Tax Alpha is the increase in portfolio performance relative to its benchmark after taxes have been considered. It is the difference between the after-tax active return and the pre-tax active return.

The tax alpha formula requires the after-tax benchmark return. Since taxes depend on the size and timing of purchases and sales and the investor's tax brackets and external gains, after-tax returns are different for every investor. A Shadow Benchmark replicates the cash flows into and out of the portfolio, but applies them to the benchmark instead. The same tax brackets and external gains are then used to calculate taxes on the benchmark.

A Direct Index is a portfolio that holds all of the constituents of an index such as the S&P 500 at their index weights. It is an alternative to holding an index ETF or mutual fund. Ironically, the motivation for a direct index is to deviate from the index in order to express personal preferences such as ESG and/or to increase after-tax returns through tax loss harvesting and other methodologies.

The Direct Index exactly replicates the pre-tax returns of the index. It requires that changes to the index constituents and weights be followed with perfect precision and timing.

The Practical Direct Index tracks the index as closely as possible with publicly available information and monthly rebalancing. The S&P 500 constituents are publicly available, but their exact weights are not. Market capitalizations are used to closely approximate the actual weights. Constituent changes often occur mid-month, but are not applied until month-end. Similarly, dividends are not reinvested until month-end. These differences introduce a modest amount of tracking error versus the actual index.

The Direct Index with Tax Loss Harvesting (TLH) strategy is a direct index portfolio that implements tax loss harvesting in order to realize capital losses and defer capital gains.

In addition to tax loss harvesting, the Direct Index with TLH and Yield Reduction strategy seeks to reduce dividends in favor of capital gains. Dividends are paid and taxed immediately, whereas capital gains may be deferred indefinitely. This strategy creates additional tax alpha but also increases tracking error.

The distinction between the three types of indices is in how they are taxed. Index ETFs do not realize capital gains and losses. All gains and losses are unrealized. Mutual Funds must distribute all realized gains to shareholders each year but may not distribute realized losses. Realized losses must be carried forward by the fund to the following tax year. With a Direct Index, any realized losses may be used to offset realized gains from other sources. If there are no gains to offset, up to $3,000 of realized losses may be dedcuted against ordinary income each year with any excess carried forward to the following year.

The External Gains assumption is a major and controversial factor in the calculation of tax alpha.

Without any external gains, deductible capital losses are limited to $3,000 per tax year. With the assumption that unlimited long-term gains are available for offset, all realized losses are used to avoid taxes at the preferential long-term gains rate. The assumption of unlimited short-term gains leads to the largest estimates of tax alpha. With this assumption, all realized losses are used to avoid taxes at the higher short-term gains rate.

Without external gains, tax loss harvesting can only exploit the $3,000 loss allowance. The significance of the loss allowance to tax alpha decreases as portfolio size increases.

The After-Liquidation performance assumes that the portfolio is liquidated at the end of the simulation. All unrealized gains become realized and are taxed at the Liquidation Capital Gains Tax Rate.

A liquidation capital gains tax rate that is lower than the long-term capital gains tax rate may be used to reflect the discounting of taxes on a liquidation that occurs beyond the end date and/or the possibility that the basis of the portfolio will be "stepped-up" at death.

A 130/30 strategy is a portfolio that is 130% long and 30% short. The net market exposure is still 100%, just like a long-only strategy. 130/30 strategies increase the opportunities for tax loss harvesting, especially in multi-year rising markets.

The Passive 130/30 strategy does not use stock-specific return forecasts, or alphas. The Smart Beta 130/30 strategy uses a simple "smart beta" (size, value, momentum, and quality) return forecasting or "alpha" model for stock selection. The AlphaWorks 130/30 strategy uses our proprietary machine learning driven alpha model.

The standard formula for Tax Alpha penalizes an active strategy for the additional taxes resulting from its pre-tax excess return. Active Tax Alpha corrects for this shortcoming by adding (Pre-Tax Active Return x Benchmark Effective Tax Rate) to the standard Tax Alpha.

Please request our Active Tax Alpha whitepaper for additional details.