A Turbo-Charged System (at Exactly the Wrong Time?)

  • SumoMe

The “system” I will write about here involves only a slight modification to one written about here by Dane Van Domelen on 10/23/2017. Any”credit” for anything that is good about this system should be given to Mr. Van Domelen.  The interesting thing – besides the surprisingly profitable, albeit volatile, long-term results – is the fact that I hated the idea when I first read Mr. Van Domelen’s article.

The Original Idea

In the linked article, Mr. Van Domelen’s proposed buying and holding tickers UPRO and TLT.  UPRO is an ETF that tracks the daily change for the S&P 500 Index times 3.  Ticker TLT tracks the long-term treasury bond.  I am not a huge fan of buying 3x funds and holding them for any length of time as my own personal feeling is that – “at some point they get killed.”  Also, while long-term treasuries have made investors a lot of money over the past 30+ years, some day interest rates will embark on a new uptrend – and long-term treasuries will likely get hit hard, or so it seems to me.

Mr. Van Domelen also tested his theory “starting in 2009.”  I must admit that I rolled my eyes when I read that.  “Of course it’s done well since 2009” I thought, “long-term bonds are up since that time and the S&P 500 has more than tripled in price”.  Still, as a “student of the market” I decided to test the idea out “a little further back.”

The results were very interesting.

The Original Test

For testing purposes I used monthly index total return data from the PEP Database from Callan Associates.  Using index data I was able to test back to a starting date of 12/31/1972.  For testing purposes I used:

*Monthly Total Return data for the S&P 500 Index (multiplied x 3)

*Monthly Total Return data for the Barclay’s Bloomberg Treasury Long-Term Index

*The portfolio rebalances at the start of each year to start the year holding:

1/3rd in SPX times 3

2/3rds in Long-Term Treasuries

*I also deducted 1% per year in “fees” (deducted pro rata on a quarterly basis)

Using this approach generated the results in Figure 1:

Measure System SPX
$1,000 on 12/31/1972 grow to==> $930,724 $81,330
Average Annual % +(-) +19.9% +13.0%
Median Annual % +(-) +19.7% +14.6%
Standard Deviation of Annual Returns 21.4% 16.9%
Ave/StdDev 0.93 0.77
Worst 12 months % +(-) (-32.2%) (-43.3%)
Maximum % Drawdown (-42.2%) (-50.9%)

Figure 1 – Original “SPX x 3 plus Long-Term Treasury” system results; 12/31/1973-9/30/2017

Classic “high risk, high return” results.  Almost 20% annually but a large standard deviation and very large drawdowns.

Other Tests

In testing other data I determined that a portfolio of:

*1/3rd SPX times 3

*2/3rds 7-10-Year Treasury times 2

Performed even better.  At least in the eye of this beholder. In addition, I personally prefer intermediate-term treasuries generally speaking as they tend to be less volatile than long-term treasuries.

There is an ETF (ticker UST) that tracks the 7-10 Year treasury times 2.

So in a nutshell, in this test:

*Each year on December 31st the portfolio resets to:

1/3 SPX (x3)

2/3 7-10 year treasuries (x2)

*Each month the SPX portion of the portfolio gains or loses the monthly total return for SPX times 3.

*Each month the 7-10 treasury portion of the portfolio gains or loses the monthly total return for the Bloomberg Barclays Treasuries – Intermediate Index times 2.

*1% in fees is deducted annually (with fees deducted quarterly)

Using this hypothetical approach, $1,000 invested in 1972 would be worth $3,134,089 through September 2017 versus $81,330 invested in the S&P 500 Index.  Figure 2 displays a chart showing the growth of equity and Figure 3 displays the particulars.2Figure 2 – Growth of $1,000 using “System using 7-10 yr. Treasuries (x2)” (blue line) versus SPX (red line); 12/31/1972-9/30/2017

