//
Research

Latest Post

XIV: 30% upside in 2 to 6 months

Trade: Long XIV (short term trade)        Current price: 31.86        Expected price: 41.46        Upside: 30%        Time horizon: 2 to 6 months

VIX measures 30 days implied volatility of the S&P 500 options. It rises when markets are down and can increase significantly in a day depending on the velocity of the fall. When VIX remains normal to low, we see contango in the VIX futures, which means future prices are above spot price. In contango, future prices (which are at a premium to spot price) can be shorted and covered on or before the contract expiration to make a profit (aka roll yield), as future prices decline to converge with spot price when they get closer to expiration date (see figure below).
When VIX spikes, we see backwardation in the VIX futures. In backwardation future prices trade below spot prices and one can play exactly opposite to contango (go long futures and cover the position on or before the contract expiration) to make a profit.

Contango Backwardation

Contango bckwardation

However, for volatility (VIX), there are two important points to consider:

1) Mean-reversion: VIX is the most mean-reverting thing and even in the worst recessions, it hasn’t stayed at elevated levels. It would spike up and then go back to its mean or lower than mean. Below table shows percentage of times VIX has spiked since its inception in Jan 1990 (24 years of historical data), and has remained at those levels.

VIX probability table

From the table above, it is clear that when VIX level reaches 22 or more, shorting VIX becomes a high probability bet, at least historically. As of 10 October 2014, VIX closed at 21.24. Historically VIX has been at or above those levels only 34% of the times, which means approximately 66% of the times its been below 21.24. Also, at these elevated levels probability of increasing even higher decreases exponentially. For example, as evident in the table above, historically VIX of 22 and above has been recorded only 31% of the times; but 24 and above only 23% of the times. In essence, if VIX spikes up at levels above 22ish, the probability that it will go even higher reduces significantly. As long as there is no imminent recession in the short term, the spikes do not remain at elevated levels and mean revert soon. Even if S&P 500 index does not increase, overtime volatility would decline, which means VIX would go down.

2) Contango: The other advantage of shorting VIX (volatility products) at these high levels is that, historically, VIX futures have been in contango more than 70% of the time and in backwardation less than 30%. As mentioned above, Contango is good for shorting VIX futures. As you can see in the below chart, from 2009 until current date (October 2014), VIX has mostly been in Contango. That means shorting VIX would mostly make money due to roll yield. As of 10 October 2014, VIX was in backwardation, closing at 21.24. Backwardation in the volatility exists during turbulent times, and the VIX spikes and exceeds the price of one or more of its futures contracts. However all turbulent events, big or small, are temporary. When they fade VIX returns to its normal level.

VIX central Contango most of the time

source: Vixcentral.com

Strategy: XIV is an ETN that tries to replicate the inverse of the daily performance of VIX. XIV shorts varying amounts of the front and second month VIX futures contracts and covers at expiration. Because of this, XIV collects roll yield, which can be positive or negative depending on whether the VIX term structure is in contango or backwardation, respectively. Since more than 70% of the time VIX is in contango (as mentioned in point 2 above), XIV generates positive yield most of the times.

However, since XIV is an inverse ETN and is rebalanced every day, this results in a quite a bit of time decay. XIV has a high volatility, which also means bigger time decay losses.

That said, XIV can be very interesting for short-term trades if you buy it on VIX spikes and sell it in a few months when volatility subsides, due to its mean-reversion nature. Going long XIV when VIX is in backwardation has proven to be an extremely lucrative strategy as backwardation in volatility market do not last that long. With a drastic VIX spike, as we saw on 10 October 2014, VIX entered in backwardation, closing at 21.24. As mentioned above, historically VIX has been at or above those levels only 34% of the times, which means approximately 66% of the times its been below 21.24. This means shorting VIX (volatility related products) or going long XIV is a high probability bet. Current fall in the market is a correction and VIX should revert back to lower levels. If it does, XIV could see an upside of 30% or more in 3 to 6 months.

The trade could benefit in 2 ways,

  1. VIX mean-reverting from its spike and hence price of XIV increasing
  2. Lower trading and hence less volatility in the end of the year months (November and December), resulting in Contango for most part for VIX and hence positive roll yield, benefiting XIV on the up side. Historically, October has been the most volatile month for VIX since its inception in Jan 1990 (24 years of historical data), declining significantly in November, December and January.

Lastly, the below chart shows a 50 day 3 standard deviation event for XIV, which has never sustained and prices have always risen from such a dramatic decline (mainly due to VIX mean-reversion and that historically VIX futures have been in contango more than 70% of the time).

