For the group prices average 9x dcf as opposed to 8x dcf only 6 months ago. Some (VNR, LGCY, BBEP) trade above 10x dcf. This naturally raises questions how much higher the prices can go.

Looking back in history can give us clues. Back in 2007 BBEP and LINE traded at about 5% distribution yield, so perhaps 19x dcf, as opposed to 9% yield/10 x dcf now. Can we go back there?

If you plot against each other: BBEP, LINE, APA, CHK and OIL you discover they all had a sharp peak about 7 years ago:

APA and CHK had a peak in June 2008, corresponding to the peak in OIL (and the "peak oil" brouhaha). This is what you would expect. But, mysteriously, the peak in BBEP and LINE came a whole year earlier (May 2007). By the time of "peak oil" BBEP and LINE, though experiencing a bit of an up blip, were both down 30%+.

This seems a little weird. Does anyone around here remember what happened? Why were the upstreams so sharply up in 2007 and so sharply down by 2008 despite "peak oil"?

(Interestingly, following the 2009 crash, LINE went on to recover much of its May 2008 peak during 2010-2011 when it traded at 7% yield - right until the Hedgeye attack last fall which took it down to its present valuation; but BBEP never climbed back anywhere near its 2007 highs).

]]>For the group prices average 9x dcf as opposed to 8x dcf only 6 months ago. Some (VNR, LGCY, BBEP) trade above 10x dcf. This naturally raises questions how much higher the prices can go.

Looking back in history can give us clues. Back in 2007 BBEP and LINE traded at about 5% distribution yield, so perhaps 19x dcf, as opposed to 9% yield/10 x dcf now. Can we go back there?

If you plot against each other: BBEP, LINE, APA, CHK and OIL you discover they all had a sharp peak about 7 years ago:

APA and CHK had a peak in June 2008, corresponding to the peak in OIL (and the "peak oil" brouhaha). This is what you would expect. But, mysteriously, the peak in BBEP and LINE came a whole year earlier (May 2007). By the time of "peak oil" BBEP and LINE, though experiencing a bit of an up blip, were both down 30%+.

This seems a little weird. Does anyone around here remember what happened? Why were the upstreams so sharply up in 2007 and so sharply down by 2008 despite "peak oil"?

(Interestingly, following the 2009 crash, LINE went on to recover much of its May 2008 peak during 2010-2011 when it traded at 7% yield - right until the Hedgeye attack last fall which took it down to its present valuation; but BBEP never climbed back anywhere near its 2007 highs).

]]>The whole futures curve of VIX has been trading at a higher premium to "spot" than usual - a kind of iv effect, I suppose: perhaps the last spike really scared people, or maybe it's Ukraine.

This is more paradoxical than usual, but even the usual (when there is less of a futures premium on VIX) is paradoxical: VIX itself is calculated from the relative prices of calls and puts on the S&P, which themselves reflect market participants expectations of the future. There is implied volatility built into those options. Elevated VIX futures prices therefore mean that VIX futures traders build additional implied volatility into their estimate of future volatility and this perforce goes to zero at expiry because VIX futures are cash settled - yes - but based on the price of the S&P options strip at expiry. ("The final settlement value for VIX futures shall be a Special Opening Quotation (SOQ) of VIX calculated from the sequence of opening prices of the options used to calculate the index on the settlement date. Etc." See here).

This seems to create an opportunity for arbitrage for a well capitalized fund. Long the strip, short the future. If one could figure out a way to hold some version of VIX spot, the trade would be similar to writing a covered call but better - and risk free. The trade would be placed for credit and the credit would be equal to cost of underlying plus time value built into the future (the amount of the contango). 1 month contango usually runs 3 to 6 percent. Hard to believe no one has thought of that?

The difficulty lies in not being able to hold VIX itself and perhaps in it being impossible to construct a combination of S+P options that reflect it; or, if such a position is possible to construct, perhaps in having to constantly roll it thus incurring a profit-destroying cost. A simpler strategy might perhaps be to long S+P options and short VIX via VIX ETF derivatives.

]]>The whole futures curve of VIX has been trading at a higher premium to "spot" than usual - a kind of iv effect, I suppose: perhaps the last spike really scared people, or maybe it's Ukraine.

This is more paradoxical than usual, but even the usual (when there is less of a futures premium on VIX) is paradoxical: VIX itself is calculated from the relative prices of calls and puts on the S&P, which themselves reflect market participants expectations of the future. There is implied volatility built into those options. Elevated VIX futures prices therefore mean that VIX futures traders build additional implied volatility into their estimate of future volatility and this perforce goes to zero at expiry because VIX futures are cash settled - yes - but based on the price of the S&P options strip at expiry. ("The final settlement value for VIX futures shall be a Special Opening Quotation (SOQ) of VIX calculated from the sequence of opening prices of the options used to calculate the index on the settlement date. Etc." See here).

