Diving into Predictive Markers of Corporate Failure. Martin M. Zorn Tuesday, June 12, :00 to 10:30am Session 27040

Transcription

1 Diving into Predictive Markers of Corporate Failure Martin M. Zorn Tuesday, June 12, :00 to 10:30am Session 27040

2 Macro Factors A Risk Road Map Default Prepayment Mortality Spreads Cash flows Market values GAAP, IFRS9, and CECL Value at risk and own default probability Credit and interest rate hedges Risk neutral valuation of all assets & liabilities Trading volume, if any Regulatory reporting 2

3 Reduced Form Model Application Orange: 1 Year PD, Green: 10 year PD, Blue: Traded Credit Spreads 3

4 Credit Model Synthesis Logistic Regression is the Maximum Likelihood Estimation for 0/1 Default/No Default P t = 1 1+e α σ n i=1 βi X i Altman s Z-score can be thought of as the linear equation in this exponent. The Merton probability of default can be an incremental explanatory variable. Other Merton inputs like the value and return volatility of company assets can be explanatory variables as well. Reduced Form Models can take the explanatory variables of any other model into account. It is an all encompassing general approach to credit analytics. 4

5 Short-term Risk is Benign Longer-term Risk Should Be a Concern 5

6 The Markers of Corporate Failures Part I Net income/assets Cash/assets Working capital/assets Equity returns Equity returns relative to the market index Size Age of firm Trailing industry default rate EBITDA/assets Financial ratios relative to implied market value of assets Implied firm asset values Implied firm asset volatility Industry Location Time of year Any other financial ratios GDP growth Unemployment rate Volatility Term structure of interest rates VIX index Case-Shiller home price index Oil prices Gold prices Foreign exchange rates S&P 500 index returns Russell 2000 index returns TSX index returns FTSE index returns Nikkei index returns Any other macro factor 6

7 Credit Formulas Out with the Old In with the New Hilscher, Jens, Robert A. Jarrow, and Donald R. van Deventer, The Valuation of Corporate Coupon Bonds, Cornell University and Kamakura Corporation Memorandum, March 11, 2018.

8 The Markers of Corporate Failures Part II Inputs Used Daily term structure of U. S. Treasury zero coupon bond yields Daily term structure of KRIS default probabilities Actual traded bond NPVs (price + accrued interest) Bond-implied recovery rate (derived) Bond-implied liquidity discount (derived) Inputs Avoided Credit Ratings Credit Scores CDS traded prices (not disclosed and subject to manipulation) CDS volumes (not disclosed) Yield to maturity (flawed model) Credit Spreads (no explanatory power) Merton Model (outdated)

9 Spread and Default: Lehman Brothers Reminder In contrast to spreads, bond prices for Lehman on the day bankruptcy was filed had already converged to an average of with a standard deviation of This pattern can be seen over and over again in the bond market.

10 It was not just Lehman Brothers- SHLD

11 Methodology for Reduced Form Modeling Strategies Digital Coupon Securities A building block coupon security pays $1 on the coupon payment date if and only if the company has not gone bankrupt before then, zero otherwise. Digital All or Nothing Principal Security The all or nothing principal security pays $1 on the scheduled maturity date if and only if the company has not gone bankrupt before then, zero otherwise Digital Recovery Securities The recovery security for period j during the bond s lifetime pays $1 at the end of period j if and only if the company has not gone bankrupt before the start of period j and does go bankrupt during period j, zero otherwise.

12 Accuracy in Fitting Traded Bond Prices: Issuer Level

13 Examples

14 To Manage the Risk You Must Understand the Risk

15 Consistent Risk Management Decisions Reduced Risk Management Costs Comprehensive Accounting, Regulatory and Disclosure Compliance Improved Returns 15 Copyright 2018, All Rights Reserved, Kamakura Corporation