XIV 3 std dev 10 Oct

source: Yahoo finance

Disclosure: Long XIV

Disclaimer

Recent Articles

  • It is very common saying that “market cannot be timed”.  Our models below prove that wrong showing a very high probability of timing long-term market tops and bottoms. We believe in combination of quantitative and fundamental analysis. For timing the market, we try to understand the market as science and constantly experiment by back-testing combination of diverse data-points. Below models convert our belief in cyclicality, irrational human behavior and market inefficiencies into science by developing 6 proprietary signals (5 contrarian and 1 trend following) indicating long-term market tops and bottoms. When back-tested, one or more of our market timing indicators predicts every fall of more than 15% in S&P since 1965. Our belief is that market-timing is the key to making big returns and beating the market. Getting one big event right or following one big trend for years and avoiding a downturn can generate significant alpha. Below are the models:   Interest rates: This indicator only gives sell signals based on a proprietary formula. It tells when to go on defensive mode and hedge the portfolio from potential losses. The indicator has captured major downturns in S&P such as 1966 (-18% after signal in 8 months), 1969 (-26% after signal in 12 months), 1973 (-39% after signal in 15 months), 1980 (-10% after signal in 1 month), 1981 (-19% after signal in 15 months), 2000 (-46% after signal in 25 months) and 2007 (-45% after signal in 29 months). It has given 10 signals since 1962, of which most were well-timed and some a little early in the cycle. The indicator is updated on a monthly basis and is a great source to hedge or make an asymmetrical bet. Below is the chart (circled are the ones when the signal did not work).   Speculation: This is a contrarian indicator which gives buy and sell signals, based on a proprietary formula. It gauges speculation in the market. It gives sell signal when speculation reaches its peak, and vice-versa. The indicator has 100% accuracy when backtested. It gave S&P 500 sell signals in 1972 (-41% after signal in 26 months), 1983 (-7.4% after signal in 10 months), 2000 (-45% after signal in 32 months), 2007 (-51% after signal in 20 months) and buy signal in 2009 (61% after signal in 25 months). The indicator is updated on a monthly basis. Below is the chart:   Business cycle: This is a contrarian indicator which gives buy and sell signals, based on a proprietary formula. It gauges tops and bottoms of a business cycle. The indicator gave S&P 500 buy signals in 1957 (34% after signal in 14 months), 1970 (16% after signal in 13 months), 1974 (42% after signal in 19 months), 1980 (24% after signal in 9 months), 1981 (38% after signal in 20 months), 1990 (36% after signal in 15 months), 2002 (22% after signal in 15 months) and 2008 (22% after signal in 18 months). It gave S&P 500 sell signals in 1958 (did not work), 1966 (-17% after signal in 8 months), 1972 (-40% after signal in 30 months), 1978 (did not work), 1983 (-8% after signal in 9 months) and 2007 (-51% after signal in 20 months). The indicator is a great way of making an asymmetrical bet or hedging the portfolio. It is updated on a monthly basis. Below is the chart (circled are the ones when the signal did not work).   Fundamental economy: This is a contrarian indicator which gives buy and sell signals, based on a proprietary formula. It gives long term bull or bear market indicators. It gauges overheating and irresponsible spending (sell signal) and extreme conservativeness and low points (buy signal) in the economy. The indicator gave S&P 500 buy signals in: Early 1962: 2% in 12 months but S&P gaining approximately 40% in a continued bull market for next 4 years, Mid 1993: 1.3% in 12 months but a continued bull market for next 7 years until 2000, with S&P gaining more than 200%, Early 2009: 35% in 12 months and a continued bull market until now (2013), Late 2010: 19% in 8 months and a continued bull market until now (2013). while the sell signals were given in: 2000 (-38% after signal in 23 months) and 2007 (-50% after signal in 19 months). This indicator is updated on a monthly basis. Below is the chart.   Price growth: This is a contrarian indicator which gives only sell signals. It gives warning for extreme growth in price. The sell signals were recorded on: February 1929: -18% after signal in 18 months November 1936: -44% after signal in 18 months April 1956: -17% after signal in 19 months August 1987: -27% after signal in 4 months July 1998: -14% after signal in 1 month One wrong signal was given in July 1997 (circled). The indicator is updated on a monthly basis. Below is the chart.   Trend following: This is a trend following indicator which gives buy and sell signals, based on a proprietary formula. Basic goal for this signal is to catch the big up-trends in the market and avoid/short significant falls and downturns. It has done a great job by producing a return of 11.7% cagr, as compared to the S&P return of 7.1%, since April 1994 until November 2013 (approximately for last 20 years). Below is the chart:
Follow

Get every new post delivered to your Inbox.