Brainscape's Knowledge GenomeTM

Browse over 1 million classes created by top students, professors, publishers, and experts.


See full index

Section 4: Liquidity Risk > Credit Portfolio Management: Ch 6 > Flashcards

Credit Portfolio Management: Ch 6 Flashcards

1
Q

Option Adjusted Spread (OAS)

A

Measurement of the spread of a fixed-income security rate and the risk-free rate
of return, which is adjusted to take into account an embedded option

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

Uses of Spread Decomposition

A
  • Can help portfolio managers understand spread movements better
  • Investors can use spread decomposition for:
    • Hedging
    • Forecasting future OAS changes
    • Developing alpha strategies
  • The portfolio strategy will depend on the reason why the credit spread moved. For example, whether the wide spreads are due to large expected default losses, high liquidity cost or a high-risk premium
  • The portfolio can ride out periods of high liquidity cost and risk aversion if a buy-and-hold strategy is used.
  • If the wide spread reflects an increase in expected default losses, you may need to reposition or hedge its portfolio.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

Risk Premium in Spread Decomposition

A
  • Represents the market-level risk premium and not a risk premium specific to each bond
  • Represents the variation of the market risk premium that is not already embedded in the other variables (e.g. LCS, CDS)
  • When the intercept explains a relatively high proportion of OAS, it suggests that systematic factors rather than bond-specific factors are driving spreads
    • When intercept getting larger and coefficients getting smaller
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

Expected Default Cost in Spread Decomposition

A
  • Part of the OAS is driven by the possibility of default and the recovery level of defaults
  • Two Models used to quantify expected default cost: CDS and CDP/CRR
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

Credit Default Swap (CDS) Model

A
  • Expected Default Costit = CDSit
  • Use market-quoted five-year CDS as a measure of expected default cost
  • Generally the five year point is most liquid
  • Restrict analysis to CDX Index
    • CDX Index - an index comprised of the most liquid North American entities that trade in the CDS market
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

CDP/CRR Model

A
  • Expected DefaultCostit = CDPit * (1 - CRRit)
  • CDP = Conditional Default Probability
  • CRR = Conditional Recovery Rate
  • CDP, CRR are not market variables
  • Calculated from a firm-specific macroeconomic model
  • Computed independently of a bond’s OAS
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

Expected Liquidity Cost

A
  • The Expected Liquidity Cost calculated will use the Liquidity Cost Score (LCS) model from Chapter 5
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

Overlap b/w Expected Default Cost and Expected Liquidty Cost

A
  • Another portion of the OAS is from the degree of uncertainty associated with the timing, magnitude, and recovery of defaults and liquidity costs
  • The author states that credit bond spreads are generally much larger than is justified by their subsequent default and recovery experience
  • If an investor seeks to hedge the default or liquidity components separately, then the contribution to OAS in bps is relevant
  • If an investor is analyzing current market compensation for taking on additional amounts of expected default or liquidity cost, the coefficients provide that information
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

Spread Decomposition Methodology/Models

A
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

Interpreting Results of Spread Decomposition Methodology

A
  • First add the expected default cost component through CDS or CDP/CRR
  • Then add LCS
  • Should see that adding LCS does not substantially change the CDS or CDP/CRR coefficient and an improvement in R-squared if LCS is a useful explanatory variable
  • Expect statistically signficant coefficients both with positive signs for beta
    and lamda
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

Multicollinearity in Spread Decomposition

A
  • When regressors are added in one-by-one in a multi-step regression analysis, the coefficients should not change substantially when new regressors are added.
  • For example, the CDS coefficient should be similar in the CDS-only model and the LCS & CDS model.
  • If there are substantial changes to coefficient values of existing regressors when new regressors are added, this could indicate multicollinearity problems.
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

Disadvantages of the Spread Decomposition Methodology

A
  • Liquidity and default are unlikely to be completely independent of each other, so multicollinearity may be a concern. Recall that multicollinearity is when two or more predictor variables in a regression model are highly correlated
  • If the CDS model is used, the CDS market may not necessarily be liquid and therefore cannot always be considered as a pure default proxy
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

Advantages and Disadvantages of CDP/CRR Model

A

Advantages of CDP/CRR Model:

  • CDP and CRR are independent of market spreads
  • Larger sample, because CDP/CRR data spans over a larger supply of tickers than the number of CDS in the CDX

Disadvantages of CDP/CRR Model:

  • Market CDS spreads are likely to be more closely related to OAS than to a modeled default probability estimator. This advocates the CDS model over the CDP/CRR model
  • In the dataset the author uses, the CDP/CRR coefficient changes (more significantly than the CDS model) when LCS is included as an additional variable.
    • stronger multicollinearity b/w CDP/CRR and LCS
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

High Yield Spread Decomposition

A
  • Compared to investment grade bonds, default risk will play a relatively bigger role for the spread decomposition of high yield bonds
  • Liquidity cost provides additional explanatory power in high-yield spreads only during period of market stress in the data set the author analyzes
  • The intercept is larger for high yield, suggesting that risk aversion, unrelated to liquidity/default characteristics, may drive a large proportion of yield spreads
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

Applications of Spread Decomposition: Ex-Ante Analysis

A
  • Bond’s market OAS can be compared with estimated OAS using the parameters from the spread decomposition model
  • If the actual OAS is wider than the estimated OAS, the bond may be trading too wide. This difference will be captured in the initial residual ˆhit
  • This could cause a trading signal to reverse the mispricing and (theoretically) more closely align the actual and estimated OAS
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

Applications of Spread Decomposition: Hedging

A
  • Hedging could be added to help neutralize spread changes
  • Bond default could be hedged through the issuer’s CDS
  • Liquidity risk could be hedged through VIX futures
  • Neither is perfect (or close to perfect). The book simulates a hedging strategy, and the results reduce volatility somewhat - but there is still substantial portfolio volatility. One explanation is basis risk
  • Hedging is employed to neutralize spread changes, so changes in OAS/CDS/LCS are relevant
17
Q

Alternative Spread Decomposition Models - I

A
  • Bond-level liquidity risk premium - the additional premium investors demand as compensation for the risk that the actual liquidity may be different than the expected liquidity cost as measured by LCS
  • Can use bond’s LCS volatility over the past 12 months to approximate the bond-level liquidity risk premium level
  • The bond-level liquidity risk premium is captured in the LCSVolit term
  • Results show that the LCSVolit term is significant, but also that it absorbs about 20% of the LCS coefficient
18
Q

Alternative Spread Decomposition Models - II

A
  • Another specification is to use the changes in LCS and CDS spreads
  • This suggests that changes in liquidity and default measures directly impact contemporaneous returns
19
Q

Alternative Spread Decomposition Models - III

A
  • One might guess that outliers could drive the results seen in this reading. The author uses log-spreads to see if the results change substantially. If they do, this could suggest that outliers were driving the results
  • The conclusions are the same with both the general and log-spread models, indicating that outliers are not driving the results

Section 4: Liquidity Risk (7 decks)

Brainscape helps you reach your goals faster, through stronger study habits.
© 2024 Bold Learning Solutions. Terms and Conditions