A quick post to present the performance results of the TAA strategies that are part of the Economic Pulse and Quant Pulse Newsletters. I’ll also discuss how the performance compares to other TAA strategies. Let’s dive in.
First, let’s look at the TAA strategies that are part of the Economic Pulse Newsletter. The table below shows various performance statistics for the strategies and how they rank versus the universe of strategies tracked by AllocateSmartly. For returns, I show performance for 2020, 20 year, and 10 year performance. The rest of the figures, such as vol, UPI, trades, are for the full 20 year period.
As for the rankings, I ranked my strategies versus all of the strategies tracked on AllocateSmartly over the periods of 2020, 10yr, and 20yr returns. I then combined the 2020, 10yr, and 20yr return rankings to get an overall ranking. As an example, the VOL-COMP Global TAA strategy ranked 3rd in 2020, 1st over the last 10yrs, and 1st over the last 20yrs among 50 or so strategies ranked, including all of mine. The strategies highlighted in green are my strategies that are available on AllocateSmartly.
All in all, the Economic Pulse strategies did very well in 2020 and over the long term are ranked among the top TAA strategies. Also, here is a list of the strategies that are ranked above the 60/40 benchmark across the three periods I considered; 2020, 10yr, and 20yr performance. You can find the detailed stats on these other TAA strategies on AllocateSmartly.
Of course, it’s not all about returns so below are the rankings using UPI (Ulcer Performance Index), a better measure of risk-adjusted performance than the Sharpe or Sortino ratio. Also, on UPI the Economic Pulse TAA strategies do quite well all while keeping the number fo trades relatively low.
Now, let’s look at the performance of the more aggressive TAA strategies that are part of the QuantPulse membership. These strategies are newer than the ones on Economic Pulse and use a daily model based on the VIX futures curve to make risk-on and risk-off decisions. Because of the more aggressive and frequent risk management the models can get more aggressive by allocating to leveraged ETFs. Basically, the aggressive versions of these strategies are seeking to maximize return while keeping risk and drawdowns similar to or less than SPY. Alternatively, using the non leveraged ETFs yield better than 100% stock returns while keeping risk quite low.
For more information on the Vol Curve model take a look at these introduction posts, here and here. The table below shows the performance of these models over various time periods and the other relevant portfolio stats. The limited history of these strategies, back to 2008, is due to the limited history of the certain VIX futures products that are used in the models.
OK, that about does it for this quick roundup of model performance for 2020. In the next week or so I’ll follow this up with an update on quant portfolio performance for 2020 and then I’ll post my annular long term portfolio performance comparison like I do every year.
9 Comments
Marcin Jagodzinski · January 12, 2021 at 5:45 am
In 2020 the day of month selected for rebalancing was very important. For most strategies trading on the last day of month was the best choice. Choosing different day could easily wipe half or more profits. Anyway those results are awesome!
paul.novell@gmail.com · January 12, 2021 at 10:42 am
Very true. But, also important to note, that over the long term, the “End of Month Effect” as it is called, has been show to be significant across markets and time periods.
Paul
Marcin Jagodzinski · January 12, 2021 at 11:01 am
I’m running a portfolio using 7 TAA strategies, divided almost equally into 3 tranches, rebalanced on 7th, 14th, 21st trading day. And the difference in 2020 between this portfolio and and portfolio rebalanced on 21st day is vast: it underperforms by 60%. So you can imagine how underperforming the 7th/14th tranches have to be to lower the overall performance to such degree. I’ll check the exact numbers later.
I’m not sure if this is “End of Month Effect”. There were some critical moments and if strategy failed to rebalance in some shore time window, it suffered. The most critical was last day of February, bottom of first wave of Covid correction. Most strategies rebalanced to more defensive assets on that day. Compare to strategy rebalancing on 14th trading day which would probably stay risk on for 14 more days, during which SPY lost 20% or more…
paul.novell@gmail.com · January 13, 2021 at 12:46 am
Yes, very true. 2020 really bifurcated strategy performance due to the speed of the move, both down and up. One criteria I use to judge a TAA strategy is its consistency in returns across days of the month, as measured by (Max Return – Min Return)/ Avg Return. It is kind o fa measurement of timing luck. I think it is an important part of strategy design. Some strategies have very little variance in returns across days of the month, < 10%. While others, often those with very fast signals, have much large variance in returns across the days of the month, greater than 50% in some cases. Paul
Brad · January 13, 2021 at 5:50 am
Exactly. Corey Hoffstein calls this “timing luck” and has several papers he wrote showing how important this is. I did most of my rebalancing in mid March so got the bad end of timing luck this time around. To minimize that I am splitting my portfolio into four tranches that run through the rules on the 1st, 8th, 16th, and 24th. Thank goodness for low commission brokers 🙂
Brad · January 12, 2021 at 5:55 am
Hi Paul. It has been awhile since I have read your blog so I have a lot of catching up to do 🙂
If I remember correctly some of your TAA strategies used the unemployment rate (UER) being above a moving average to move all or art of a strategy to cash. It looks like the UER has been above most of the moving averages since March 2020 despite the market moving up a lot since then.
Do you still use this indicator and if so was that a drag on performance for this year? Would you make any tweaks to the system based on this or was 2020 just a fluke year when unemployment was high but the stock market rallied? Thanks!
paul.novell@gmail.com · January 12, 2021 at 10:48 am
Hey Brad, the unemployment rate is part of a set of indicators I use in SPY-COMP. But the indicators are only half the story. My model uses the status of the economic indicators plus the market trend to make investment decisions. So, even when an indicator is flashing warning signs, the price trend of the SP500 has to confirm that. In 2020, the market trend turned positive by the end of May 2020, and it’s been risk-on ever since.
Paul
Kevin M · January 12, 2021 at 8:41 am
Hi Paul, thank you for continuing to provide this information to us followers. It is helpful and highly informative. I’m having mixed feelings about 2021 and get the sense domestic and global change and uncertainty will have a drastic negative impact during the year and perhaps beyond. Have you formulated an opinion yet? I’m so apprehensive I almost believe an annuity for the short term might yield the best return in what I think will be a disastrous financial year. Thank you as always.
paul.novell@gmail.com · January 12, 2021 at 10:55 am
Hi Kevin, this is one of the many reasons I use 100% rules based models to make my investment decisions. My overarching goal
in my investment process is to NOT let my opinions influence my investment decisions. The two don’t mix well together. So, I work
really hard to pick the right model that meet my investment goals and then I let the models do the work.
Paul
Comments are closed.