Machine Learning Fair Value Corporate Bond Strategy
The investment universe consists of the Intercontinental Exchange global investment grade (G0BC) index constituents. Only bonds denominated in USD, bonds from the Financial, Utility, or Industrials sectors are included. Remove Cash, Quasi & Foreign Government, Covered/Securitized bonds. Each month, estimate the fair value of the option-adjusted spread (OAS).
ASSET CLASS: bonds | REGION: United States | FREQUENCY: Monthly |MARKET: bonds | KEYWORD: Corporate Bonds
I. STRATEGY IN A NUTSHELL
Universe: USD-denominated IG bonds (Financials, Utilities, Industrials) from ICE G0BC index, excluding government and securitized bonds. Estimate fair OAS using ridge + adaptive LASSO regressions on bond and issuer characteristics. Value signal = market OAS – fair OAS. Long top quintile, equally weighted, monthly rebalanced.
II. ECONOMIC RATIONALE
Captures undervalued bonds: market OAS > fair OAS signals underpricing. The strategy relies on robust ML-based fair value estimation, using multiple financial and market variables to identify mispriced bonds reliably.
Hendrik Kaufmann, Deka Investment GmbH; Philip Messow, Robeco Asset Management; Frederik Wisser, Quoniam Asset Management GmbH; [Next Author], EBS Universität für Wirtschaft und Recht – EBS Business School – Department of Finance and Accounting
<Abstract>
Leveraged ETFs and market makers who are active in option markets must adjust imbalances arising from market movements. Establishing delta-neutrality may cause either return momentum or reversal depending on the sign and size of the imbalance vis-a-vis market prevailing liquidity. We find that a large and negative (positive) aggregated gamma imbalance, relative to the average dollar volume, gives rise to an economically and statistically significant end-of-day momentum (reversal). We compare this channel to the rebalancing of leveraged ETFs and find that the effect generated by leveraged ETFs is economically larger. Consistent with the notion of temporary price pressure, the documented effects quickly revert at the next day’s open. Information-based explanations are unlikely to cause the results, suggesting a non-informational channel through which leveraged ETFs and option markets affect underlying stocks towards the market close.