Measure System SPX
$1,000 on 12/31/1972 grow to==> $3,134,089 $81,330
Average Annual % +(-) +23.1% +13.0%
Median Annual % +(-) +21.3% +14.6%
Standard Deviation of Annual Returns 21.4% 16.9%
Ave/StdDev 1.08 0.77
Worst 12 months % +(-) (-27.8%) (-43.3%)
Maximum % Drawdown (-39.9%) (-50.9%)

Figure 3 – “Updated System” versus SPX;12/31/1972 to 9/30/2017

Before anyone gets too excited based on the outsized hypothetical returns, there are a number of “reality checks” to consider:

*These results are “hypothetical” based on index returns and NOT real-time results from actual trading

*These results involve “extremely volatility”.  The annual standard deviation of returns is 21.4% and the maximum drawdown is -39.9%.  On 3 separate occasions – 1973-74, 1987 and 2008 – drawdown exceeded -28%.  Not everyone is built to endure these types of declines and stay with it.

*The market has had an exceptional run up of late and some day interest rates will rise in earnest – so now may be the exact wrong moment to even be looking at something this aggressive.

A Trading Portfolio

I am in no way, shape or form recommending that anyone trade this portfolio.  As always it is – in the immortal words of Rod Serling – “submitted for your approval.”

It is my usual method though to try to shape ideas into actual trading methods – even if just for further analysis. So for trading purposes:

*1/3 in ticker UPRO (SPX x 3)

*2/3 in ticker UST (intermediate-term treasury x 2)

*Rebalanced annually

*1% fee deducted per year (0.25% deducted quarterly)

The first full month that both of these ETFs generated real-time returns was February of 2010. The results generated since that time – versus the S&P 500 Index – appear in Figure 4.

4Figure 4 – Growth of $1,000 using total return data for UPRO and UST (blue line) versus SPX (red line); 1/31/2010-9/30/2017

Summary

Bottom line: The hypothetical results are excellent – and very compelling.  At the same time, the real world risks are huge.

This is absolutely “high risk/high reward” territory.  The danger is that an investor inured to all market risk by the recent non-stop advances by the major stock averages will see something like this as a way to “shoot for the stars.”  But the downside risk here is huge!  Consider that in 1987 this “system” lost -39.9% in just three months time (also 1973-74 witnessed a -29% decline and 2008 a -30% decline).  The reality is that most investors will not “stay the course” in the face of this type of volatility.

The fly in the ointment, the monkey in the wrench, the pain in the, uh, well you get the idea.  Just remember that if “past is prologue” this system could lose -39.9% in the next 3 months (ala the 1987 decline).

Who’s ready to ride?

Jay Kaeppel

Disclaimer:  The data presented herein were obtained from various third-party sources.  While I believe the data to be reliable, no representation is made as to, and no responsibility, warranty or liability is accepted for the accuracy or completeness of such information.  The information, opinions and ideas expressed herein are for informational and educational purposes only and do not constitute and should not be construed as investment advice, an advertisement or offering of investment advisory services, or an offer to sell or a solicitation to buy any security.

3 thoughts on “A Turbo-Charged System (at Exactly the Wrong Time?)

  1. Could you send me the yearly performance since 1972? I’m more interested in having to explain down years when SPX up.. I could handle elevated volatility for that much better return.. Also, did you include dividends and interest in comparing performance?

    1. Ryan, Unfortunately,no I can’t. The problem is that I haven’t quite figured out how to replicate XIV returns. Prior to the time that XIV started trading real-time I used index data for the index that XIV is designed to track. However, that data only went back to January 2008. So kind of dead in the water at the moment in terms of trying to go back farther. Truth be told this approach scares the heck out of me – but the big returns are compelling. Hopefully not like a moth to the flame…..Har. Jay

  2. Jay,

    Now switch your hedge to 3x leverage, so TYD, and buy 40% UPRO only (at EOM) when its above its 10 month SMA. If it is you’re 40% UPRO + 60% TYD, if not 100% TYD.

    You may want to try 50/50 to reduce the influence of the bond bull a bit.

Comments are closed.