AAL A&B Tranche LiftIndex Analysis
Historical LiftRatio Analsysis
| Expected Return for Scenario | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| EETC | Tranche | Treasury | Yield Spread | Expected Spread | Combined | Credit | Asset | Lift Ratio | ||
| AAL_16-1 | A | 2.25 | 1.27 | 1.27 | 0.05 | 0.02 | 0.46 | 2.77 | ||
| AAL_16-1 | B | 1.89 | 2.41 | 2.41 | 0.04 | 0.09 | 0.32 | 7.62 | ||
| AAL_16-2 | A | 2.21 | 1.27 | 1.27 | 0.05 | 0.02 | 0.3 | 4.28 | ||
| AAL_16-2 | B | 1.85 | 2.43 | 2.43 | 0.01 | 0.1 | 0.17 | 14.25 | ||
Value is a relative thing for risky assets and a one measure used by investors is the Sharpe Ratio where you divide the expected annual return by the annual standard deviation of return to get a measure of the amount of income received per unit of risk. This works in one dimension easily but it’s complicated when looking at the output from LiftRisk for EETC’s because there are two dimensions for risk: credit and asset. In order to asses this Ruxton developed the LifRatio which looks at equally likely 3 scenarios in the “bad” region of outcomes: depressed asset value with base credit assumptions (which we call Asset SD), increased default rates with base asset assumptions (which we call Credit SD), and depressed asset values with increased default rates (which we term Equal Weight SD). We use a 1 standard deviation in the asset and credit dimension as the base in order to make it somewhat comparable to Sharpe. We then run the three scenarios and to determine which produces the largest drop in income and then use that as the denominator to divide the expected return to get income per unit of risk. We call this the LiftRatio and, just like Sharpe Ratio, the higher the better.
Recently, the historical vol scenario has the B tranches outperforming the A’s on a relative risk vs. return basis by multiples of return. Some of this is reduced risk at the front end due to a lower volatility, and some is due to the upward sloping yield curve. The seems preferred over A on a risk vs return basis but, on the other hand, if the investor values duration or extra seniority initially, the A Tranche might be preferred for duration.
Time Behavior of B vs A Tranche
The graph shows the B structure showing more expected loss given default at the beginning of the transactions life, but his is also when the periodic default rates tend to be lower. This is seen by the size of the bubbles, which are calibrated to the number of defaults that experience a payout in the simulation analysis. The number of payouts for the A tranche are lower, as expected, through year 7, but then there is the cumulative affect of the additional 5 years.
Total Return Performance of the EETC’s
| Expected Return for Scenario | ||||||
|---|---|---|---|---|---|---|
| EETC | Tranche | Yield | Credit | Asset | Combined | |
| AAL_16-1 | A | 3.53 | 3.51 | 3.07 | 3.48 | |
| AAL_16-1 | AA | 3.41 | 3.41 | 3.39 | 3.41 | |
| AAL_16-1 | B | 4.3 | 4.21 | 3.99 | 4.27 | |
| AAL_16-2 | A | 3.48 | 3.46 | 3.18 | 3.43 | |
| AAL_16-2 | AA | 3.31 | 3.31 | 3.3 | 3.31 | |
| AAL_16-2 | B | 4.27 | 4.18 | 4.1 | 4.26 | |
The absolute outperformance in the expected case in all of the scenarios implies that we would be better served owning the B tranche over the A or AA for total return, though capital or duration factors might take alter the preference.