TAA Investing

大圣电竞赛事下注手机APP下载

TAA strategies look amazing on paper. After all, what is not to like about higher absolute and higher risk-adjusted returns over a buy and hold portfolio? But in the real world when implementing TAA strategies there are a few common barriers that limit and make difficult their implementation for many investors. I call it the “Yeah, but..” or the Pain factor. The three most common barriers to adoption and implementation I hear from investors are taxes, trading, and turnover. In general, TAA strategies generate higher taxes, more trades, and higher turnover than a buy and hold portfolio. In my experience these factors matter a lot more than it would appear on the surface, even in the face of the very large benefits that such strategies can provide. I was wondering if there was a simple way to quantify this Pain Factor and help investors choose amongst the wide variety of TAA strategies? In this post I describe my attempt at quantifying this Pain Factor.

Let’s start with the data. I pulled the relevant data for the TAA strategies tracked on Allocate Smartly . Fortunately they have all the relevant data for the analysis. The three data points I’m interested in for the Pain Factor are the amount of strategy gains that are long term capital gains (vs short term capital gains, interest, and dividends), the number of trades per year, and the percent turnover of the strategy. I pulled the data over the last 20 years along with some other key performance metrics. To calculate the Pain Factor I did a simple rank ordering of long term capital gains (higher percentages are better), trades (lower is better), and turnover (lower is better), and did a sum of the rankings to get an overall figure of merit, which I then rank ordered. The table below shows the resulting rankings.

Note: CG = Capital Gains. Short term and long term CG do not add up to 100%. The rest of the gains are either interest or dividends. Negative numbers for short term capital gains means that they are losses. 

The first thing that jumps out is that buy and hold, i.e the 60/40 benchmark, is not as good as most people think. Yes, there is basically zero work to implement the strategy, outside of re-balancing, but the tax performance of the strategy is not as great as many investors think. This is because 40% of the strategy, being in bonds, is always taxable. The next thing to point out is that while the average TAA strategy (yellow row at the bottom of the table) confirms the general idea that TAA strategies generate significantly higher trades, turnover, and taxes than buy and hold, there is quite a range amongst TAA strategies. For example, 9 of the TAA strategies in the table generate a higher percentage of their gains from long term capital gains than buy and hold – meaning they are more tax efficient. For tax free accounts you would simply ignore the tax columns and focus on trades and turnover.

Let’s look at trades and turnover next. These are related but slightly different. Number of trades is obvious. The higher number of trades makes the strategy more difficult to implement for most investors. Turnover is basically the number of trades times the size of those trades as a percentage of the overall portfolio. The average TAA strategy trades just over twice per month (26.1 times per year) and has a turnover of 250%. However, those figures vary quite a bit. From 0.7 trades per year for Lethargic Asset Allocation and Growth Trend Timing to 96 times a year for Varadi’s Min Correlation Portfolio. There are quite a few strategies where only a “little extra work” gets you most of the benefits of TAA. I put a little extra work in quotes because I think it varies by investor. A while back I did a post on selecting your first TAA strategy where I defined a simple threshold of 12 times per year (1 time per month) as a good starting point for a first TAA strategy. Using that same threshold here gives an investor 15 strategies to choose from.

Now, let’s put it all together and look at the overall rankings. Surprisingly, there is a TAA strategy comes out on top of buy and hold in the Pain Factor, Lethargic Asset Allocation, which as the name implies gives you the benefit of TAA for not too much extra effort. And there are quite a few that are not to too far away from buy and hold in terms of effort but have much better results. For example, to point out a few of the more well known strategies, Traditional Momentum (aka Dual Momentum from Gary Antonacci) comes in at #12 and also, Faber’s Global Asset Allocation 13 (aka GTAA13 which I have often written about on this blog) comes in at #13. GTAA13 has a unique feature where the short term capital gains are actually losses which can be written off against current income thus making the strategy even more tax efficient. Also, my own SPY-COMP strategy comes in at #6 on the list.

That’s about as far as I wanted to go in today’s post. After this analysis, the next obvious step for me is to take into account the Pain Factor and a Performance Metric to calculate a bang for the buck metric, or Performance per unit of Pain. In other words, given the pain of implementing a TAA strategy what do you get out of it? I’ll take a look at that in the next post. For the performance metric I’m going to calculate one figure of merit using Annual Return and one with UPI. Basically, one for investors in wealth building mode and one for investors in portfolio withdrawal mode.

What do you think of the Pain Factor? Would you calculate it differently? Let me know in the comments if you have other ideas.

Tagged ,

15 thoughts on “ The Pain Factor in TAA Strategy Implementation

  1. Great posting, Paul. I’ve tried to restrict the effect of multiple trades by employing such methodologies with my IRA funds. I’ll be fascinated to see your follow-up postings.

  2. How do you measure “trades per year”? Most TAA strategies trade once a month. Is it dependent on the number of instruments traded?

    1. Number of holding does influence the number of trades. The other factor is what signals the model uses to change asset allocation, e.g the use of a shorter momentum period would generate more trades than a longer momentum period. Trades is simply the total number of changes to the asset allocation over the period of time divided by the number of years (in this case 20 years) which gives you the average number of trades per year.

      Paul

      1. – I live in a jurisdiction where there’s no difference between long term/short term gains. There’s some advantage if strategy tends to hold assets for longer than 1 year, but it’s slim
        – I trade several TAA strategies and I never paid any attention to “number of trades”. Why should I care? Fees are not very significant, and doing one more trade is a matter of additional 1 minute. I’m AS subscriber, so a I’m getting target allocations from Allocate Smartly and after 5-10 minutes I’m done.
        – I don’t care too much about turnover, for the same reason.

        I may have different perspective, but personally never considered number of trades / turnaround as a pain points.

        1. Thanks for your perspective. That is an unusual situation in terms of taxes. For many investors, more trades is more work, more touch points with the portfolio, and thus more chances to make behavioral mistakes, i.e. get thrown off the strategy by the market action.

          And of course, as I’ll cover in some following posts, why trade more for the same, or lower results.

          Paul

          1. It think that when number of trades is above some number (20? 30? 40?) it’s no longer feasible to do FIFO analysis by hand. I wrote a script to do this. And then it doesn’t make big difference if it’s 50 or 500 trades.

            I’m trading 7 TAA strategies tranched into 3 tranches. So the number of trade is pretty big.

  3. Thanks, Paul, great post.

    One clarification regarding taxation of GLD in Lethargic Asset Allocation, which has a static allocation of 25% to GLD. GLD is considered a “collectable” for tax purposes, and the long term cap gain rate is 28%. So the LT cap gain rate for LAA is below the 80% listed. But still a very tax efficient strategy.

    I wish you’d included your other strategies in the table, maybe in your upcoming “next step” post?

    Thanks,

    Dave H

    https://www.spdrgoldshares.com/media/GLD/file/SPDR-Gold-Trust-Tax-Information-2019.pdf

    1. Thanks Dave. Good point. Yeah, I’ll have my other strategies ranked there. I’ll release them to newsletter subs first.

      Paul

  4. A corrections regarding my post above, the LT cap gain percentage for LAA will be the same, but the after-tax return will be slightly lower.

    Dave H

  5. Blargh. Looks like there’s no such thing as a free lunch. My strategy has some of the best risk/reward metrics around, but boy does it come with a give on turnover. Feels like this is something that needs to be executed systematically and in a tax-free retirement account.

Comments are closed.