Level 2: Post what your remember here...

ibrahim-1987

Active Member
Regarding the Q about Monte Carlo vs delta Normal vs HS
VaR was HS>dn>mc

I think programming error was the answer
Bkz mc simulation is more accurate than dn, so it will produce more accurate measure, & dn has a an important deficiency, which is it's reliance on standard deviation, which is not capable to response quickly to change in the portofolio.

So I think a programming error made mc to be > than dn.
 

FRM_Exam

Member
Can anyone pls comment on the question - A manager wants to hedge his fixed income portfolio with an instrument which has negative duration? Is it

a) Buy the put option on the zero
b) Buy the put option on IO

I have a strong feeling after reading the above that b option is wrong now - I guess I got this one wrong :(
 

emer

New Member
B is correct

Hi Ibrahim,

it is not; Option A is correct. An IO has negative duration, it will increase in value when yield rise. Hence, the put on the IO will profit when yield go down. Thus will have positive duration. Apply the same argument for the put on the zero, and you'll see that this combination will have negative duration.

Regards,
J
 

FRM_Exam

Member
Hi Emer - Isn't there another way to look at this question?

If you concentrate on the wording - it says - with an instrument which has negative duration.

Let us divide and look

So if you want to hedge - ofcourse you will buy a put option
And the second part wanting to use a negative duration instrument - i.e use an IO strip

Any comment welcome?
 

FRM_Exam

Member
apologies what I meant by above is that - IO strip is an instrument (maybe that is what they refer to) which has a negative duration itself
 

emer

New Member
apologies what I meant by above is that - IO strip is an instrument (maybe that is what they refer to) which has a negative duration itself

No there is no other way. Remember mapping for options in Delta-Normal VaR and that Delta for put options is always negative.
 