This seems to create an opportunity for arbitrage for a well capitalized fund. Long the strip, short the future. If one could figure out a way to hold some version of VIX spot, the trade would be similar to writing a covered call but better - and risk free. The trade would be placed for credit and the credit would be equal to cost of underlying plus time value built into the future (the amount of the contango). 1 month contango usually runs 3 to 6 percent. Hard to believe no one has thought of that?

The difficulty lies in not being able to hold VIX itself and perhaps in it being impossible to construct a combination of S+P options that reflect it; or, if such a position is possible to construct, perhaps in having to constantly roll it thus incurring a profit-destroying cost. A simpler strategy might perhaps be to long S+P options and short VIX via VIX ETF derivatives.

]]>But, reading through the 10K, I have encountered some difficulty in trying to understanding the business plan. I am sure it is a good plan, I just would like to understand it! Can anyone help?

Here are numbers taken from ARP's latest 10K:

Gas and oil production revenues 2013: $ 266,783

Gas and oil production expenses 2013: $97,237

Clearly, production operations are cash flow positive. However,

Depletion, depreciation and amortization 2013 associated with gas and oil production: $129,729

Total SGA: $78,063

Clearly, only a portion of SGA apply to Gas and oil production. Since gas and oil production are 56% of revenues, I suppose we could assume that SGA costs associated with gas and oil production are perhaps $78,063 x .56 = $43,715.

If so, then the net result of the gas and oil production would appear to be negative $3,898 even before giving consideration to any interest expense.

Yet, ARP continues to acquire gas and oil properties. There is an entrepreneurial logic here which escapes me. If the business is money losing, why keep buying more?

Is it because:

1) SGA expenses are largely fixed and as total gas and oil business goes, they will eventually decline as share of revenues and the business will become profitable?

2) SGA expenses associated with gas an oil production are in fact a much smaller portion of total SGA expenses than the apparent 56%?

3) Gas and oil are currently underpriced and one expects prices to increase in the future? But in this case would it not make more sense to just buy reserves and hold off exploiting them until prices have risen?

4) Although depletion is said to be calculated per unit of production, it somehow overstates the rate of consumption of reserves?

5) Some other reason I have not understood?

]]>But, reading through the 10K, I have encountered some difficulty in trying to understanding the business plan. I am sure it is a good plan, I just would like to understand it! Can anyone help?

Here are numbers taken from ARP's latest 10K:

Gas and oil production revenues 2013: $ 266,783

Gas and oil production expenses 2013: $97,237

Clearly, production operations are cash flow positive. However,

Depletion, depreciation and amortization 2013 associated with gas and oil production: $129,729

Total SGA: $78,063

Clearly, only a portion of SGA apply to Gas and oil production. Since gas and oil production are 56% of revenues, I suppose we could assume that SGA costs associated with gas and oil production are perhaps $78,063 x .56 = $43,715.

If so, then the net result of the gas and oil production would appear to be negative $3,898 even before giving consideration to any interest expense.

Yet, ARP continues to acquire gas and oil properties. There is an entrepreneurial logic here which escapes me. If the business is money losing, why keep buying more?

Is it because:

1) SGA expenses are largely fixed and as total gas and oil business goes, they will eventually decline as share of revenues and the business will become profitable?

2) SGA expenses associated with gas an oil production are in fact a much smaller portion of total SGA expenses than the apparent 56%?

3) Gas and oil are currently underpriced and one expects prices to increase in the future? But in this case would it not make more sense to just buy reserves and hold off exploiting them until prices have risen?

4) Although depletion is said to be calculated per unit of production, it somehow overstates the rate of consumption of reserves?

5) Some other reason I have not understood?

]]>The following wikipedia entry explains why it is not possible to replicate vix using s+p options:

'The VIX is calculated as the square root of the par variance swap rate for a 30 day term^{[clarify]} initiated today. Note that the VIX is the volatility of a variance swap and not that of a volatility swap (volatility being the square root of variance, or standard deviation). A variance swap can be perfectly statically replicated through vanilla puts and calls whereas a volatility swap requires dynamic hedging.'

One would have to hold a square root of a position. Nevertheless, premium on vix futures does represent additional premium on top of s+p option premium - there is probably room for trading this fact even if it is not honest to goodness (risk free) arbitrage.

