Objective research to aid investing decisions

Value Investing Strategy (Strategy Overview)

Allocations for October 2024 (Final)
Cash TLT LQD SPY

Momentum Investing Strategy (Strategy Overview)

Allocations for October 2024 (Final)
1st ETF 2nd ETF 3rd ETF

Overview of Research on Asset Allocation in the Face of Disaster

| | Posted in: Strategic Allocation

Has the academic community made practical progress in specifying asset allocation approaches that mitigate adverse impacts of multi-market crises (systemic risk) on diversified portfolios? Two recent papers address this question in complementary top-down and bottom-up ways. The February 2011 version of “Asset Allocation and Asset Pricing in the Face of Systemic Risk: A Literature Overview and Assessment” by Christoph Meinerding assesses recent research linking systemic risk with asset pricing and asset allocation, with systemic risk essentially characterized by the empirical properties of contagion. The 2011 paper “Fat-Tailed Models for Risk Estimation” by Stoyan Stoyanov, Svetlozar Rachev, Boryana Racheva-Iotova and Frank Fabozzi reviews mathematical approaches for modeling return distributions that match empirical data. Based on the relevant bodies of research, these papers conclude that:

“Asset Allocation and Asset Pricing in the Face of Systemic Risk: A Literature Overview and Assessment” takes a top-down view of how economic or market shocks become crises, with conclusions as follows:

  • The key empirical properties of contagion (systemic risk) are:
    • An initial shock such as a drop in global economic growth, a change in commodity prices, or a change in interest or currency exchange rates can boost comovement of capital flows and asset prices across markets via trade linkages, financial entanglements and/or investor irrationality.
    • Contagion elevates the risk of adverse events in financial markets for a certain time period, temporarily changing the risk-return properties of asset classes.
    • Contagion increases uncertainty as well as quantifiable risk, manifested as exceptionally large dispersion in beliefs about the magnitude of a crisis and corresponding rationales for response.
    • Contagion may be essentially coincidence of extreme (gross return distribution tail) events, such that structural breaks in market correlations are inherently resistant to unbiased tests.
  • While it appears likely that the effects of contagions (systemic risk) partially or completely offset the gains from asset class diversification, the practical implications of efforts to understand the link between systemic risk and asset allocation are scarce.

“Fat-Tailed Models for Risk Estimation” takes a bottom-up view of the best way to model the empirical return distributions of financial markets, with conclusions as follows:

  • Because theory is lacking, the problem of modeling financial gross return distributions is largely statistical, with the data indicating to some degree the following characteristics:
    • Volatility clusters, with large (small) returns tending to follow large (small) returns.
    • Returns are autoregressive, with positive (negative) returns tending to follow positive (negative) returns.  In other words, returns exhibit momentum.
    • Return distributions are skewed, meaning that the upside and downside tails are different.
    • Return distributions are fat-tailed, with the probabilities of extreme profits and losses much larger than predicted by a normal distribution. Tail fatness tends to:
      • Vary from asset to asset.
      • Change over time, thinner (fatter) in “regular” (turbulent) market conditions.
      • Vary with measurement frequency, thinner (fatter) for low-frequency (high-frequency) sampling.
  • Using the Gaussian (normal) distribution or the Student t distribution (fatter tail, but fixed fatness and unskewed) on the grounds of simplicity or tractability is potentially misleading.
  • “The best approach is to choose an extended model that includes the normal distribution as a special case and then test its performance on the real data.”

In summary, evidence from surveys of relevant research indicates that academia has made little progress in finding practical ways for investors to protect even diversified portfolios from extreme events (crashes).

One straightforward, practical caution is that relying on just the mean return and standard deviation of returns (or combinations of the two) to assess high-frequency trading outcomes is probably risky.

See “The Black Swan: The Impact of the Highly Improbable (Chapter-by-Chapter Review)”, “The Fourth Quadrant: No Realm for the Normal” and “Surviving by Staying Out of the Fourth Quadrant” for additional background and discussion.

Login
Daily Email Updates
Filter Research
  • Research Categories (select one or more)