Today we discuss a new feature on the models page designed to help you understand your models overall tax/trade efficiency over time.
Momentum strategies aren’t afraid to trade. Those familiar with momentum models are no stranger to turnover, but that isn’t to say your clients understand why (or how) they see a perceived excess of trades in their accounts. Some may like to see the activity, while others may question why you feel the need to constantly rotate positions, especially if they don’t completely understand the idea of relative strength investing. Regardless of your client’s position on trading, there is one concept all of us do understand: taxes. Around the office, we like to say there are only three things certain in life: death, taxes, and the daily report being published by the end of the day. Of course, we all want to avoid paying Uncle Sam, but clients often think that more trades always equal a higher tax bill. But is that true?
Today’s feature will unveil a tool the analyst team designed to help answer that question. The idea of “tax/trade efficiency” is simple: how good is my model at realizing long-term gains, which receive more favorable tax treatment? Conversationally, we frequently discuss momentum’s tendencies to cut losers off quickly and let winners run. This, in theory, would create favorable tax treatment (this trading pattern should generate short-term losses and long-term gains) but until now there hasn’t been a way to “prove it”. The trade efficiency tool allows us to quantify this idea. Before going further, we want to stop and provide a general disclaimer: The provided score should not be construed as tax advice, but rather as a general guide as to a model’s tendencies when it comes to overall trading patterns over time.
Released on 10/30, a model’s popout will have another tab available, detailing a model’s trade efficiency. The biggest point to note is the model’s “trade efficiency score” (highlighted below) which should be interpreted as a percentage. In short, the score represents the total proportion of gains that were classified as long-term compared to the total number of gains realized over the course of the backtest. Higher score = models with more “efficient” trading tendencies. The yearly breakdown going into the score is under the “cumulative trading efficiency” tab, breaking down each calendar year into more detail. Years that realized losses (or were still carrying forward losses from previous years) did not generate a score. You can adjust the lookback period by adjusting the start/end date (highlighted below). This will adjust the total score and will help create a more personalized view. A practical use case would be to adjust the timeframe to when you started following the model, or focus on the last 10 years, etc. Remember, realized gains typically will not align perfectly with a models performance over the year… since this tab is only focused on realized gains, not inclusive of unrealized gains that impact overall model performance but have no impact on the taxes your client would hypothetically pay.
We also included a hypothetical tax liability tab (date range fully adjustable) that allows you to get a peak behind the curtain of what a tax bill might have looked like following said model based on unique long & short term tax rates, carryforward allowances, and starting investments. Again, don’t use this as explicit advice, but more so of a guideline for what you might have experienced. We also included the effective tax rate as an additional data point.
For those of you less interested in the year by year analysis or hypothetical tax liabilities over time, the cumulative trade efficiency table gives some high level details of overall trade patterns, including the total percentage of trades in each tax bucket. As you would expect, the majority of trades for our example (Large Cap Core) are short term losses as the model rotates away from weakness, followed next by the long-term gain bucket, where most of the historical gains come from (+120% on average). We also include a historical trade distribution that allows you to X-ray into any trade over time and see its unique holding period and % gain or loss.
A few closing points. This score is best used as a secondary indicator. Tax inefficient models can still perform quite well, but are perhaps best suited in qualified accounts to reduce tax drag. A general rule of thumb, models that give positions more room to run will typically be more efficient (pre-made matrix models generally hold a higher score than FSM models due to more lenient sell thresholds). Use the score as another tool in the toolbox. Many of you will have clients that have asked (or complained) about the number of trades in their account. Use this tax score to help back up the claims that momentum strategies are more efficient than they may appear.
Click here for the whitepaper. For further questions, email miles.clark@nasdaq.com.
