Author Topic: Leveraged ETF question  (Read 4417 times)

Jags4186

  • Pencil Stache
  • ****
  • Posts: 596
Leveraged ETF question
« on: October 19, 2015, 10:04:28 AM »
I've been thinking more and more about leveraged ETFs.  I understand that there are volatility risks, however is there a long term scenario where a 2x or 3x leveraged ETF would underperform the market--assuming the market goes up of course.

For example,  if I were to buy UPRO and, assuming in 15 years from the now the market will be higher than it is today (as every 15 year period has been) and I never intended to sell this position until 15 years from now, what would have to happen for me to underperform the market?  Even with beta decay and management fees, it seems that if I even get 1.5X performance vs 3x its worth being invested in the fund.

I've been toying with the idea of putting 10k of my Roth IRA into this ETF.

Thoughts?

DaveR

  • Stubble
  • **
  • Posts: 244
Re: Leveraged ETF question
« Reply #1 on: October 19, 2015, 10:31:23 AM »
UPRO 5-yr return: 32.81%
Market 5-yr return price: 32.79%
(reference: http://www.proshares.com/funds/upro_performance_and_quote.html)

So 3x leveraged ETF outperformed by 0.02% in the last 5 years. Expense ratio of 0.95%.

I can buy the market at 0.05% ER. Paying 0.90% more ER to get 0.02% extra gains doesn't add up. What am I missing?
« Last Edit: October 19, 2015, 10:43:25 AM by DaveR »

DaveR

  • Stubble
  • **
  • Posts: 244
Re: Leveraged ETF question
« Reply #2 on: October 19, 2015, 10:42:31 AM »
What am I missing?

Apparently, I'm missing basic reading comprehension. UPRO page is Market Price (not return).

S&P in same period: 13.34%
So, UPRO outperformed 19.47%

Sure ER is 19x higher, but I'm now a little intrigued.

Jags4186

  • Pencil Stache
  • ****
  • Posts: 596
Re: Leveraged ETF question
« Reply #3 on: October 19, 2015, 10:49:41 AM »
Actually after posting this I did a little bit more research.  I used SSO (2x) because it's data goes back to 2006 vs UPRO (3x) which has only been around since 2009 and is in only an up market.

If you had put in 10k January 2, 2007 on SSO and $1000 in SPY January 2, 2007  SPY Would be ahead right now:

1/2/07 - 9/30/15 returns on $1,000:

SPY: $1,621
SSO: $1,433

However, if you start with $1000 and DCA in $100 a month same time period you get different results:

SPY: $18,057
SSO: $23,543


That is the intriguing part, IMO.

beltim

  • Magnum Stache
  • ******
  • Posts: 2804
Re: Leveraged ETF question
« Reply #4 on: October 19, 2015, 11:18:59 AM »
For example,  if I were to buy UPRO and, assuming in 15 years from the now the market will be higher than it is today (as every 15 year period has been) and I never intended to sell this position until 15 years from now, what would have to happen for me to underperform the market? 

In a word, volatility.  Volatility can make your return much more or less than 3x depending on the circumstances.

If you just want leveraged returns, buying a deep in the money call (with far away expiration date) or a e-mini future gets you leverage at a lower cost without the volatility risk. 

DaveR

  • Stubble
  • **
  • Posts: 244
Re: Leveraged ETF question
« Reply #5 on: October 19, 2015, 11:58:11 AM »
In theory we'd expect to see a 2x or 3x multiple whether up or down. In practice, it's a matter of correlation. UPRO claims a 0.99 correlation...we'll assume that holds for down markets too.

So the question is...do investors have the balls to stay in a fund when losses are 2x or 3x the market? SSO survived 2008. How many funds didn't? Are we dealing with survivorship bias? If I put money in UPRO how will it tangle with the next bear that comes its way...it probably won't do me a lot of good if ProShares merges the fund into another.

There are a number of risks that this type of fund introduces (what else would you expect for 2x or 3x returns). And they are not tax efficient. E.g. 2014's 38.00% return was 21.57% after taxes on distributions and sale of shares (see prospectus).