FRM_Exam

Member
1)Implied volatility is assumed to be the center. What will be the effect? (out-the-money call value )
2) What did John Rusnak do? (made fake transactions to manipulate VaR)
3) What will the Q-Q plot look like? (I marked the one which was straight below 1 and then upward sloping for +ve values)
4) Backtesting, outcome (?)
5) Calculate ES for 96.5% confidence (straight-forward average of VaRs above 95.5%)
6) Netting arrangements, calculate exposure (straight-forward add all)
7) Suggest measure for specific exposure profile (Ans: credit triggers since bond ratings needs to be considerd and posting very high collateral is not feasible)
8) Calculate implied risk-free rate [using (1+rfr) = (1-PD)* (1+yield)]
9) What will be the common strike for 4 barrier options (no idea personally, I marked 42 guessed)
10) Calculate VaR for bond transition matrix (it was 9)
11) Calculate default rate for 3rd year(cum. default rate in yr 3 - cum. default rate in yr 3
12) Which of the following statements related to correlation are wrong (Ans: correlation is possible for few distributions only)
13) What is true about the ratios net stable and liquidy (liquidity ratio Is calculated for 30 days)
14) Which is the most liquid hedging option? (Eurodollar futures?)
15) What is true about ring fencing assets (allows SPE to issue debt at lower interest rates)
16) Calculate probability of default (20%)
17) Calculate option value (I didn't know that we had to calculate the probabilities of up and down moves. I marked B 0.8 something)
18) Calculate component VaR (wi*betai*VaRp)
19) Calculate hedge using keyrates (30k short and 3.5k short)
20) 5k short out-of-the-money calls, 5k ITM calls, some 8k forwards, calculate VaR (25% volatility, 252 days, daily at 99%), I got D (9500 something)
21) Calculate payment for Total Return Swap (Ans was 31.5, -40 mil and +8.5)
22) Which approach does not require correlation estimates (data driven approach bootstraapiing)
23) What is true about weighting schemes (I marked C, age weighted gives less weight to old VaRs)
24) Hedge using negative duration (Short put on IO mortgage strip?)
25) Enhancement required so that senior tranche has 90% protection (5 mil)
26) Calculate the amount of duration mismatch (700, D(liab) * liab - D(assets) * assets)
27) Which asset should be pledged as collateral? (given correlation matrix, I choose D
28) Which asset to add to the portfolio? (Nix, because it had the highest information ratio)
29) What amount of alpha is attributable to the benchmark? (0.18% found )
30) What will make it more beneficial to make an investment based on ARAROC? (Reducing the equity beta)
31) What is true about capital requirements under basel III? (Equity capital tier 1 must be 4.5%)
32) Which of the following will add to equity tier 1 capital under basel 3(common stock) 33) What are the most frequently used distributions for severity prob of default? (poisson and lognormal)
34) Which of the following is not true? (operational VaR is symmetric and fat tailed)
35) Which of the following is true? (Total risk = sum of risk contributions)
36) What is the capital requirement? (capital factor was 3, and a table of various confidence levels and VaR and SVaR was given, had to use 10 day 99% VaR, I got 340 something)
37) Difference between capital requirements based on drawdown if loan-equivalent is 0.6 something (difference was 20-30? I don't remember)
38) Something about assumptions changing (Ans: Operating risk increases/decreases) [don't remember this question properly]
40) Some big question with 2 custom formulas for calculating exposure to loans. We had to find the value of b and g? (b < 0, g >= 0)
41) Which of the following accounts for diversification? (internal models approach?)
42) There was a question on Irish and US credit crisis, it asked for what was a factor in Irish crisis but not in US.(securitization)
43) There was a question on Icelandic banking system.(CBOI giving bd loans backed by inter bank collterals)
44) What factor should have raised red flag regarding the fraud by Bernie Madoff.(performance fee)
45)Question with 2 bonds, each 100m, and they both form a pool that backs a senior/sub structure. Prob default of each bond is equal to p.
Then they have some answers that represent prob of default of the equity tranche or the senior tranche.(prob of no default on equity is (1-p)*(1-p) so default prob on equity is 1-(1-p)*(1-p)=2p-p^2
46) There was a market value of credit risk question. It gave an exposure of 100k and then a risk free bond with MV of maybe 980 or 970. Ans:7.3%
47) Hedge fund fee question that gave you a 2 and 20% fee structure and return numbers (?). There was a high watermark but no hurdle rate in place. I think I put down answer D...2.99? I guessed it right?
48) Merger acquisition question - gave 2 stocks and asked what positions you'd take. Buy target and short acqurere
49) PE fund question about the Mezz Capital strategy - sub and pref stock with no change in control
50) One question on through the cycle credit ratings and how they relate to asset appreciation and depreciation - possible solutions were along the lines of "...tightens credit when home prices depreciate..." Something like that.
51) Question on what signs show that an increase in lending (I think) isn't leading into an asset bubble (or something like that). Ans seems that asssets are backed by colletral repo
52) Convert arb question about cash inflow - answer was bond coupon
53. High threshold (converge to GEP)
54. Backtesting (95% vs 99 %)99%
55. Interest rate effect(O/p and U/P)
56. KMV
59. impact concentration to UL and EL: increase only UL
60. Which one is example of Wrong Way risk
61. Delay in Valuation (Model Risk): counter party linked
62. Liquidity Duration 2 stocks (my ans: 10 days)
63. Highest beta line in SA Approach (retail banking)
64 Liquidity (my ans : spread between short term & Treasury rates)
65. impact survivorship bias to Hedge Fund performance
66. 3 shocks (my ans : const. credit rtio for bank)
67. Asian Options: less volatile
68. Interest rate payments and prepayments: burn out effect
69: one on MBS exceess spread calculation:.15 %
70: which will causeincrease collateralizaition: excesss spread
71. @Question with LVaR and Elasticity: ans is increase Var by .24%
72 lowest and highet VaR limits for a portfolio: VaR1-VaR2 and VaR1+VaR2
73@Some question with VaR-Results calculated with historical simulation, delta normal and monte carlo simulation;
Consitency questioned by trader. What is the most plausible explanation for this inconsistency?
==> Answered: Different model assumptions;
74monte carloLpath dependent)


Was it short a put option on IO strip or go long a put option on IO strip? I am not sure - but I am seeing different option / answer choices in the previous conversations - where did the buy part come from?
 