]]>The following wikipedia entry explains why it is not possible to replicate vix using s+p options:

'The VIX is calculated as the square root of the par variance swap rate for a 30 day term^{[clarify]} initiated today. Note that the VIX is the volatility of a variance swap and not that of a volatility swap (volatility being the square root of variance, or standard deviation). A variance swap can be perfectly statically replicated through vanilla puts and calls whereas a volatility swap requires dynamic hedging.'

One would have to hold a square root of a position. Nevertheless, premium on vix futures does represent additional premium on top of s+p option premium - there is probably room for trading this fact even if it is not honest to goodness (risk free) arbitrage.

]]>You need to start with 500K or more, imho. I started with 250K and the first 10 years were really hard. once you go over 500K a lot of tools kick in - you can use a lot more leverage, all fees become negotiable, you have constant access to top level research, you call your broker at 2AM and get him out of bed to trade Tokyo, maybe your margin goes into red and you can negotiate around it. And, at any rate, at that level all you need is a single digit return to put food on the table, which is a lot easier than trying to do than anything double digit.

I'm actually not that great - annualized returns are around 15% over the last 25 years but they are bought at the cost of phenomenal volatility - I can be up 250% one year and down 60% the next. Which is still a nice return, but not for most stomachs - I would imagine it a psychological impossibility if you're paying a mortgage and two college tuitions at the same time.

So the critical advice is - if you dont have the capital to do it, don't try it.

]]>You need to start with 500K or more, imho. I started with 250K and the first 10 years were really hard. once you go over 500K a lot of tools kick in - you can use a lot more leverage, all fees become negotiable, you have constant access to top level research, you call your broker at 2AM and get him out of bed to trade Tokyo, maybe your margin goes into red and you can negotiate around it. And, at any rate, at that level all you need is a single digit return to put food on the table, which is a lot easier than trying to do than anything double digit.

I'm actually not that great - annualized returns are around 15% over the last 25 years but they are bought at the cost of phenomenal volatility - I can be up 250% one year and down 60% the next. Which is still a nice return, but not for most stomachs - I would imagine it a psychological impossibility if you're paying a mortgage and two college tuitions at the same time.

So the critical advice is - if you dont have the capital to do it, don't try it.

]]>Continuing the work from last post, let's take a look at another trade. It so happens that today is Thursday, the advertised day to enter this trade. As the market has not opened yet, I will use Wednesday prices to make some rought calcualtions.

Using Nasdaq historical price data and a simple excel spreadsheet we can calculate that during any 7 day trading period over the last 10 years, GLD has moved as follows:

AVERAGE | 2.75% |

MEDIAN | 2.32% |

MIN | 0.43% |

MAX | 18.45% |

SIGMA | 1.81% |

Yesterday, GLD has closed at 126.32.

Let's look at a hypothetical trade:

Sell -100 GLD FebWk4 124 Put | $0.45 | ($4,500.00) |

Buy 100 GLD FebWk4 126 Put | $1.10 | $11,000.00 |

Buy 100 GLD FebWk4 127 Call | $0.91 | $9,100.00 |

Sell -100 GLD FebWk4 129 Call | $0.34 | ($3,400.00) |

Optionsexpress Trade Analyzer gives the following p/l estimates:

124.00 | 7800.00 | 63.93% |

124.78 | 0.00 | 0.00% |

126.00 | (12200.00) | -100.00% |

127.00 | (12200.00) | -100.00% |

128.22 | 0.00 | 0.00% |

129.00 | 7800.00 | 63.93% |

Here are the probabilities calculated from the same spreadsheet:

124 - 1.80% move - probability 66.2%

124.78 - 0.86% move - probability 97.0%

126 - 0.21% move - probability 100.00%

127 - 0.58% move - probability 99.7%

128.22 - 1.47% move - probability 79.3%

129 - 2.16% move - probability 54.8%

As I discussed in my last post, this calculation does not account for trading costs. According to optionsxpress, the transation would cost $600 to initiate. Let us assume it would cost the same to close. Thus we have a transaction which on 1.5% of cases (1-(97%+100)/2] results in total loss of $13400 ($12200+$1200); 20.25% of cases [(1-88.(97.0% + 79.3%)/2 - 1.5%] results in "breakeven" or worse (that is to say, results in a loss of the amount equal or greater than the trading fee of $1200; in 60% of cases [(66.2% + 54.8%)/2] results in maximum gain of $6600 ($7800-$1200); and the rest of the time (18.75%) of cases results in an outcome ranging between $1200 loss and $6600 gain.

Let's summarize this:

1.5%: -$13,400

20.25%: between -$13,400 and -$1,200; let's say median -$7,300

18.25%: between -$1200 and +$6600; let's say median +2,700

60%: maximum gain of +$6600

Whenever we close out this trade before expiration, there will be some time value left in the options, which will skew the results a bit but I will leave this out as I have no way to quantify this. Thus on an average weekly trade a year we would expect to make:

0.015% x (-13,400) + 0.2025 x (-$,7300) + 0.1875 x $2,700 + 0.60 x $6600) = $2,773.50

or $144,222 a year.

**************

March 6, 2014

Right?

Well, wrong.

In this particular case GLD went over 129 by Tuesday, February 25, but because iv in the 129 call jumped a lot more than iv in the 127 call, profit on the trade reached only $3,200 instead of the calculated maximum of $6600. By Wednesday GLD reversed and the trade went into negative territory. The trader who did not take profit on Wednesday hoping that another day's time decay would improve his profits, and ended up closing on Wednesday, ended up losing $2,300 instead or else risked taking the maximum loss.

In other words, theoretical maximum profit calculations are seriously off - and you have to realize that when using the trade calculator; and you have to close whenever the trade goes beyond your outside strikes.

]]>Continuing the work from last post, let's take a look at another trade. It so happens that today is Thursday, the advertised day to enter this trade. As the market has not opened yet, I will use Wednesday prices to make some rought calcualtions.

Using Nasdaq historical price data and a simple excel spreadsheet we can calculate that during any 7 day trading period over the last 10 years, GLD has moved as follows:

AVERAGE | 2.75% |

MEDIAN | 2.32% |

MIN | 0.43% |

MAX | 18.45% |

SIGMA | 1.81% |

Yesterday, GLD has closed at 126.32.

Let's look at a hypothetical trade:

Sell -100 GLD FebWk4 124 Put | $0.45 | ($4,500.00) |

Buy 100 GLD FebWk4 126 Put | $1.10 | $11,000.00 |

Buy 100 GLD FebWk4 127 Call | $0.91 | $9,100.00 |

Sell -100 GLD FebWk4 129 Call | $0.34 | ($3,400.00) |

Optionsexpress Trade Analyzer gives the following p/l estimates:

124.00 | 7800.00 | 63.93% |

124.78 | 0.00 | 0.00% |

126.00 | (12200.00) | -100.00% |

127.00 | (12200.00) | -100.00% |

128.22 | 0.00 | 0.00% |

129.00 | 7800.00 | 63.93% |

Here are the probabilities calculated from the same spreadsheet:

124 - 1.80% move - probability 66.2%

124.78 - 0.86% move - probability 97.0%

126 - 0.21% move - probability 100.00%

127 - 0.58% move - probability 99.7%

128.22 - 1.47% move - probability 79.3%

129 - 2.16% move - probability 54.8%

As I discussed in my last post, this calculation does not account for trading costs. According to optionsxpress, the transation would cost $600 to initiate. Let us assume it would cost the same to close. Thus we have a transaction which on 1.5% of cases (1-(97%+100)/2] results in total loss of $13400 ($12200+$1200); 20.25% of cases [(1-88.(97.0% + 79.3%)/2 - 1.5%] results in "breakeven" or worse (that is to say, results in a loss of the amount equal or greater than the trading fee of $1200; in 60% of cases [(66.2% + 54.8%)/2] results in maximum gain of $6600 ($7800-$1200); and the rest of the time (18.75%) of cases results in an outcome ranging between $1200 loss and $6600 gain.

Let's summarize this:

1.5%: -$13,400

20.25%: between -$13,400 and -$1,200; let's say median -$7,300

18.25%: between -$1200 and +$6600; let's say median +2,700

60%: maximum gain of +$6600

Whenever we close out this trade before expiration, there will be some time value left in the options, which will skew the results a bit but I will leave this out as I have no way to quantify this. Thus on an average weekly trade a year we would expect to make:

0.015% x (-13,400) + 0.2025 x (-$,7300) + 0.1875 x $2,700 + 0.60 x $6600) = $2,773.50

or $144,222 a year.

**************

March 6, 2014

Right?

Well, wrong.

In this particular case GLD went over 129 by Tuesday, February 25, but because iv in the 129 call jumped a lot more than iv in the 127 call, profit on the trade reached only $3,200 instead of the calculated maximum of $6600. By Wednesday GLD reversed and the trade went into negative territory. The trader who did not take profit on Wednesday hoping that another day's time decay would improve his profits, and ended up closing on Wednesday, ended up losing $2,300 instead or else risked taking the maximum loss.

In other words, theoretical maximum profit calculations are seriously off - and you have to realize that when using the trade calculator; and you have to close whenever the trade goes beyond your outside strikes.

]]>