Tell me exactly why. You sound very opinionated but you didn't give any specifics as to your opinion. ...calculus, stats, freight train, power of law, math knowledge, excel...but no real specifics
When "everyone knows" somethings is going asymptotically towards 0, a herd of people tend to short it. Since there are only so many shares available to short, this can cause a short squeeze. In fact, "enterprising" longs may realize this and attempt to cause a short squeeze by refusing to lend shares. Trust me, you don't want to be on the wrong side of a short squeeze -- no matter how worthless the security, a short squeeze can still bankrupt you. Also, borrowing costs can skyrocket easily towards 100%. I think IB (or any broker) will start passing the cost on to you if they deem their own costs "excessive". Look for the fine print. I'm not saying you shouldn't short, but just keep in mind that there are reasons not to in some cases, even when (or perhaps I should say particularly when) the security is obviously worthless. As I said in my earlier post, it really doesn't need to be that way in an era of computers, blockchains, etc. -- the current system for shorting US stocks is antiquated, probably owing to the fact that it benefits certain very large brokers, aided by public perception that shorting is in itself bad, when in reality it can be an important part of price discovery (e.g. when people short fraud stocks) as well as a source of liquidity (if MM's can't short, it can get rather difficult to buy).
The power law has to do with how fast a number can become asymptotic. The reason 3x ETF have decay is because of the stochastic nature of prices which cause the implied decay. When prices are not stochastic, they can go parabolic. For example, the 3X ETF "should" move at 3x the rate of the futures contract. However, that is only true under normal prices. It's possible that a 50% increase in NG, from say 1.80 to 2.25 could leave to a 10,000% increase in the corresponding ETF. THAT is why it's not free money to short them. And that has to do with the power law.
No need to talk all stochastic or parabolic here, nor is it an issue with the ETF not tracking correctly, or a misunderstanding of calculus or stats, or a short squeeze risk (unless they're squeezing the underlying, these ETFs trade at NAV or pretty close except in a few rare exceptions) as other posters have incorrectly surmised. The entire issue is that inverse and 2X/3X ETFs return the DAILY percentage change of the index/commodity they track. That's it, nothing more. That means in some cases they significantly diverge from returning 2X/3X/inverse of the underlying over time, but it's simple math. I highly recommend you set up a simple excel spreadsheet that simulates an underlying and the response of a fund with a daily percentage change in the next column. Play around with an underlying that goes up a set percentage each day, an underlying that goes down a set percentage per day, one that goes up for some period of time then down to the same starting point, and one that oscillates randomly. You'll get an intuitive feel for exactly what the daily percentage, no math beyond the ability to calculate a percentage is required. I'll do a quick demo here if you're not so inclined, let's say you have a regular ETF and its inverse, with the inverse on a daily percentage basis while the ETF just tracks the underlying. They both start at $10 and have no tracking error. On day 1 the underlying goes up 50% to $15. The ETF is now at $15, and the inverse goes down 50% so it's at 10*-.5=5. The next day the underlying goes back down to $10. That is a 33% decline. The ETF is now at $10, and the underlying goes up 33%, or mathematically 1.33*5=$6.66. Both the ETF and its inverse did exactly what they claimed to do, but while the ETF is back at its starting point the inverse is much lower. Same issue with 2X/3X that provide daily percentage performance.
Actually the NG futures look weak, but UGAZ showed some strength. Might've been fooled. UGAZ looked good for awhile but Nat Gas futures looked awful. I am not long UGAZ.
Thanks, Sig, You explain why it might be a GOOD idea to short leveraged ETFs as long as the fees aren't high; I have seen fees for them at IB at 5%/year. If a leveraged ETF can deteriorate a lot through volatility, it will earn much more than 5%/year. I am very intrigued by the idea of shorting the 3X pairs. They both deteriorate, and shorting the same amount of both should be close to a "low- risk, sure thing." Occasional rebalancing, and a cash cushion in the account would be required. Bry
Yeah, here is your "simple" math. https://www.math.nyu.edu/faculty/avellane/LeveragedETF20090515.pdf And here is a good piece for the more retail oriented crowd "math light" http://seekingalpha.com/article/3070136-shorting-leveraged-etf-pairs-easier-said-than-done