LankyLint

Member
Delta of put is NEGATIVE

Duration of IO strip is NEGATIVE

If they wanted to hedge with a negative duration INSTRUMENT, it should be SHORT PUT ON IO STRIP

If they just wanted something with negative duration, short put (POSITIVE) * IO strip NEGATIVE = NEGATIVE, SHORT PUT ON IO STRIP

Zero coupon bond = positive duration, long put = -negative delta, overall = NEGATIVE

To me, short put sounds more reasonable since it USES a negative duration instrument and creates a negative duration hedge
 

emer

New Member
Delta of put is NEGATIVE

Duration of IO strip is NEGATIVE

If they wanted to hedge with a negative duration INSTRUMENT, it should be SHORT PUT ON IO STRIP

If they just wanted something with negative duration, short put (POSITIVE) * IO strip NEGATIVE = NEGATIVE, SHORT PUT ON IO STRIP

Zero coupon bond = positive duration, long put = -negative delta, overall = NEGATIVE

To me, short put sounds more reasonable since it USES a negative duration instrument and creates a negative duration hedge

Hi again,

I see, but there was a trick from GARP-side: It was not "short" as in "sell short" but it was "Short maturity put option on...with 1 month expiry".

Regards,
J
 

David Harper CFA FRM

David Harper CFA FRM
Subscriber
Thanks for the great feedback and input on this thread! On a positive note, it seems like the question samples were well-distributed across the topics? For example, if i read correctly, several Basel III questions, at least three current issues (Allied Irish bank, Irish vs US, Icelandic), implied volatility, even key rates queried, even Eurodollar futures made an appearance!! (I realize not to price, but you see why I included some P1 instruments in my mock?). Some immediate, specific reactions:

1.
@LeeBrittain: In regard to "David- I feel like your notes and questions prepared us well for the exam. One area that I don't think was covered in your notes but was on the exam was the question about the liquidity duration of a portfolio given certain share position sizes and % that could be sold in a day without effecting the price."
Thank you so much! But "liquidity duration" is not a concept anywhere in P2 readings. Of course "liquidity" and of course "duration." I can infer a definition from the comments here, but I'm skeptical that it deserves to be called a "duration." If anyone can cite a source, I'd love to know.

2.
About the DV01 gap (see here)
1. I thought Q.79 (Duration dollar matching qustion) had an issue with the question itself.
Liability Value = 200 M; Modified Duration = 20
Asset Value = 220 M; Modified Duration = 15
What is the gap in DV01.
Answers: 700 M, 1,000 M, 1,100 M, 1200 M
I did 700M but how can DV01 could be so high and 100's of millions.


I see conflicting phrasings of the question. If the question is the difference in DV01 (it is awkward to ask for the "gap" in DV01 [sic]), troubleshooter must be correct, the difference in DV01 implied by this setup is DV01(A) - DV01(L) = (200*20/10000) - (220*15/10000) = 0.07. I would like this question if it asked for the "net DV01" (eg). But if the answer is $700 million, that is NOT DV01, but rather the net dollar duration (aka, net value duration). Or difference/gap in value/dollar duration.

3.
I would be interested to see the full question on implied volatility, I am a little concerned about it, but it's hard to judge. It looks like the question displayed an downward sloping implied volatility skew. Skew would connote downward sloping but sounds like the plot showed anways implied volatility as a decreasing function of strike; i.e., Hull's figure 19.3. In which case:
  • If we price the option using the implied volatility given by an ATM option (i.e., compute model price using ATM implied volatility) , clearly BOTH the ITM call and OTM put will be undervalued (i.e., market price > model price). This is the point of crashophobia explanation for the equities implied volatility skew: OTM puts are bid up so their implied volatilities is higher (similar ITM call matches per put-call parity, but the explain starts with OTM puts as expensive catastrophe insurance!). Shannon has expressed this with a statement that I 100% agree with: "If the volatility of the ATM option was used to price the options, but in actuality there was a skew, this means that the ITM calls and OTM puts are both undervalued because the volatility used to price them was lower than the actual volatilities." Yes, true, always true.
  • The wording used here is crucial. Here is s statement, from above, that makes does NOT sense: "if the pricing was using implied volatility, then compared to volatility used, which one will be undervalued." It is impossible to "price an option using its own implied volatility" or, by circular definition, you just get the option's traded price. We can say, with validity, for example instead: "the (model) price of OTM put using the implied volatility of an ATM option"
4.
In regard to the negative duration/IO strip, I would be interested to confirm the question is specifically "A manager wants to hedge his fixed income portfolio with an instrument which has negative duration? Is it a) Buy the put option on the zero, b) Buy the put option on IO ..."?

With respect to INSTRUMENTS, a zero coupon bond must have positive duration (i.e., positive Mac duration = maturity); an IO strip must have a negative duration; and, to my thinking, a put on a zero must have a negative duration (rate up --> zero down --> put up); and a put on an IO strip must have positive duration (rate up --> IO strip up --> put down)

But this is why the difference between (modified, effective) duration and DOLLAR DURATION matters; it is analogous to percentage Greeks which are sensitivities [~ duration] and POSITION Greeks, which are needed for hedged [~dollar duration]. Start with simple: long a zero-coupon bond is positive duration and positive dollar duration. What is short a zero-coupon bond? The *instrument* remains positive duration, but DOLLAR (aka, value or position) duration is negative as value duration = Price * mod duration. So, short a zero coupon bond = -P * +duration = negative dollar duration; i.e., if rates go UP 1.0 unit, it gives the estimated position GAIN in short position.

So what's a put on an IO strip? I agree with emer: w.r.t IO strip, an increase in rate --> increase in price (i.e., the IO strip is the rare case of negative "instrument" duration).
  • A put on an IO strip is a derivative, an instrument itself, with positive duration (rate up --> IO strip up --> put down)
  • A long put on an IO strip has positive position duration (i.e., +Q * +D = + position duration). But a short put on an IO strip has negative position duration: -Q * +D = - negative position duration (rate up --> put down --> short put gains)
HOWEVER, I am concerned the question is imprecise, because if the question is to "hedge interest rate exposure" or "reduce exposure to interest rates," this amounts to DECREASING the absolute value of the position (value) duration (or put another way, it depends on the underlying). For example:
  • If the initial (underlying position, eg) is net long T-bonds (long dollar duration), the position is exposed (the risk) is to an increase in rates. Hedges here can include short bond, long IO strip, long IO call, or short IO put (i.e., IO put decreases in price, as rates increase).
  • But if the initial exposure is short T-bonds (negative dollar duration), the hedge can include long bond, short IO strip, short IO call, or long IO put
    (Answer B could be correct: a long IO put is a derivative on an underlying instrument [IO] with negative instrument duration but positive position duration, and that can reduced the interest rate exposure to an original net short position. Phew.:eek: ).
Thanks,
 

troubleshooter

Active Member
@ES - It was 95.5, I just wrote the questions in a hurry
Great...

@Probability of default - I remember getting 20% with another formula. Frankly, I don't even remember the formulas anymore!
I am almost certain it's less than 20% due to rfr being > 0.

@Option value - Based on what others have said, I think it was 0.52 (not sure)

@Hedge using keyrates - Sorry I meant one long one short* Only one option had the correct numerical values (d?)
Yeah. I think we both got this one down.

@Total return swap - Two libor rates were given, we had to use the one at the start of the year, not the end.
You are right mate. I screwed this one up big time. If I was taking Level I and II at the same time (I took Lvl I in 2009), I would have gotten it right. Damn... Such a blunder I am kicking myself.
 

LankyLint

Member
Don't be too hard on yourself. Everyone makes silly mistakes.

I think I used the Market Value = Face value * (1 - PD) / ( 1 + rfr)? Did you use the same formula?
 

troubleshooter

Active Member
Don't be too hard on yourself. Everyone makes silly mistakes.

I think I used the Market Value = Face value * (1 - PD) / ( 1 + rfr)? Did you use the same formula?
Yes. MV = 800,000 FV = 1,000,000

(1 - PD) / ( 1 + rfr) = 0.8
rfr has to be 0 to have PD = 20%...
rfr = 5% gives you PD = 16%
 

David Harper CFA FRM

David Harper CFA FRM
Subscriber
@Shannon: in regard to:

I have a question for David:

Is it worth writing GARP about inconsistencies or errors? With the amount of studying that I put in, if I fail because of all of the errors I will be beyond furious.