Back to your original thinking... over 15yr horizon, since the market averages "up" then a leveraged fund should be "up" too. I'd have to crunch the numbers (I'd do a MC simulation) to see what .95ER and tax effects have. In the end, the roller coaster would be... a wild ride. Whether or not it would be a fun ride depends on your definition of fun.

NorCal

  • Pencil Stache
  • ****
  • Posts: 616
Re: Leveraged ETF question
« Reply #6 on: October 19, 2015, 12:21:11 PM »
Leveraged funds are leveraged to the DAILY price movements of the underlying index, not, the long term performance.  This is a fundamental difference.  If you want to day trade, these can be a good vehicle.

If you want to invest for the long term, these are a horrible investment.  You need to understand what these products are made up of.  Last I checked (years ago), these were all made up of futures contracts.  And just like any other futures contract, you'll get eaten alive by the roll-yield if you's simply buying it and expecting it to make a positive return over many years.  Also, futures don't pay dividends, which have historically made up ~50% of the long term returns on equities.

In fact, I once read an article by a sophisticated investor who created a "risk-free" trade by shorting both the long and short leveraged ETF of the same index.  It was guaranteed to return a small rate of return over time.

beltim

  • Magnum Stache
  • ******
  • Posts: 2804
Re: Leveraged ETF question
« Reply #7 on: October 19, 2015, 12:25:26 PM »
In theory we'd expect to see a 2x or 3x multiple whether up or down. In practice, it's a matter of correlation. UPRO claims a 0.99 correlation...we'll assume that holds for down markets too.

No.  The question is not correlation, it's volatility.  Leveraged ETFs are rebalanced daily.  Let's take a look at some examples:
Example 1: Underlying index up 5 days in a row, 2.0% each day.
Index after 5 days: (1.02 ^ 5) = 1.104, so up 10.4%.
3X Leverage: (1.06 ^ 5) = 33.8%.
33.8 / 10.4 = 3.25
When the market has multiple up days in a row, the leveraged investment returns more than 3x the underlying investment.

Example 2: Underlying index down 5 days in a row, 2.0% each day.
Index after 5 days: (0.98 ^ 5) = 0.904, so down 9.6%.
3X Leverage: (0.94 ^ 5) = 0.734, so down 26.6 %.
26.6/ 9.6 = 2.77
When the market has multiple down days in a row, the leveraged investment returns less than 3x the underlying investment.

Example 3: Index day 1 up 2.0%, day 2 down 2.0% day 3 up 2.0%, day 4 down 2.0%, day 5 up 2.0%, day 6 down 2.0%.
Index after 6 days: (1.02) ^3 * (.98) ^3 = .9988, so down 0.12%.
3X Leverage: (1.06) ^3 * (.94) ^3 = .9892, so down 1.08%.
1.08 / 0.12 = 9
When the market has volatility, the leveraged investment has MUCH greater losses than the index (in this case, 9x).

And these examples are over 5-6 days.  Over months or years, the results are much less predictable, but depend on the amount of volatility.  Leveraged ETF returns are path-dependent - that is, they can't be predicted from just the starting and ending points of the underlying index.

DaveR

  • Stubble
  • **
  • Posts: 244
Re: Leveraged ETF question
« Reply #8 on: October 19, 2015, 01:11:29 PM »
Thanks for the examples beltim. Short and sweet, without the hassle of running some simulations like I might normally do. (Though I still will probably play around a little because I like learning.)

beltim

  • Magnum Stache
  • ******
  • Posts: 2804
Re: Leveraged ETF question
« Reply #9 on: October 19, 2015, 01:19:30 PM »
Thanks for the examples beltim. Short and sweet, without the hassle of running some simulations like I might normally do. (Though I still will probably play around a little because I like learning.)

