore, even though the short run mean may be small compared to the short run volatility, over the long run, the mean will become sizeably large compared to the volatility.

Would like to ask on the section 5 unit outline part (d) Describe commonly used equity and interst rate model (and their limitations) including ....and on compound poisson is that Levy Process? Where can I find information on this model? and on Copulas.

part (a) Crtize the following modelling methods:

1) Single period vs multiple period - which model this refer to?

2) Actuarial vs capitl markets - what is the capital market modelling method?

3) Formula-based model; is this refering to like Black's model those that can be expressed in closed-form?

Thanks.

campbell

04-29-2007, 09:01 PM

Hmmm, I don't remember compound Poisson showing up in the equity models we've seen. I know about it, I just don't remember seeing it in the readings. Most of the equity models show up in the Investment Guarantees text.

By the way, the "criticize the various methods" doesn't mean you should talk about =particular= models, but rather is related to why you'd pick a model with one property or the other.

For example, for equity models you can have formula-based like lognormal in that the quantities you're interested in have closed-form solutions. Almost all the models we come across have formulas of some sort, but the quantity of interest can't be gotten except by simulation. So what's the benefit of one style over the other -- why might you want formula-based, and why might you want one that involves simulation (consider the different credit default models, from the Risk Management text, as an example).

sx06103

05-05-2007, 09:30 PM

The exact line is:

I don't have my study notes at home, but I recall Relative VaR being discussed explicitly in SN 105.

Could anyone please point out where "Relative VaR" is defined? Which SN? Which page? I couldn't find it. Thanks a lot.

Could anyone please point out where "Relative VaR" is defined? Which SN? Which page? I couldn't find it. Thanks a lot.

Somewhere in the Risk Management Text Ch. 5 =)

rekrap

05-06-2007, 12:43 PM

Could anyone please point out where "Relative VaR" is defined? Which SN? Which page? I couldn't find it. Thanks a lot.

Somewhere in the Risk Management Text Ch. 5 =)

http://www.soa.org/education/course-catalog/spring-exam-session/2007/edu-investments.aspx

Risk Management,Crouhy, M., Galai, M.R., 2001, Irwin/McGraw Hill, Chapters, 7–12

Not only is that chapter is not on the syllabus, but also even it does not discuss explicitly discuss "relative VaR".

The best I can tell (http://www.sal.tkk.fi/Opinnot/Mat-2.108/pdf-files/ejav04.pdf):

Usually it is more informative to relate the maximum possible change in portfolio value to its expected value, E(ΔV). For this purpose, relative VaR can be defined (Jorion 2001): Rel VaR = E[ΔV] - Abs VaR

So, I would look again to the Jorion SN for more. Mine are at work, so still no help here... or maybe it's on an older version which the SOA hasn't noticed yet.

Quick question, anyone knows what does it mean by on the run and off the run treasuries?

Thanks.

rekrap

05-06-2007, 08:13 PM

Quick question, anyone knows what does it mean by on the run and off the run treasuries?

Thanks.

"On the run" Treasuries are the most currently traded (recent issues) in the market... "off the run" are all the rest.

Thanks rekrap.

One more thing, what is economic surplus?

rekrap

05-07-2007, 11:46 AM

Thanks rekrap.

One more thing, what is economic surplus?

Economic Surplus is simply the difference between the Assets and Liabilities. Babbel defines it in Ch 4 as the difference between its market value assets and the risk-free present value of its liabilities, using a duration-matched discount rate for the liabilities.

What is fat-tailed distribution?

Thanks.

rekrap

05-08-2007, 10:30 AM

What is fat-tailed distribution?

Thanks.

http://www.amex.com/dictionary/charts/chart77.gif

That's a nice graph :-p

thanks.

vBulletin® v3.7.6, Copyright ©2000-2014, Jelsoft Enterprises Ltd.