Market Tremors: Quantifying Structural Risks in Modern Financial Markets

Description

Since the Global Financial Crisis,  the structure of financial markets has undergone a dramatic shift. Modern markets have been “zombified” by a combination of Central Bank policy, disintermediation of commercial banks through regulation, and the growth of passive products such as ETFs.  Increasingly, risk builds up beneath the surface, through a combination of excessive leverage and crowded exposure to specific asset classes and strategies.  In many cases, historical volatility understates prospective risk. This book provides a practical and wide ranging framework for dealing with the credit, positioning and liquidity risk that investors face in the modern age.  The authors introduce concrete techniques for adjusting traditional risk measures such as volatility during this era of unprecedented balance sheet expansion. When certain agents in the financial network behave differently or in larger scale than they have in the past, traditional portfolio theory breaks down.  It can no longer account for toxic feedback effects within the network.  Our feedback-based risk adjustments allow investors to size their positions sensibly in dangerous set ups, where volatility is not providing an accurate barometer of true risk. The authors have drawn from the fields of statistical physics and game theory to simplify and quantify the impact of very large agents on the distribution of forward returns, and to offer techniques for dealing with situations where markets are structurally risky yet realized volatility is low.  The concepts discussed here should be of practical interest to portfolio managers, asset allocators, and risk professionals, as well as of academic interest to scholars and theorists. Acknowledgements Contents List of Figures List of Tables 1 Introduction “Zombification” of Modern Markets The Challenge to Investors An Analogy with Waiting Times Qualitative Features of Zombification The Dilemma for Institutional Investors External and Network Risks Vulnerability Not Predictability The Two Asset Base Case Textbook Description of Risk A Parallel Universe The Full Network Model Steering a Middle Course 2 Financial Networks in the Presence of a Dominant Agent Reviewing the Core Ideas Contrast with Reactive Approaches An Emphasis on Practicality Buying Options in Fragile Markets The Standard Approach to Modeling Risk The Mean Field Interpretation Indistinguishability: A Hidden but Crucial Assumption From Mean Fields to the Entire Financial Network Computational Issues in the Full Network Model Mean Field–Majority Agent Interactions Quantifying Feedback from the Dominant Agent The Basic Mean Field—Majority Agent Algorithm Estimating Price Impact: Practical Considerations Properties of the Modified Distribution Reflections on Parameter Stability Contrast with Impact-Adjusted VAR Algorithm Requirements Searching for Dominant Agents Other Likely Culprits: Banks, Dealers and Market Makers More Reflections on the Central Banks Conclusions Addendum: Complimentary Models to Ours The JLS Model Limitations of the LPPL Model Mean Field Games with a Majority Player 3 Exchange-Traded Products as a Source of Network Risk Easy Access for the Retail Investor Premium/Discount to NAV Major Players in the ETF Space How the Creation/Redemption Process Works Where It Can All Go Wrong ETNs: Debt Trading on Equity Exchanges Safety Considerations Levered Exchange-Traded Products Where is the Bid? Underperformance of Leveraged ETFs and ETNs Path Dependence over Longer Horizons Conclusions Endnote: The Effect of Mean Reversion on Levered and Inverse Exchange-Traded Products 4 The VIX “Volmaggedon”, with Exchange-Traded Notes Destabilizing the Market Initial Demand for Long VIX ETNs VXX Implementation Details Sound Reasons for Launching the VXX Impact of the VIX Futures Term Structure on Carry Costs of Rolling a Long VIX Futures Position Speculators Complete the Market Short VIX ETNs Enter the Fray The Pre-conditions for a Crisis How Many Contracts to Buy, Conditional on a Volatility Spike The TAS Order Book Price Impact: Relevance of the Market Microstructure Literature Price Impact: A Matter of Perspective Specifying the Price Impact Curve More Insights into Square Root Impact A Surprising Analogy Where Cost and Mark to Market Risk Converge Estimating Follow Through During the Volmageddon The XIV Goes to 0 on Volmageddon Day Strong Conclusions, with a Few Caveats 5 Liquidity Fissures in the Corporate Bond Markets Dangers in the Bond-ETF Feedback Loop Framing the Problem for High Yield ETFs OOM Estimates for the US High Yield Market The ETF Premium/Discount: A Potential Sign of Instability Sketching Out a Danger Scenario Specifying the Mean Field Dominant Agents in the Corporate Bond ETF Space Forces Driving an Extreme Discount to NAV Limit up and Down Thresholds for Various Securities Deciding How Many Shares to Buy Contraction in Dealer Balance Sheets Interpreting the Order Book Data Correctly A Survey Approach to Estimating Impact Toxic Leakage into Mutual Funds Flow-Performance Curves for Bond Funds Quantifying the Relationship Between Performance and Flows Expected Outflows in State 8 Postscript: Tackling the High Yield Bond—ETF—Mutual Fund Feedback Loop During the Covid-19 Crisis in Q1 2020 6 Market Makers, Stabilizing or Disruptive? Part One: The Corona Sell Off and the GEX Institutional Demand for Bond-Like Equity Structures The Impact of Dealer Hedging on Price Action Aggregate Gamma Exposure and the GEX Potential Hotspots for the GEX Statistical Properties of the GEX Incremental Value of the GEX Dealer Positioning and the Q1 2020 Crisis Part 2: A Qualitative Model of Market Maker Impact Pinning Arises from Dealer Hedging Discrete Time Pinning Model Simulated Paths Near Expiration Generalizing the Model Statistical Results: Fat Tail Generation and Potential Whipsaws Non-technical Summary Technical Endnotes 7 The Elephants in the Room: Banks and the “Almighty” Central Bank Part 1: Central Bank Policy and Forward Credit Spreads Raw Size of the Banking System Central Bank Levers The Fed’s Historical Reaction Function Typical Impact of Rate Cuts and QE Multi-factor Regression Results Part 2: The Single Greatest Predictor of Long-Term Equity Returns Institutional Strategic Asset Allocation Templates Bond Supply Varies as a Function of Yields and Issuance Estimating the Quantity of (Bonds + Cash) Over Time Specifying the Equity Supply Indicator Variable Empirical Results A Comparison of Post-GFC Valuations Concluding Thoughts 8 Playing Defense and Attack in the Presence of a Dominant Agent Sizing Positions Sensibly Identifying Pressure Points in the Market Exploiting Products with Fragile Design Features References Index