Leveraged Daily Return ETFs

[jschembs]

Has anyone looked into puts on the ETFs designed to deliver leveraged daily returns for the underlying? My understanding of the math is the volatility decay dooms most of these to nearly worthless over the long run, so am I missing something that buying longer-dated puts at a certain price is almost a no-brainer?

[ItsAValueTrap]

It's a little worse than that.  They have to trade illiquid products every day.  This generates massive transaction costs over a year and is the reason why you have to pay a few to several percent to short them.

 

I've written about it on my blog:  http://wp.me/p1mOGr-dX

I have no idea if the put options are a good idea. 

 

If markets exhibit excess volatility most of the time, then it would make sense that these ETFs will lose money most of the time (and in a small number of cases, make a ridiculous amount of money).  This would affect the distribution of future prices.  You'd also have to make a forecast about the cost of the borrow.  It could be that the options are priced pretty close to what they should be; the leveraged ETFs aren't that difficult to analyze.  Of course I could be wrong because I haven't really looked into playing the options on these ETFs.

[AchilliesValue]

I've actually spent some time thinking about this strategy as well.  In particular buying puts on TNA.  I ended up not doing it and I think there are two main risks.  The implied volatility on the puts is very high so while the underlying security gets killed in time decay your also getting killed in the instruments time decay.  I think the other risk is that for this thesis to work you need the ETFs to in effect "blow up" and I was concerned about being able to cash out my position if that in fact happens.  You could exercise your now deep in the money options but then your essentially writing off everything but the intrinsic value of the option.

 

I think this could work really well but you'd almost have to get the timing perfect.  I've never been good at that so I passed.

[bz1516]

The performance of these funds is dependent on the volatility environment they operate in.  In periods of extreme volatility they will decay rapidly along the lines that the standard narrative on them describes.  However in periods of low volatility the decay is more or less inconsequential.  The pattern of volatility is also very important.

 

In the current period of very low volatility there is no reason these funds cannot be held for long periods.  A rule of thumb I use is they are safe to use if the VIX is below 25 and unsafe if the VIX is above 40.  The VIX is only loosely related to the kind of volatility that cause these funds to decay so it is truly a rule of thumb and not a clear rule.  The point being the value in holding the fund will be greater than any decay if the VIX is below 25 and conversely the decay will outweigh the intended value of the position if the VIX is over 40.

[RichardGibbons]

The decay from volatility is built into the option prices.

 

For instance, back in June, when UVXY was trading at about $70, I bought 2015 $70 strike puts at the ask for something like $47.  Clearly they were priced to include the various forms of decay from which they suffer. 

 

Now that UVXY is at $20, they're trading for 52-56, though I sold them a while back, because, while I still think these will almost always be profitable, I've since decided that other short volatility ideas are better, and I didn't want too much exposure to short volatility.

[AchilliesValue]

Now that UVXY is at $20, they're trading for 52-56, though I sold them a while back, because, while I still think these will almost always be profitable, I've since decided that other short volatility ideas are better, and I didn't want too much exposure to short volatility.

 

Do you mind sharing some of your other short volatility ideas?  I thought about trying to be clever and look companies with high stock compensation that in theory would be forced to report much lower GAAP earnings as volatility rises, but I get the impression they can cook the implied volatility a bit and anyway most the companies that have high stock compensation exclude it in "adjusted earnings" anyway so it doesn't impact the price performance as much as it should. 

[RichardGibbons]

The whole key to my volatility strategy is that the futures are mostly in contango, and therefore ETFs built by rolling one month into the next constantly lose money, and the inverse of those ETFs constantly gain money.

 

The simpler of my two positions now is a ZIV long.  As a result of contango in the medium-term futures on which is is based, I think it will deliver strong, but often bumpy returns.  I think it will be comparable to the UVXY short, but less volatile, with fewer extreme outcomes, and also gives me the option of tax-deferred compounding for years.

 

The more complex is long SVXY (basically the same as XIV), long way out of the money SVXY puts, selling 10% of the shares every time it goes up 20%, and buying occasionally as it falls when I feel like it. 

 

SVXY has a simple average return of something like 80% per year.  However, the challenge is that there is a big difference between the simple average and the compounded average, because it will go to zero at some point, and will fall dramatically every so often.  (e.g. look at the stock chart for XIV at the end of the year 2011.  This sort of drop is a normal occurrence, not an anomaly.)

 

So, I want to get those high simple returns, but not have my compounded average returns go to zero when, say, someone detonates nuke on New York City, and short term volatility doubles overnight.  (I think it's very, very difficult for medium term volatility to double overnight, which is why I'm not so concerned about ZIV.)

 

Therefore, the puts exist so that when the VIX futures double and SVXY falls to 0 in a single day, I recoup my entire investment.  I'll also probably sell them after a long down-trend like the 2011 example above, because if the VIX futures are already really high, it's basically impossible for them to double overnight.

 

There are a couple of other non-obvious negatives.  First, if volatility is volatile, (i.e. SVXY bounces around a lot)  SVXY loses money -- though hopefully not as fast as it gains from contango.  Second, the futures are likely to go into backwardization (i.e. the opposite of contango where rolling futures hurts SVXY) when volatility goes very high.  Thus, SVXY will constantly be losing money when volatility is high.

 

So, there's the irony that, when volatility is high and SVXY is most likely to gain from reversion to the mean, it's likely also losing money daily from backwardization.  And, when volatility is low, contango tends to be high and you'll be making the most money every day from contango, but it's also the time when you're most at risk of losing money as a result of reversion to the mean, and also when a disaster will have the most impact on your position.  (That's one of the reasons why I have the sell 10% of every 20% gain rule.  A 20% gain is most likely to occur when volatility is low and I'm greedy for contango, but the rule forces me to take some profits off the table, reducing my risk and providing gunpowder for when volatility increases.)

 

It's also worthwhile noting that I haven't done these sorts of volatility trades for very long, so I don't know if it will work, but it seems plausible to me.

 

[RichardGibbons]

Bumping this since Gio asked in the painful Fairfax thread.

 

The main difference between my thoughts then and my last post are.

• I shouldn't have said the ZIV long is comparable to a UVXY short.  I'm not sure why I said that, because it isn't very similar at all, and I knew it then.  It's similar in the sense that they will mostly move in the same direction, but the magnitude of the moves and the long term results will be very different.
  
• To me now, ZIV seems superior to SVXY, just because of its lower volatility and far lower chance of going to zero.  That makes it much more tax-efficient, less complex, and less impacted by volatile volatility. I'll still do SVXY trades occasionally, but will be more opportunistic (i.e. caring more about mean reversion and less about contango).