Please Note: Blog posts are not selected, edited or screened by Seeking Alpha editors.

Buy And Hold S&P 500 Bull 3x ETFs (E.g. UPRO): High Risk, Massive Returns

|About: ProShares UltraPro S&P 500 ETF (UPRO)

Summary

3x daily S&P 500 leveraged ETFs are generally viewed as high-risk investments best suited for intraday speculative trading by professionals.

Since its inception in 2009, UPRO has experienced 49.8% annualized growth and has tracked 3x S&P 500 daily gains extremely well (correlation = 0.996).

Over the past 65 years, a zero-tracking error 3x S&P 500 ETF would have averaged 29.4% annual gains (range -85.5% to 191.7%) and 1.69% monthly gains (range -68.0% to 54.0%).

Buy-and-hold positions in UPRO should drastically outperform the S&P 500 in years when the index gains 5% or more, but risk drawdowns upwards of 95% and complete loss of investment.

Basic Background on Leveraged ETFs and UPRO

Investopedia.com defines a leveraged ETF as "an exchange-traded fund that uses financial derivatives and debt to amplify the returns of an underlying index." ProShares UltraPro S&P 500 ETF (NYSEARCA:UPRO) is a bullish leveraged ETF that aims to multiply the daily return of the S&P 500 by a factor of 3. UPRO's 3x multiple works on daily gains, so investors should not expect to realize 3x the S&P 500's growth over any extended period of time.

Warnings and Technical Issues

The focus of this article is on the merits of long-term investment in a 3x S&P 500 ETF. In my experience, the mere suggestion of holding a leveraged ETF for longer than part of one day is often met with outrage from people familiar with the funds. The consensus among stock market analysts seems to be that leveraged ETFs are best suited for intraday speculative trading by professionals (e.g. Bob Pisani at CNBC, Joe Light at The Wall Street Journal, Justin Kuepper at Yahoo! Finance). There are some good reasons to exercise caution with leveraged ETFs:

  1. Leveraged ETFs are designed to achieve a daily multiple of the underlying index. All bets are off on what the multiple might be over a week or a month.
  2. They lose value in periods when the index fluctuates but has no net change. This is due to the somewhat unintuitive behavior of amplified percentages. For an illustrative example, see this SEC alert.
  3. They can experience massive drawdowns when the underlying index has significant losses. For example, S&P 500 declined 19.4% from April 29 to Oct. 3, 2011; UPRO declined 51.7%.
  4. If the S&P 500 ever reaches an intraday loss of 33%, you can lose your entire investment in UPRO. (Since 1950, S&P 500's biggest loss from open to low was 20.5%, on Black Monday, Oct. 19, 1987).
  5. There is no guarantee that UPRO's daily gains will be exactly 3x the S&P 500's ("tracking error").

There are even more technical issues associated with leveraged ETFs and UPRO in particular (fully described in its prospectus). Basically, there are numerous reasons why UPRO might suddenly exhibit erratic behavior and deviate from its 3x goal. Potential investors should be aware of these issues. In this article, I operate under the assumption that UPRO will continue to track the S&P 500 as well as it has during its first 5.7 years. Finally, potential investors should note that UPRO has an expense ratio of 0.95%.

Outline of Article

Now that we've got background and technical issues out of the way, here's a basic outline of the meat of the article. In the first section, I look at the performance of UPRO since its inception in 2009. I conclude that UPRO has done an excellent job tracking 3x daily S&P 500 gains and has had tremendous growth during this bull market, but has also suffered severe drawdowns.

In light of UPRO's excellent tracking, I focus the second section on historical gains for a hypothetical "zero tracking error" 3x S&P 500 ETF, going all the way back to 1950. My conclusion: incredible growth (16.1% annualized), nauseating drawdowns (up to 97.7%). Histograms show that the hypothetical 3x S&P 500 ETF actually does average roughly 3x greater monthly and annual gains than the S&P 500.

Finally I summarize my perspective on leveraged ETFs and how they might fit into an investor's portfolio.

UPRO's Performance in its First 5.7 Years

A natural concern with UPRO is whether it accomplishes its stated goal of multiplying daily S&P 500 gains by 3. UPRO has historically had minimal tracking error (below). Interestingly, the mean daily gain for UPRO was actually 3.26 times greater than the mean daily gain for S&P 500 (0.206% vs. 0.063%).