For instance the cash flow problem. If the question was "what leads to a positive cash flow?" the answer would have been obvious if you know what is going on. Asking about an "increase" in cash flows is just flat out not correct. The coupon is obviously a positive cash flow but it certainly is not an increase in cash flows.

Same thing I mentioned earlier about the whole continuous vs discrete on #4. The If we used yield = PD*LGD + r we get one answer and if we use 1+r=(1-PD*LGD)(1+y) we get another. The intro said to use continuous so I did it but seeing all of the other mistakes they made I can picture them forgetting what they actually told us to do.

I would really like to give some of them at GARP a piece of my mind.

Anyone else who thinks there were lots of errors should join me in being mad.

We deserve better.

I have three thoughts:
  1. Yes, absolutely: In my opinion, there can be no doubt about your question: given the time and expense involved (from a pure customer perspective) and the importance of the exam's reputation and its long-run health, I consider this sort of feedback necessary, appropriate and fully justified
  2. HOWEVER, first step should be to give the benefit of the doubt: nobody has the actual exam documents; even comments on this thread show how recollections vary; and the wording/phrasing can turn the difference (a single few words misunderstood can confuse an otherwise good question). I would suggest, in most cases, lacking the full question, there may not be enough information?
  3. I will be glad, if you like, to itemize the "dubious questions" for organized submission, to complement anything you submit? I can start a thread in the same forum here http://forum.bionicturtle.com/forums/2012-errata-and-garp-2013-frm-recommendations.76/
    i.e., I was already going to submit, to GARP, an updated itemization of issues with the 2012 Practice Exams, I can add an itemization of dubious exam questions ... but i think if the criteria is high confidence, so far it's a short list?
Let me know? Thanks!
 

troubleshooter

Active Member
Thanks for the great feedback and input on this thread! On a positive note, it seems like the question samples were well-distributed across the topics? For example, if i read correctly, several Basel III questions, at least three current issues (Allied Irish bank, Irish vs US, Icelandic), implied volatility, even key rates queried, even Eurodollar futures made an appearance!! (I realize not to price, but you see why I included some P1 instruments in my mock?). Some immediate, specific reactions:

1.
@LeeBrittain: In regard to "David- I feel like your notes and questions prepared us well for the exam. One area that I don't think was covered in your notes but was on the exam was the question about the liquidity duration of a portfolio given certain share position sizes and % that could be sold in a day without effecting the price."
Thank you so much! But "liquidity duration" is not a concept anywhere in P2 readings. Of course "liquidity" and of course "duration." I can infer a definition from the comments here, but I'm skeptical that it deserves to be called a "duration." If anyone can cite a source, I'd love to know.

2.
About the DV01 gap (see http://forum.bionicturtle.com/threads/level-2-post-what-your-remember-here.5923/page-2#post-17598)
1. I thought Q.79 (Duration dollar matching qustion) had an issue with the question itself.
Liability Value = 200 M; Modified Duration = 20
Asset Value = 220 M; Modified Duration = 15
What is the gap in DV01.
Answers: 700 M, 1,000 M, 1,100 M, 1200 M
I did 700M but how can DV01 could be so high and 100's of millions.

I see conflicting phrasings of the question. If the question is the difference in DV01 (it is awkward to ask for the "gap" in DV01 [sic]), troubleshooter must be correct, the difference in DV01 implied by this setup is DV01(A) - DV01(L) = (200*20/10000) - (220*15/10000) = 0.07. I would like this question if it asked for the "net DV01" (eg). But if the answer is $700 million, that is NOT DV01, but rather the net dollar duration (aka, net value duration). Or difference/gap in value/dollar duration.