No problem.  For some real-world examples, the last 6-months returns:
SPY: -2.37%
UPRO: -9.17%
Ratio = 3.87

1 year:
SPY: 7.71%
UPRO: 21.1%
Ratio = 2.74

Apr 9, 2010 - Sep 30, 2011:
SPY: -2.93%
UPRO: -19.75%
Ratio = 6.74

Telecaster

  • Handlebar Stache
  • *****
  • Posts: 1544
  • Location: Seattle, WA
Re: Leveraged ETF question
« Reply #10 on: October 19, 2015, 01:36:17 PM »
I've been thinking more and more about leveraged ETFs.  I understand that there are volatility risks, however is there a long term scenario where a 2x or 3x leveraged ETF would underperform the market--assuming the market goes up of course.

For example,  if I were to buy UPRO and, assuming in 15 years from the now the market will be higher than it is today (as every 15 year period has been) and I never intended to sell this position until 15 years from now, what would have to happen for me to underperform the market?  Even with beta decay and management fees, it seems that if I even get 1.5X performance vs 3x its worth being invested in the fund.

I've been toying with the idea of putting 10k of my Roth IRA into this ETF.

Thoughts?

I think if the market trends sideways for a few years--which it has done many times before--you will run a sizable risk of blowing your fingers off.  Here's why:

Let's say you invest $100 each in SSO and SPY.   Let's say the market goes up 5%, so you have you $105 in SPY and $110 in SSO (because it doubled the market).  The next day, the market drops 5%.   In SPY you have $99.75 (5% of $105 is $5.25).  And in SSO you have $99.50.     Now do that same experiment two more up/down trading days, and SSO gets even farther behind.  Now continue that experiment for a year.   You wind up losing some in SPY, and a huge amount on SSO.    If there are long periods of sideways trading or a bear, you need a monster bull to get back to where you were.   

If fact, let's move away from hypothetical.   I just pulled up the chart on Yahoo!.  Since inception in July 06 (or max on the chart anyway), SSO is beating SPY, so that's good.  But note that SSO was trailing from 2007 until May 2014, and it took a raging bull market in order to make up those lost gains.  One of the longest bull markets in history in fact, and you needed almost every inch of that bull just to match the index. 



 


mrpercentage

  • Handlebar Stache
  • *****
  • Posts: 1236
  • Location: PHX, AZ
Re: Leveraged ETF question
« Reply #11 on: October 19, 2015, 07:21:06 PM »
You might wan to read this. Im pretty sure this guy is a genius.

http://seekingalpha.com/article/3581376-the-real-cost-of-hedging-with-leveraged-etfs

DaveR

  • Stubble
  • **
  • Posts: 244
Re: Leveraged ETF question
« Reply #12 on: October 19, 2015, 10:19:17 PM »
You might wan to read this.

Good stuff. Beta slippage, plus shows UPRO, SSO (both long) as well well as a few short ETFs during 2015's volatility.

And I fired up R for a few [hundred thousand] sim run. You see the effects pretty clearly. If you're good at market timing, then leverage is pretty sweet. But I'm a lousy market timer, so I'll stick to boring ol' indexing for now.

NorCal

  • Pencil Stache
  • ****
  • Posts: 616
Re: Leveraged ETF question
« Reply #13 on: October 19, 2015, 10:31:33 PM »
You might wan to read this. Im pretty sure this guy is a genius.

http://seekingalpha.com/article/3581376-the-real-cost-of-hedging-with-leveraged-etfs

Excellent article.  My advice:  If you don't understand that article, don't invest in these products. 

If you do understand that article, you should have better investment methods than leveraged ETF's.

hodedofome

  • Handlebar Stache
  • *****
  • Posts: 1211
  • Age: 39
  • Location: Texas
Re: Leveraged ETF question
« Reply #14 on: October 20, 2015, 09:45:10 PM »
As with most methods the best time to invest in it, is after its sucked as an investment. Difficult emotionally to do, but if you find a year or two where a leveraged ETF has vastly underperformed the underlying index, that may be a good time to get in.

Single Stock Futures may be a better way to get some leverage for those with smaller accounts and can't do the e-mini.


Sent from my iPhone using Tapatalk