Figure 2: Daily gains, UPRO vs. 3x daily S&P 500, over UPRO

Now let's look at UPRO's growth. It turns out an investor would have done tremendously well buying UPRO the day it debuted in 2009 and holding it ever since. Consider its growth compared to the S&P 500 over UPRO's lifetime (Table 1). Annual growth of 49.8% turns $10,000 into $1 million in approximately 11.5 years. On the other hand, UPRO was prone to drawdowns of over 50%, while the S&P 500 never lost more than 20% of its value.

Table 1. Performance of UPRO and S&P 500 from June 25, 2009, to Feb. 23, 2015
  Total Growth Annualized Max. Drawdown
UPRO 890.9% 49.8% 51.7%
S&P 500 129.2% 15.7% 19.4%

Hypothetical growth of $10,000 in UPRO vs. S&P 500 since UPRO's inception is shown below. After a couple of big drawdowns, UPRO took off and reached $99.2k compared to S&P 500's $22.9k.

 Figure 1. UPRO vs. S&P 500, June 25, 2009, to Feb. 23, 2015.

Historical 10-Year, Annual, and Monthly Gains, 3x S&P 500 ETF vs. S&P 500

Huge gains during a 6-year bull market do not automatically render UPRO or other leveraged ETFs a good investment going forward. We would like to consider historical performance going back further than 2009 to get an idea of how UPRO would perform in various market scenarios. While UPRO has only been around since 2009, we can easily go back and see how a hypothetical 3x S&P 500 ETF with no tracking error would have performed throughout the history of the S&P 500. This analysis is justified by the excellent tracking of UPRO to 3x S&P 500.

Table 2 shows total performance of the 3x S&P 500 ETF vs. S&P 500 from Jan. 3, 1950, to Feb. 23, 2015. The 3x S&P 500 ETF had an incredible 16.1% annualized growth over this 65-year period. It was also incredibly risky, at times losing up to 97.7% of its value.

Table 2. Performance of 3x S&P 500 ETF and S&P 500 from Jan. 3, 1951, to Feb. 23, 2015
  Growth of $10k Annualized Growth Max. Drawdown
3x S&P 500 ETF $169.55M 16.1% 97.7%
S&P 500 $1.29M 7.7% 56.8%

While we're at it, let's look at hypothetical growth of $10k over the full 65-year period (below). The 3x S&P 500 ETF's final value of $169.55M was actually less than half of its maximum, $408.30M, achieved in March of 2000.

Figure 3: Hypothetical growth of $10k in S&P 500 (top) and in 3x S&P 500 ETF (bottom).

Then there's the drawdown issue. From 2000 to 2002, the 3x S&P 500 ETF dropped from an all-time high of $408.30M to $35.91M, a 91.2% drawdown. It increased to $223.81M only to lose 95.7% of that value, falling to $9.56M by early 2009. While these drawdowns are incredibly severe, the post-drawdown value of $9.56M is still over 7.5 times higher than the S&P 500's final value of $1.26M.

The performance of the 3x S&P 500 ETF from 2000 to 2007 is concerning. By Oct. 2007, the S&P 500 had fully recovered from its 50.9% drawdown from March 2000 to Oct. 2002, while the 3x S&P 500 ETF had only reached 54.6% of its March 2000 peak. This is a striking example of how a leveraged ETF can deteriorate in a period when the underlying index experiences no net change.

Table 3 shows how an investment in a 3x S&P 500 ETF would have fared against the S&P 500 during each decade or partial decade since 1950. The 3x leveraged ETF outperformed the S&P 500 in 5 out of 7 decades. In the 1970s, S&P 500 gained about 16% while the 3x ETF lost 10%. Mean final balance across the seven decades was $23.3k for the S&P 500 and $128.3k for the 3x ETF. The worst decade for the 3x ETF was the 2000s, when it lost 90% of its initial $10k. Maximum drawdown was higher for the 3x ETF every decade and ranged from 52.6% to 97.7%.

