Research shows tax management can boost returns from factor investment

Key Points

  • Factor investing typically leads to increased taxes in a portfolio
  • Taxable investors are largely unaware of portfolio tax questions and managers typically strive to maximise pre-tax returns
  • Tax management adds value across all six factor strategies tested over a 10-year horizon

Taxes necessarily erode the alpha of any investment strategy. Tad Jeffrey and Rob Arnott asked the question back in 1993: “Is your alpha big enough to cover its taxes?” Active management and factor investing typically require high turnover and, therefore, increases taxes to the portfolio. A value strategy’s success, for example, depends on buying stocks with low valuations and generating alpha by selling them as they appreciate. For taxable investors, these capital gains events create tax liabilities that reduce their investment strategy’s excess returns.

Financial literature has focused on tax-exempt investing. Studies of after-tax performance must rely on assumptions that can render the results less broadly relevant, as taxation remains a fairly customised feature of investment returns. Taxable investors are largely unaware of portfolio tax questions and managers typically strive to maximise pre-tax returns. Customisation is generally ill-suited to research reports written for a general investment audience and consequently the literature on tax management is pretty thin. 

A new study in the CFA Institute Financial Analysts Journal, by Lisa R Goldberg, Pete Hand, and Taotao Cai, explores the tax efficiency of an indexed investment and six different factor tilts. Essentially, the authors asked whether tax management – the practice of disciplined loss-harvesting within a defined investment objective – within each of these seven common factor investments was worth the effort in alpha.

The authors compare the after-tax active returns of a tax-managed index-tracking portfolio and six factor-tilted portfolios with those of corresponding strategies managed without consideration for taxes (‘tax-indifferent’ strategies). The research stated that four factor-tilted portfolios – value, value momentum, small value, and multi factor – have a targeted forecast beta of one (‘beta-1’ strategies) and two strategies, quality and minimum volatility plus value, have below-market target betas (‘lower-risk’ strategies). The Russell 1000 index is the universe and benchmark for all portfolios except the small-value portfolio, which uses the Russell 3000 index. The benchmarks are taxed as exchange-traded funds (ETFs) that make no capital gains distributions.

“Taxable investors are largely unaware of portfolio tax questions and managers typically strive to maximise pre-tax returns

The tax management implemented in this research goes beyond simple tax awareness with regard to the choice of investments and investment vehicles. Tax management in this context involves the ongoing management of loss harvesting – that is, delaying capital gains in favour of short-term capital losses and benefitting from any tax rate differential between long-term and short-term gains.

To accommodate period dependence, the authors compute returns for 52 rolling quarterly 10-year periods and reported on the full distribution of the resulting after-tax returns. 

Additional tax alpha
Overall, this research shows that value is, indeed, added by tax management across the indexation and all six factor strategies tested over the 10-year horizon. This added value in after-tax active returns ranged from 2.16% per year for the indexation strategy to 1.5% for the quality factor strategy – a substantial gain.

By comparing each tax-managed portfolio with its tax-indifferent equivalent, the authors were able to attribute the after-tax returns into three components – the factor alpha, the tax alpha and the pre-tax residual. The research found that factor alpha is the same for both the tax-managed and tax-indifferent portfolios and so it is the tax alpha and pre-tax residual that indicate the additional benefit to these strategies of tax management.

heidi raubenheimer

               Heidi Raubenheimer

Figure 1 shows all three components of the after-tax returns for these seven tax-managed portfolios. The research finds the highest average tax alpha occurred in the indexed strategy: 2.26% a year and the beta-1 strategies (value, value momentum, small value and multi-factor) have significantly higher tax alphas than the lower-risk strategies. Each of these beta-1 strategies captures between 70% and 80% of the tax alpha of the indexing strategy, amounting to between 1.59% and 1.89% in excess returns a year. The lower-risk strategies have fewer securities and roughly two-thirds of the volatility of the beta-1 strategies, which implies reduced opportunities for loss harvesting. Their tax alphas were only 0.68% and 0.27% for quality and minimum volatility strategies, respectively.

The authors based their results on an investor in the highest tax bracket of a no-state-tax state. But they varied these assumptions to illustrate the customisable nature of these results and the importance of this type of analysis. The full article shows that investors in a lower tax bracket naturally gain less from tax management and those paying both federal and state taxes gain more.

The authors observe that lowering the capital gains tax rates significantly reduces the tax alpha for the index strategy and the beta-1 factor-tilted strategies, while diminishing the quality-tax alpha and erasing the tax alpha for minimum volatility plus value. Introducing a high state-tax rate increases the tax alpha for the index-tracking strategy and the four beta-1 factor-tilted strategies; the tax alphas for the lower-risk strategies are minimally affected.

The authors note that tax alphas for the factor-tilted portfolios are largely the result of managing the tax rate differential and so depend heavily on having an abundance of short-term gains to offset.

The results in figure 1 assume that portfolios are passed on tax-free at the end of the investment horizon – like an estate or donation – so it is essentially the impact of ongoing tax management that is being measured. To evaluate the further consequences of liquidation taxation, the authors calculated the after-liquidation tax alphas, given the embedded capital gains liabilities in the portfolio and the difference in the after-tax return of each portfolio and its benchmark. The median indexation tax alpha drops from 2.16% to 1.30% after liquidation. The beta-1 strategies lose about half a percentage from their tax alphas but the quality and low-volatility tax alphas actually increase slightly. These portfolios have slightly higher tax alphas because they have fewer capital gains – hence, lower tax liabilities – versus their benchmarks.

screen shot 2019 03 29 at 15.41.54

The researchers then flip the question of the benefit of tax management over to interrogate the trade-off between factor investing and tax in terms of a ‘hurdle rate’. They ask how much additional alpha a factor-tilted portfolio must deliver to compensate investors for not investing in an indexed portfolio. 

As expected, figure 2 indicates the hurdle rates are higher in a donation taxation scenario than a liquidation scenario for the beta-1 factor investments. Hurdle rates were generally smaller for the beta-1 factor strategies than the lower risk strategies, which require a bigger factor alpha to compensate taxable investors. 

The findings of the study suggest tax management can have significant benefits when integrated into the portfolio management of indexed and factor strategies for taxable investors. The authors argue that indexed strategies yield the highest tax alpha, whereas lower-risk strategies tend to have the lowest tax alphas, and tax alphas are highly dependent on investor-specific factors. The research shows that high federal and state tax rates increase tax alpha relative to lower tax rates, and liquidation taxes also reduce tax alpha. 

Furthermore, because compounding contributes to alpha over the long term, the authors’ maintain tax alphas tend to be lower for longer horizons and that choice of factor strategy is also important in determining the extent of the benefits of tax management. This research can help investors and their advisers, as it demonstrates how they can calculate and consider the hurdle rates for each strategy alongside its tax indifferent equivalent when choosing an investment strategy.  

Heidi Raubenheimer is managing editor of the CFA Institute Financial Analysts Journal