What Works Best?
Several readers have asked, of all the active investing/trading strategies investigated by research summarized in the blog, which ones that individual investors can practically implement work best. With reservations (because of all the uncertainties, statistical biases/criticisms and contradictory evidence to which the research is subject), here are some best guesses:
First, some precepts ranging from obvious to arguable:
- Asset prices repeatedly cross above and below their long-term trends.
- Driven by long-term real economic expansion and inflation, these long-term trends are generally up.
- Departures from trend derive from tendencies of financial markets, which are social systems, to overreact or underreact systematically to new information.
- Confirmation bias helps sustain misreactions.
- Misreactions to good/bad information may be asymmetrical in both degree and speed.
- Misreaction may be path dependent, such that similar new information may lead to different misreactions under different market conditions.
- Variations in prices are highly correlated for similar assets (which form asset classes), but substantially uncorrelated most of the time for different asset classes.
- Individual investors can access asset classes (diversified across components) via exchange-traded fund (ETF) and mutual fund proxies.
Strategic diversification across asset classes harvests uncorrelated volatility via periodic rebalancing to some basic class allocation weights. Periodic rebalancing, independent of any ability to time asset classes, tends to move funds from asset classes that are above their long-term trends to those below their long-term trends.
While expanding the number of asset classes held and increasing the rebalancing frequency have theoretical (frictionless) benefits, there are real-life trade-offs between: (1) number of asset classes held and rebalancing frictions; and, (2) rebalancing frequency and rebalancing frictions.
There are analytical methods for optimizing asset class allocation weights (such as Modern Portfolio Theory). However, empirical results suggest that simple equal weighting is competitive, perhaps because the inputs to analytical methods involve unreliable estimates of asset class returns, volatilities and mutual correlations.
For simplicity and friction avoidance, individuals might consider a strategic allocation involving perhaps five to ten equally weighted asset classes (via low-fee ETFs or mutual funds) and annual rebalancing. This largely passive approach requires little effort after setup.
Recent research relevant to strategic diversification includes:
- “Performance and Risk of Equity Strategy Indexes”
- “Overview of Risk-based Investment Allocations”
- “Volatility-based Equity Market Allocations”
- “Tests of Strategic Allocations Based on Risk Metrics”
- “Mean-Variance Optimizations Versus Equal Weight”
- “Mean-Variance Investing Basics”
- “Fundamentals of Portfolio Weights and Rebalancing”
- “Mean-Variance Optimization Versus Equal Weight”
- “How to Beat Equal Weight Asset Allocation?”
- “Diversifying Across Strategic Allocation Strategies?”
- “Combining Sharpe Ratio and Pairwise Correlation for Diversification”
- “Alternative Portfolio Efficiency Measures”
- “Asset Class Diversification Effectiveness Factors”
- “Adaptive Asset Allocation Policy”
- “University Endowment Performance: Strategic versus Tactical Allocation”
- “Translating Risk Strategies into Common Factors”
- “Harvesting Equity Market Premiums”
- “Asset Allocation Strategy Horse Race”
- “Alternative Global Equity Diversification Approaches”
- “Diversifying with Equity Volatility Exposure?”
- “Overview of Research on Asset Allocation in the Face of Disaster”
- “Liquidity in Asset Selection and Asset Class Allocation”
- “Hedges and Safe Havens Across Asset Classes”
- “An Era of Unstable Risk Premiums?”
- “Optimal Asset Class Allocations”
See the “Strategic Allocation” category for more.
For investors seeking an active edge, there is evidence supporting belief that the price paths followed by assets and asset classes in varying relative to trend are partly predictable, supporting belief in strategies based on intermediate-term momentum and longer-term reversion to value. These strategies may be complementary aspects of the same sloshing of capital from asset class to asset class, and from asset to asset within class. Because of misreaction dynamics and asymmetries, it appears that investors must be more nimble to exploit momentum than reversion to value, and to exploit overvaluation than undervaluation. Many technical trading rules relate to attempts to detect investor misreaction.
Individual investors can implement momentum and/or value allocation strategies for asset classes (again, via low-fee funds, keeping search and trading costs down). These strategies achieve strategic diversification to some degree via the number of (attractive) asset classes held at one time and, further, by varying the asset classes held over time. Relevant research includes:
- “Asset Allocation Combining Momentum, Volatility, Correlation and Crash Protection”
- “Combining SMA Crash Protection and Momentum in Asset Allocation”
- “Alternative Asset Class ETF Momentum Allocations”
- “Melding Momentum, Diversification and Absolute Return”
- “Simple Asset Class ETF Momentum Strategy”
- “Industry/Asset Class Momentum Over the Long Run”
- “A Few Notes on The Ivy Portfolio: How to Invest Like the Top Endowments and Avoid Bear Markets”
- “Combining Value and Momentum Across Asset Classes”
- “The Decision Moose Asset Allocation Framework”
- “Combined Value-Momentum Tactical Asset Class Allocation”
- “Asset Class Momentum Strategy”
- “Asset Allocation Based on Trends Defined by Moving Averages”
In summary, strategic diversification and momentum and value strategies applied at the asset class level via low-fee funds may be among the best approaches for individual investors.
Investors who can trade very efficiently may be able to exploit:
- Cultural practices (when people pay taxes, rules for capital gains taxation, when people make retirement fund contributions, political cycles) that appear to affect asset prices systematically, thereby producing calendar effects.
- The premium paid by those who hedge volatility with equity options and commodity futures. Evidence supports belief that these hedgers are willing to overpay for derivatives to obtain protection against volatility of underlying assets.
To reiterate, the above observations are just guesses regarding the kind of practical investing/trading strategies that work best. Data snooping bias and defects in statistical assumptions are pervasive (see Investing Demons), and these shortcomings generally suborn overstatement of expected returns. Also, “best” is also a function of individual goals and constraints.