Does separating the active (alpha) and passive (market exposure, or beta) components of an overall equity investment strategy, thereby isolating turnover, reduce overall tax burden? In their May 2018 paper entitled “The Tax Benefits of Separating Alpha from Beta”, Joseph Liberman, Clemens Sialm, Nathan Sosner and Lixin Wang investigate the tax implications of separating alpha from beta for equity investments. Specifically, they compare two quantitative investment strategies:
- Conventional long-only – overweights (underweights) stocks with favorable (unfavorable) multi-factor exposures within a single portfolio.
- Composite long-short – allocates separately to a passive (index fund) portfolio and to an active long-short portfolio targeting multi-factor exposures but with no exposure to the market.
They design these competing strategies so that aggregate exposures to the market and target factors, and thus pre-tax returns, are similar. They consider three target factors: value (60-month reversion) and momentum (from 12 months ago to one month ago), together and separately; and, short-term (1-month) reversal only separately. Their base simulation model has: 8% average annual market return with 15% volatility; 2% average incremental annual return for each target factor with 4% volatility; and, 180% annual turnover for value, momentum and value-momentum and 1200% annual turnover for short-term reversal. Their test methodology involves 100 iterations of: simulating a multifactor return distribution of 500 stocks; then, simulating portfolios of these stocks with monthly factor rebalancing for 25 years. They assume long-term (short-term) capital gain tax rate 20% (35%) and a highest-in, first-out disposition method for rebalancing. Based on the specified simulations, they find that:
- Gross average annual pre-tax return for the long-only (composite) strategy is 9.5% (9.7%). The composite strategy beats the long-only strategy in 77 of 100 simulations, because the former can apply negative weights to stocks with the weakest factor exposures.
- Turnover of the long-only strategy incurs capital gain taxes for both active and market components of returns, whereas the composite strategy focuses turnover on factor alpha and enables deferral of capital gain on passive market exposure. Specifically:
- Post-tax, the multi-factor composite strategy beats the long-only strategy in all 100 simulations, with gross average annual outperformance 1.3%.
- Moreover, taxes for the composite (long-only) strategy decrease slightly (increase strongly) with strategy returns.
- For the factors separately, the composite strategy outperforms the long-only strategy by average gross annual post-tax 1.4% for value, 1.1% for momentum and 2.3% for short-term reversal.
- The composite strategy also simplifies changes in investment managers by investors.
- Findings are generally robust to alternative factor models, costs of trading and shorting and liquidation taxes.
In summary, evidence from simulations indicates that strategies that separate alpha from market beta (isolating turnover to factor exploitation) have material tax advantages over conventional long-only factor tilt strategies.
Cautions regarding findings include:
- The portfolio strategies outlined above are beyond the reach of most investors, who would bear fees for accessing them via funds. Fees may vary by strategy type.
- As noted in the paper, the stock market modeling/simulation approach used in the study makes simplifying assumptions (in addition to those cited above), including: the market factor is uncorrelated with the value, momentum, and short-term reversal factors; the correlation between value and momentum factors is -0.70; all stocks have constant market beta of 1.0; and, annualized stock-level volatility is a uniform 25%. Using real market data may affect findings.
