I'm not referring to reverse splits. I'm referring to the fact there is tracking error and that error is multiplied by both the leverage and the reverse splits. From a mathematical perspective, it would not matter if the high on the ETF was 10 or 10000, the max gain is 100%. In order to truly replicate the payoffs of the reverse position you have to dynamically replicate the ETF as if it were an option. As the ETF goes down in price, you would have to sell more shares. In fact, you would have to sell an exponential amount of shares at exactly the right price to match the payoffs of the corresponding long. This is how options are priced btw. You can perform this function in excel by creating two columns. Start with $100. Initiate a long position in column A and a short in B. Go to any data website and dump 5 years or whatever of data into excel. Copy and paste the log daily changes. Now drag your columns down. Sum your net p&l. Get back to me with result. Now, to make this exercise more interesting, do the exact same procedure for the long corresponding ETF. Dump the data, sum the cash flows. You will NOT get the same numbers. According to you these cells should match. But they will be off by a magnificent amount. Has nothing to do with reverse splits. Reverse splits don't affect cash flows.
We're talking about a stock here. Ticker symbol TVIX. If you went to your broker in October and had them buy $1049 of TVIX, then did nothing, your position in TVIX today will be worth $3.26. End of story. Same as if you bought AAPL or any other stock that experienced splits or dividends during the time you held it. What goes on inside TVIX's holdings is irrelevant except to the extent that it drives the fall in the stock's traded price. In between then and now the (for example) 100 shares of TVIX you bought at $10.49 per share would become one share of TVIX worth $3.26, but at the end of the day your position went from $1049 to $3.26. The reverse splits don't impact anything regarding the value of your holding, you go into the close on the day of a 1:10 reverse split, for example holding 100 shares worth $1 each and come out holding 10 shares worth $10 each. You have $100 worth of TVIX stock both before and after, nothing is magnified by anything. I think you're overthinking this and thinking about how the TVIX doesn't track its underlying over time because it rebalances daily for a percentage change, but that doesn't change the actual price at which the stock TVIX changes hands and the split adjusted impact of that change in price to someone who buys and holds the stock.
You completely misunderstood my post. Forget about the reverse splits. Back them out. Run the exercise in excel, you get the same results. Better yet, go to their website. Download their pricing model. Put that into excel. Go back and grab the data from 2008 when the vix went to 90 or whatever the high was. Synthetically re-create the TVIX since I don't believe it was trading then. Tell me what you see. Hint: this has already been done by quants and the white papers are published. But it's a good learning exercise. I'm not trying to be snide here, but this is the best way to learn is to roll up your sleeves, get a pot of coffee going and build some models. You learn a lot more that way then looking at a chart and saying, oh wow, look at the high on this thing, it's at 1,450 and it's trading at a buck. I'm not directing this at you, but the general population that replaces charts where math and models need to be.
Again I think you're talking about what goes on inside TVIX and we're talking about the quoted price of the security, we're probably in violent agreement overall. No offense taken on the models, I've spend a ton of time nestled up with TVIX and the other VIX ETFs and Excel and agree with your assertions on how it (doesn't) track VIX as well as the impact of daily rebalancing. BTW, how about those VXUP and VXDN ETFs. Talk about a lack of modeling and maths, it's been fun to watch the reality replace the hype on those guys!
Let me further add, I believe off the top of my head (maybe wrong here) but TVIX went through a major formula change a few years ago to improve the tracking error. The chart of course does not show that. But please read up on it, there are some blog posts about it. It doesn't change the math, just the pricing function.
There was good research on this I saw awhile back and if I find it, I'll post it. I understand what you are saying about price, but the question that has to be asked, is can we replicate the cash flows. The easiest (well not that easy) way is to try to build your own ETF that replicates the one in question, TVIX. Use the same log daily changes in prices, that part is easy. Then go back to the period in question on the chart (obviously the 2008 highs in vol) and see if you can actually capture that cash flow. As you said, the share price is simple math. But you understand what I'm saying right, I should be able to buy 100 shares of TVIX say in 2007 when the vix was at 12 or whatever, ride through the credit crisis, and sell at the high (everyone on ET does right) and capture that exact cash flow. If we run that exercise and can't do it (and we can't) then we know there is an error in the model. This is a tedious exercise of course and it's one of the reasons so many new volatility ETFs are created. The promise of being able to "capture" that vol. It's VERY VERY hard. Trust me. A lot of people are spending a lot of time trying to do it. And some probably have built their own models that can but those models will NEVER make it to the public in the form of an ETF. You get the off the shelf junk that doesn't work.
It would have been nice if you had explained decay to the original poster. Your reply to them did sound a little snide. JMHO FWIW.
Not yet, and probably not for a while. if you believe in the JUNE Fed rate hike, then there is a nice move coming. But I think we go sideways until December.
Careful 2012 TVIX Disaster Issuer Credit Suisse stopped issuing TVIX shares in February 2012. Due to the limited supply, the TVIX started to rise as premiums between the net asset value (NAV) and the market price had risen as high as 90%. Retail investors were in for a rude awakening on March 22, 2012, when the TVIX collapsed 29.3% and proceeded to fall another 29.8% the following day. The TVIX fell from $14.43 to a low of $7.16, a drop of over 50% in 48 hours. Even more shocking was that the price collapse was completely unrelated to the underlying moves of the VIX Index. In fact, the VIX Index actually rose higher on that second day. Immediately after the sell-off, Credit Suisse released a statement that it would resume issuing shares again. The suspicious timing of the sell-off and news release led to many class action lawsuits. This is a cautionary tale for investors to always check the premium between the market price and NAV of an ETN product. Most importantly, the TVIX pricing is completely market driven without structured pricing mechanisms in place. It is not a product for long-term investment, nor for unsophisticated investors. Read more: (TVIX) VelocityShares Daily 2x VIX Short-Term ETN: Who is Invested? | Investopedia http://www.investopedia.com/article...-shortterm-etn-who-invested.asp#ixzz48TyICiKx Follow us: Investopedia on Facebook http://www.investopedia.com/article...hortterm-etn-who-invested.asp?partner=YahooSA
This is a risk with any ETN. At some size the issuer can't hedge the underlying any more and closes the creation unit process, allowing the note to trade at huge premiums to NAV.