Table 3: Performance of hypothetical zero-tracking error 3x S&P 500 ETF and S&P 500 for each decade since the 1950s.
Decade Security Growth of $10k Annualized Growth Max. Drawdown
1950s S&P 500 $35.9k 13.6% 21.5%
  3x ETF $311.8k 41.1% 54.1%
1960s S&P 500 $15.4k 4.4% 28.0%
  3x 500 ETF $26.6k 10.4% 64.6%
1970s S&P 500 $11.6k 1.5% 48.2%
  3x ETF $9.0k -1.1% 88.3%
1980s S&P 500 $33.4k 12.7% 33.5%
  3x ETF $133.6k 29.4% 80.0%
1990s S&P 500 $40.8k 15.0% 19.9%
  3x ETF $372.9k 43.3% 50.8%
2000s S&P 500 $7.7k -2.6% 56.8%
  3x ETF $1.0k -20.5% 97.7%
2010s S&P 500 $18.6k 12.8% 19.4%
  3x ETF $43.4k 33.0% 52.6%
Averages S&P 500 $24.1k1 8.2% 34.7%1
  3x ETF $142.5k1 19.4% 72.6%1

1 Calculation excludes 2010s since this was a partial decade.

Finally, let's look at the distribution of annual and monthly gains (below) to get an idea of historical gains over more realistic holding periods. The 3x S&P 500 ETF averages much greater annual and monthly gains than the S&P 500, but also has much greater spread. Some relevant statistics are also provided in Table 4. Notice that the 3x S&P 500 ETF averaged 3.34 times greater annual gains and 2.95 times greater monthly gains than the S&P 500.

Figure 4: Annual gains from 1950 to 2014 for S&P 500 and 3x S&P 500. Vertical lines indicate means.

Summary of annual and monthly gains of 3x S&P 500 ETF and S&P 500 from Jan 3, 1951, to Feb. 23, 2015.
Duration Security Mean (Std. Dev.) Min, Max
Monthly gains S&P 500 8.8% (16.5%) -37.6%, 44.2%
  3x ETF 29.4% (55.5%) -85.5%, 191.7%
Annual gains S&P 500 0.57% (4.07%) -23.1%, 16.6%
  3x ETF 1.69% (12.4%) -68.0%, 54.0%

A few extra notes regarding annual and monthly gains:

  1. The 3x ETF has positive growth in 97.9% of years and 97.2% of months that the S&P 500 had positive growth.
  2. In years when the S&P 500 gained at least 5%, the 3x ETF never gained less than 17.1%.
  3. In months when the S&P 500 gained at least 1%, the 3x ETF never gained less than 2.3%.

Final Thoughts

To reiterate valid concerns with leveraged ETFs, potential investors should remember that investments in these funds risk huge drawdowns, potential loss of entire investment, and deterioration during periods of no net movement of the S&P 500. There is no guarantee that UPRO or any other leveraged ETF will consistently achieve its intended daily multiple going forward.

That said, I find buy-and-hold positions in UPRO very exciting. How can you not be impressed with 49.8% annualized growth over 5.7 years? I like the idea that you can realize such huge gains by betting on the market as a whole rather than trying to pick the one-in-a-hundred stock that is about to explode. The past 5.7 years have admittedly been a strong bull market, but even over the past 65 years a hypothetical 3x S&P 500 ETF averaged 29.4% annual gains. That's remarkable.

In my view, concerns about leveraged ETFs declining when the underlying index has no net change are somewhat overblown. While analysts often provide toy examples of how such deterioration can happen, historically there has only been one year since 1950 when the S&P 500 grew but the hypothetical zero-tracking error 3x ETF declined (1987). It rarely happens.

With potentially huge drawdowns or loss of your entire investment, you probably shouldn't put too much of your retirement savings in a fund like UPRO (I currently have 8% in UPRO). But in my view, a buy-and-hold position in UPRO is a great option for investors who don't mind taking on substantial risk for a real opportunity at tremendous growth.

Disclosure: The author is long UPRO.

The author wrote this article themselves, and it expresses their own opinions. The author is not receiving compensation for it (other than from Seeking Alpha). The author has no business relationship with any company whose stock is mentioned in this article.

Additional disclosure: The author obtained historical stock prices for S&P 500 and UPRO from Yahoo! Finance, and used R to perform analyses and generate figures. This material is based upon work supported by the National Science Foundation Graduate Research Fellowship under Grant No. DGE-0940903. Any opinion, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation.