Exam Feedback November 2015 Part 2 FRM Exam Feedback

[Tanya Vasileva]

@tanya: could you tell me the exact answer for the jensens inequality question as I am getting only 0.47bps and not any of the answers even going by the formula..

with the example substitute for u and d 6% and 4 % and instead of 10% the exam gave 5%
and then take the # between left and right side

 

[a.lesnar]

Good morning
There was a question at the beginning of the exam concerning a firm in financial distress, it was asked whether senior and junior debt would increase or decrease.
I cannot remember the question word by word, but there was no indication on whether correlation or default probability would change or not (they were therefore constant).
Basically speaking if I remember correctly, we were asked to state how would senior and subordinated debt react when a firm goes into financial distress.
I was tempted to choose increase/decrease, but I selected decrease/decrease, although I cannot find explicit evidence in the theory for this.
I just assume investors decrease debt holdings for distressed firms by logic. There was no shift in correlation to make me infer that senior debt would increase.
Why would there be more demand for senior debt for a firm that is going in worse conditions, if the investor can simply go safe and flight to quality (and purchase bond from safer issuers)?
Thanks

 

[Toy]

a.lesnar, I think when a firm is in distress, the subordinated debt behaves like equities. When it is in good times, it behaves like bonds. So I think chose that senior debt will decrease and subordinated debt will increase. I think it is the behavior of the subordinated debt that is ambiguous. Not 100% sure my answer is correct and I have not gone back to check.

 

[Toy]

I meant to write "when it is in good times it behaves like senior bonds"

 

[Tanya Vasileva]

yes, in fin. distress the subordinated acts like equity, so decrease and senior increase

 

[a.lesnar]

I also selected increase senior/decrease subordinated, but my doubt concerned the fact that there was no assumption on correlation or probability of default on the question. Thus I changed my answer to decrease/decrease, since I am not totally convinced by logic that the senior debt would be positively impacted by a negative turmoil in the company. The semantics of a couple of questions were in my opinion a bit confusing

 

[Toy]

This is a question from the Stulz reading. This is not from the readings on the tranches of CDO (where correlation between tranches is a factor).

In fact Bionic Turtle appears to address this. Here is a question from Stulz:


404.2. Assume a firm with only two classes of debt: senior debt consisting of a zero-coupon bond with face value of $100.0 million; and subordinated debt consisting of a zero-coupon bond with face value of $50.0 million. Consider the following two statements:

I. An increase in firm (asset) volatility necessarily implies a decrease in the value of the subordinated debt

II. As time to maturity decreases, the value of the subordinated debt necessarily increases (pulls to par)

Under the Merton model for credit risk, which is (are) true statements?

a. Neither

b. I. only

c. II. Only

d. Both are true


Here is Bionic Turtle's answer:


404.2. A. Neither. In Stulz's model, subordinated debt is valued as the difference between two call options, c(V,F,T) - c(V,F+U,T). The impact of volatility and maturity is therefore ambiguous and tends to be different for a high-value firm (when subordinated debt act more like debt: increase in volatility tends to decrease value and decrease in maturity tends to increase price via pull to par) than for a low-value firm (when subordinated debt acts more like equity: increase in volatility tends to increase value and decrease in maturity tends to decrease value)

 

[Jo_]

yes I understand and I agree on the formula, but the variable on the axis was (1-benefits) as expressed in the problem, that is why I chose answer D (decreasing line) instead of B

That was my reasoning also. I guess we all agree on the concept, it's a question of how the Y-axis was defined.

 

[dundaking]

I think I chose both decrease. I kinda remember something about how mezzanine debt at high levels of default rates acts like senior bonds when pd increases holding correlation constant. The opposite is true i.e. it acts like equity when pd rates are low? Given t was in 'times of distress' I therefore took the assumption pd rates were high...

 

[dundaking]

For the benefits of diversification, this is what annoys me about GARP. The quesion doesn't really test your understanding of risk management. It's one of their 'sneaky' questions for want of a better term. Any risk manager should know that diversification benefits plateaus with an increase in the number of securities held in the portfolio. The introduction of ambiguity into the question where diversification benefits was defined as being 1-netting factor is unnecessary in my opinion... Oh well

 

[Twol84]

Yeah GARP is very far behind CFAI in the ways questions are worded, practice papers are designed, and sometimes even on how topics are chosen. Tuckman deserved at least ONE question outside of Jensen's inequality.

I think that question wanted to test that the candidate understood that netting benefits are finite. D (the decreasing graph) depicts the marginal benefits of netting, but not the actual benefits, which increase before hitting a constant peak.

 

[Toy]

with respect to this:

dundakingNew Member
New
I think I chose both decrease. I kinda remember something about how mezzanine debt at high levels of default rates acts like senior bonds when pd increases holding correlation constant. The opposite is true i.e. it acts like equity when pd rates are low? Given t was in 'times of distress' I therefore took the assumption pd rates were high...

This is different from CDO tranches (where you have senior tranches, mezzanine and equity tranches). Trading in CDO tranches is essentially taking a view on how correlation moves. The dynamics of pricing debts within the capital structure of a single bank in the Merton approach is different.

 

[Esbringa]

There is a table in the swechesser where is explained:

If volatility raise:

Equity increase
Senior Debt decrease
Subordinated Debt increase / decrease depends on the firm value.

 

[Toy]

Does anybody remember the exact question on the netting benefit? I can't. I thought question was about netting benefit and chose (b). If the question was about netting factor, then
the netting factor should decrease as the number of trades netted increase, all things being equal. However if the question was about the relationship between netting benefit and number of trades netting set, then netting benefit will increase with the number of trades in the netting set and tapers off. (Bionic Turtle reading on netting, note that I could not copy the BT formula and manually typed it in).



Netting factor


A majority of netting may occur across instruments of different asset classes that may be

considered to have only a small correlation. One should note that this would still create a

positive benefit. We derive the following formula for the “netting factor” with respect to

exposure under the assumption that future values follow a multivariate normal distribution:



Netting factor = {sqrt(n+ (n-1)p)}/n








  

Where (n) represents the number of exposures and rho-bar is the average correlation. The

netting factor represents the ratio of net to gross exposure and will be +100% if there is no

netting benefit and 0% if the netting benefit is maximum.

 

[Sidewinder]

Don't know the question exactly but I thought to have in mind that the y-axis was "Diversification benefit"

 

[a.lesnar]

Don't know the question exactly but I thought to have in mind that the y-axis was "Diversification benefit"

yes exactly that's why I corrected from B to D (where the line decreases): if I remember correctly, the Y-axis was explicitly expressed as (1-netting factor).
Concerning my previous statement on the question for the financially distressed firm, I found at page 138 of Schweser material book 2 a table (Figure 3): when the probability of default increases (holding correlation costant) the Credit VaR of senior debt increases. However, both senior and subordinated mean values decrease.
If I remember correctly, the question asked for the value of the bonds, not the credit VaR, am I eventually right?
Thanks

 

[William_Jose]

yes exactly that's why I corrected from B to D (where the line decreases): if I remember correctly, the Y-axis was explicitly expressed as (1-netting factor).
Thanks

Since the Y-axis mentions benefit=1-netting factor, the choice will be (b),ie., the curve first rises exponentially and then flattens. If it was only netting-factor, the plot would be like (d). You can check the plot by entering some numbers in excel,entering that formula in a column and then plotting that (taking correlation rho as some arbitrary constant value, say, 0.5).

Concerning my previous statement on the question for the financially distressed firm, I found at page 138 of Schweser material book 2 a table (Figure 3): when the probability of default increases (holding correlation costant) the Credit VaR of senior debt increases. However, both senior and subordinated mean values decrease.
If I remember correctly, the question asked for the value of the bonds, not the credit VaR, am I eventually right?

Yes, it asked about how values of subordinate debt and senior debt would behave. Subordinate debt would behave like equity and hence would decrease with decrease in volatility and senior debt value would increase with fall in firm volatility.

 

[William_Jose]

with the example substitute for u and d 6% and 4 % and instead of 10% the exam gave 5%
and then take the # between left and right side

In this problem, I did ((1/1.04+1/1.06)*.5)/1.05=0.907. The zero rate comes out to be around 4.9% and hence the convexity is just .47 bps. I was hence not getting either 0.84 or 1.82 bps in the exam and this just threw me off for a while. Please clarify.

 

[a.lesnar]

Since the Y-axis mentions benefit=1-netting factor, the choice will be (b),ie., the curve first rises exponentially and then flattens. If it was only netting-factor, the plot would be like (d). You can check the plot by entering some numbers in excel,entering that formula in a column and then plotting that (taking correlation rho as some arbitrary constant value, say, 0.5).

Yes, it asked about how values of subordinate debt and senior debt would behave. Subordinate debt would behave like equity and hence would decrease with decrease in volatility and senior debt value would increase with fall in firm volatility.

Thanks for the information, but I am convinced there was no hypothesis on the volatility, neither on correlation. The question just stated that the firm went into financial distress. I just found out that in Schweser material the mean value of both senior and subordinated bonds decrease (page 138 book 2). The Credit VaR of the senior bonds increase, but their mean value decreases, If I remember correctly the Credit VaR was not cited in the question
Thanks

 

[Tanya Vasileva]

In this problem, I did ((1/1.04+1/1.06)*.5)/1.05=0.907. The zero rate comes out to be around 4.9% and hence the convexity is just .47 bps. I was hence not getting either 0.84 or 1.82 bps in the exam and this just threw me off for a while. Please clarify.

the right side is 1/1.05^2 = 0.90702948
the left side is 0.5*(1/1.04+1/1.06)/1.05=0.90711176