3.
I would be interested to see the full question on implied volatility, I am a little concerned about it, but it's hard to judge. It looks like the question displayed an downward sloping implied volatility skew. Skew would connote downward sloping but sounds like the plot showed anways implied volatility as a decreasing function of strike; i.e., Hull's figure 19.3. In which case:
  • If we price the option using the implied volatility given by an ATM option (i.e., compute model price using ATM implied volatility) , clearly BOTH the ITM call and OTM put will be undervalued (i.e., market price > model price). This is the point of crashophobia explanation for the equities implied volatility skew: OTM puts are bid up so their implied volatilities is higher (similar ITM call matches per put-call parity, but the explain starts with OTM puts as expensive catastrophe insurance!). Shannon has expressed this with a statement that I 100% agree with: "If the volatility of the ATM option was used to price the options, but in actuality there was a skew, this means that the ITM calls and OTM puts are both undervalued because the volatility used to price them was lower than the actual volatilities." Yes, true, always true.
  • The wording used here is crucial. Here is s statement, from above, that makes does NOT sense: "if the pricing was using implied volatility, then compared to volatility used, which one will be undervalued." It is impossible to "price an option using its own implied volatility" or, by circular definition, you just get the option's traded price. We can say, with validity, for example instead: "the (model) price of OTM put using the implied volatility of an ATM option"
4.

In regard to the negative duration/IO strip, I would be interested to confirm the question is specifically "A manager wants to hedge his fixed income portfolio with an instrument which has negative duration? Is it a) Buy the put option on the zero, b) Buy the put option on IO ..."?

With respect to INSTRUMENTS, a zero coupon bond must have positive duration (i.e., positive Mac duration = maturity); an IO strip must have a negative duration; and, to my thinking, a put on a zero must have a negative duration (rate up --> zero down --> put up); and a put on an IO strip must have positive duration (rate up --> IO strip up --> put down)

This is why the difference between (modified, effective) duration and DOLLAR DURATION matters; it is analogous to percentage Greeks which are sensitivities [~ duration] and POSITION Greeks, which are needed for hedged [~dollar duration]. Start with simple: long a zero-coupon bond is positive duration and positive dollar duration. What is short a zero-coupon bond? The *instrument* remains positive duration, but DOLLAR (aka, value or position) duration is negative as value duration = Price * mod duration. So, short a zero coupon bond = -P * +duration = negative dollar duration; i.e., if rates go UP 1.0 unit, it gives the estimated position GAIN in short position.

So what's a put on an IO strip? I agree with emer: w.r.t IO strip, an increase in rate --> increase in price (i.e., the IO strip is the rare case of negative "instrument" duration). But a long put on an IO strip is an instrument with positive duration, similar to a long bond position: an increase in rate --> decrease in put value. So, long IO put should have positive duration.

HOWEVER, I am concerned the question is imprecise, because if the question is to "hedge interest rate exposure" or "reduce exposure to interest rates," this amounts to DECREASING the absolute value of the position (value) duration (or put another way, it depends on the underlying). For example:
  • If the initial (underlying position, eg) is net long T-bonds (long dollar duration), the position is exposed (the risk) is to an increase in rates. Hedges here can include short bond, long IO strip, long IO call, or short IO put (i.e., IO put decreases in price, as rates increase).
  • But if the initial exposure is short T-bonds (negative dollar duration), the hedge can include long bond, short IO strip, short IO call, or long IO put
    (Answer B could be correct: a long IO put is a derivative on an underlying instrument [IO] with negative instrument duration but positive position duration, and that can reduced the interest rate exposure to an original net short position. Phew.:eek: ).
Thanks,

For 2 above, I am pretty sure they said DV01 gap. And it is just plainly wrong choices starting at 700 M. I think GARP is a bit amateurish compared to CFA Institute to be honest. In all three levels of CFA, I never found a single question to be erroneous. Back in 2009 when I did Lvl 1, they had one question that was sort of impossible binomial tree and other one (box spread) was not in course. I think they should get their QA doing their jobs..

For 3, No, David. Quesiton was pretty specific in saying that the imp. volatility from the lower side is being used to price all the options... I did the out of the money put but I think I got it wrong. But it is quite possible that there was more than one possible option. If someone remebers the question correctly, should write to GARP if there was an issue.

For 4, your first paragraph is accurate. But they made it confusing by saying that Short maturity put on IO Strip, which sounds like you are shorting the put but you are actully going long on "short maturity" IO Strip. I find this kind of questions silly if you are trying to test someone's knowledge. I think most of folks know that IO strip has -ve duration but would have gotten it wrong with misleading wording like "Short matury put on IO Strip" rather than "Buy a short maturity put on IO Strip" or something like that or just delete the word short as it can be confused with going short...
 
Top