(1)I’m creating my CDO potfolio, is credit risk of synthetic CDO less than Cash cdo? on the other hand, in synthetic CDO, I take the risk but no legal ownership of underlying asset, but in cash CDO, I have legal ownership, so in all, which one has more risk, cash cdo or synthetic cdo? also in synthetic cdo, I can sell CDS, which gives me premium and better off
(2)I use equity tranch to reduce mezzanine tranch risk, does it mean both extension risk and contraction risk of mezzanine tranch is reduced after introducing equity tranch?
can I say equity tranch provide credit protection of mezzanine tranch, because eqity tranche get principal and interest payment after mezzanine tranch?
(3) in my risky portfolio, my analyst forcast alpha is 20%, can I adjusted the alpha by multiplying 20% with R square(correlation between relized alpha and forcast alpha)
(4). in factor potfolio, is beta=1? because other factor sensitive are all zero. in tracking portfolio, is beta=1? because it is same as index risk. so can I say the risk of tracking portfolio is same risk as factor portfolio?
(5)in maket neutral strategy, I long IBM and short microsoft, is the beta of maket neutral portfolio zero?
(6) my portfolio VaR is very high, does it mean the credit risk of my CDO portfolio is very high? VaR measures left tail risk, does it also meausre right tail, which means I have very high alpha return?
Thanks作者: adehbone 时间: 2013-4-8 16:09
1. I would say the synthetic CDO would hold more risk since there is always the possibility of the counterparty not fulfilling their end of the deal.
2. Credit risk to the senior tranches will be reduced if credit tranching is employed. In that case, the subordinate tranches would absorb the losses first.
3. I don’t understand why you would want to multiply r^2 by alpha, it is the coefficient of determination.
4. Beta is 1 for that specific factor, while all the other factor sensitivities are 0. The factor portfolio is a special type of tracking portfolio in which the goal is to have a beta of 1 for a specific factor and 0 for the rest of the factors. That is not necessarily the case for a tracking portfolio, where you are trying to track a benchmark.
5. The goal of a market neutral strategy is to reduce portfolio beta to 0, assuming IBM and Microsoft can perfectly hedge each other. In reality, this is extremely difficult to do since beta is constantly changing.
6. I can’t say I’ve heard of VaR measuring credit risk specifically. It just seems as a superior method of measuring downside risk as opposed to maximum drawdown.作者: maxmeomeo 时间: 2013-4-8 16:09
my dear friends, can I have your opinion? thanks作者: mkytz15 时间: 2013-4-8 16:10
3) Yes.
4) In a factor portfolio, there is only one Factor with Unit sensitivity to that factor. So, yes, In factor portfolio Beta for that factor is 1.
In a tracking portfolio, there are many factors. So, asking if beta for the portfolio is 1, is not a valid question. Beta is the sensitivity to a given factor. So, it is like asking: is sensitivity of a portfolio 1? I will then ask you: sensitivity of the portfolio to what?
For a multifactor model, we need to come out of the mould of CAPM model, where beta is always the sensitivity to Market Risk. In a multifactor model, there are many other priced risks, including market risk, thus, many sensitivities for each of those factors.
5) Yes, a market neutral portfolio will have sensitivity to market risk (beta) as 0. In your example, you will have to long some amount of money in IBM and short some amount of money in microsoft to get a market neutral strategy. These 2 amounts would not necessarily be same, if that is what you implied.作者: luckygiftvn 时间: 2013-4-8 16:11
Where are you currently located francisgy?作者: bdavi77962 时间: 2013-4-8 16:11
spirit has no fixed accomodation, salary can be transfered to anywhere.I’m covering global equity/conv bond/derivatives. I will tell you the location once June exam result is released作者: johnnyBuz 时间: 2013-4-8 16:11
francisgy wants us to share jobs with him and answer his awfully worded questions but he wont tell us where he is located, a little give and take here partner作者: mfleming1983 时间: 2013-4-8 16:12
My dear friends,
pls provided your linkedin link, I will invite you to join my network. you will know which nice place I’m in作者: noel 时间: 2013-4-8 16:12
alternatively, we can keep touch through: francisgycfa@gmail.com, Genius, don’t be shy作者: bchadwick 时间: 2013-4-8 16:13
Re Q6
My understanding is that VAR is simply the probability of the maximum drawdown occuring作者: edgeon 时间: 2013-4-8 16:13
(5) Also note the difference between beta-neutral and delta-neutral for market-neutral strategies.
(6) There is no such thing as vanilla “VaR”. VaR must defined over a time frame and event likelihood. For example, a 1-day 1% VAR would be say $10MM. This means, 1% of days you can expect to lose at least $10MM, maybe more.
Contrast this with 1-day 5% VAR, which would be lower, say $5MM. Again, this just says that on 5% of days you can expect to lose at least $5MM, maybe more.
5-day 1% VAR could be higher or lower (than 1 / 1%), depending on your return distribution.
