Tuesday, March 29, 2011

Is There a Relationship Between Oil Price and Gasoline Price in the US?

With all the turmoil in the middle east, there has been a lot of talk about oil price and the resulting rise in Gasoline prices.  I downloaded daily spot prices and got yearly averages from 1986 through 2011 YTD.   The price categories are:
  • Cushing, OK WTI Spot Price FOB ($/Barrel)
  • Europe Brent Spot Price FOB ($/Barrel)
  • New York Harbor Conventional Gasoline Regular Spot Price FOB ($/Gal)
  • U.S. Gulf Coast Conventional Gasoline Regular Spot Price FOB ($/Gal)
As you can see in the graph below, Cushing Crude Price remained relatively flat from 1986 to 1999 at which time prices began a steady rise.


The next graph is for Gulf Regular Gasoline Price which follows a similar pattern on the same dates. 


These similar patterns suggest that correlations may help to see if anything has been changing in the relationship between Crude and Gasoline prices.  In the graphs below, the Cushing Crude price and U.S. Gulf Gasoline price are correlated for the period 1986 to 1999 with a resulting very strong correlation of 94% (R square).


In the table below, you will see that the correlations between the Cushing (US) and Brent (Euro) Crude price correlation weakens after 2003 while Gulf and NY Gas prices remain very strong in all three timeframes.  Looking further into the correlation between Cushing Crude Price and either Gulf or NY Gas prices, you find these correlations weaken to around 70 after 2003 (highlighted in red in the table).  This makes me want to look deeper into why the change in relationships after 2003.

1986 - 19992000 - 20022003 - 2011
Cushing Crude to Brent Crude99.099.495.8
Gulf to NY Regular Gasoline99.499.899.8
Cushing Crude to Gulf Gasoline94.096.370.1
Cushing Crude to NY Gasoline99.497.769.3

Using a best fit exponential growth line to the price data, you can determine the stable and consistent Compound Annual Growth Rate (CAGR).  In the graph below, the best fit growth line for Cushing Crude Price has been found, and this same technique is applied to to the 3 time periods for both Cushing Crude Price and Gulf Gas Price Growth and summarized as CAGR in the table following.




1986 - 19992000 - 20022003 - 2011
Cushing Crude Price Growth0.2%-7.2%6.1%
Gulf Regular Gasoline Price Growth0.2%-7.2%10.9%


So what happened in 2003 to cause the Gasoline price in the US to grow at a rate almost twice that of Cushing Crude Price.  California, in 2003 outlawed MTBE which was an additive in US gasoline in improve octane rating.  This step was taken when MTBE was found in groundwater and to be carcinoginic.  This lead to the use of Ethanol in gasoline to increase octane rating and finally the Federal Government's intervention (support) of ethanol production.  This is claimed to help reduce our dependance on foreign oil with a stabilization in gas prices.  However, the result has instead been the doubling of gas price growth relative to oil prices and the corresponding increase in corn prices to produce ethanol which in turn is driving up food prices in the US. 

Thursday, March 10, 2011

Annual Deficits and the National Debt

With all the talk about deficits and the debt, I got curious about the relationship between the two.  Currently the US Debt is about $13,528 Billion through 2010 ($14,193 Billion currently).  Thinking this was the sum of the annual deficits, I decided to add up all the annual deficits since 1862, and got only $8,010!  So where did the other $5,518 Billion come from??

Supplemental appropriations began in earnest in about 1950, in response to several natural disasters where the federal government stepped in to support the states who had typically covered the costs of these events.  The federal government was now in the insurance business, spending non-budgeted money in support of the states recovering from these disasters.  Now supplemental appropriations regularly come up outside the regular budget process. Most supplemental appropriations fall into two categories: defense supplementals, needed to fund the costs of military actions, and non-defense supplementals, almost all of which go to the cost of emergency response to national disasters like hurricanes and floods.  Below is a recent graph of the Sum of Deficits along with the Sum of Supplementals from 1950 to 2010.


As you can see, this is becoming a very large contributor to the National Debt and is done outside the normal budget process.  Then I got curious to see how the size of the supplementals compared to the defecit each year.  Below is a graph of the supplementals as a percent of the deficit from 1960 to 2010.  There are some negative numbers in those years when there was a budget surplus. 


In the years from 1975 to 1996, the supplementals were 38% of the deficit.  But in 2002 to 2010, supplementals now are almost equal to the deficit at 95% !  To note, in the years 1997 to 2001 (the white space between the two graph areas), the total deficit was $-536B and the total supplementals were $1,124B, for a net add to the National Debt during these 5 years of $588B.  So much for the "balanced budget" years of Clinton.

Seems to me that we should be hearing some talk about the Federal budget process including ALL of the defense spending, probably requiring a war reserve, and also getting out of the disaster insurance business.  Keeping all this spending as part of the budget process would put more pressure on Congress to fix the big debt creators rather than just pushing them outside the budget.

Thursday, March 3, 2011

The Real "Stats" of Professional Sports

Professional sports standings, playoffs and finals are often in the news so I was curious about the real statistics of the stats.  I looked into MLB, NFL, NHL and NBA using the most recently completed season for each sport and their final game.  I was interested to know if the winning teams were statistically better than their league competitors or was the whole season just chance, as in a coin flip!  I am using a 99.7% confidence level for my conclusions or said another way, there is a 3 in a 1000 chance I could be wrong!  Also remember, that in any single game where you win or loose, there is a 50/50 chance that either team will win, UNLESS one of them is uniquely better.

NFL plays a 16 game season.  There was one team that WAS statistically better that all the others, which was New England with a .875 season.  Of interest, there was also a statistically poor team with a .125 record which was Carolina.  All the other teams were statistically equivalent, meaning their wins and loses were just chance. (How else would the bookies make money!).   But alas, the super bowl was played not by the "best" team, but just the lucky ones.

MLB plays a 162 game season, and as is typically the case, no team had a truely winning season.  Philadelphia had the best record at .599, but they were not statistically better than any other team.  There was a statistically poor team with a .352 record which was Pittsburgh.  The World Series was played by two lucky teams and the winner was again a flip of the coin, since winning in 5 games is not statistically meaningful.

NHL plays an 82 game season, and just like MLB, there was not a superior team, statistically speaking.  But, also like MLB, there was a low performer which was Edmonton with a .330 record.  The Stanley Cup was played by two lucky teams with a coin flip winner as a 6 game final is not statistically relevant either.

The NBA plays an 82 game season also, but in this case there is something very different!  There were 4 teams that were statistically unique, 2 in the East and 2 in the West.  They were Cleveland (.744), Orlando (.720), Dallas (.671) and LA Lakers (.695), who went on to the Finals against Boston (.610).  The 7 game series again did not establish a statistical winner.  Also of interest is that there were 7 teams who all had statistically low seasons below .335.  I would assume from these results that the bookies have a more difficult time making as much money as they do in the other sports.

So the adage of "any team on any given day" does seem to ring true, except in basketball!  Watching the games sure is a pretty exciting way to watch a coin flip.

Tuesday, March 1, 2011

"Manufacturing's Dismal Decade" - A Look at the Numbers

The US Trade Deficit has been REDUCED by 37.7% after the 2008 recession!  Is this possible??

Pat Buchanan posted this column on 2/25/11, titled Manufacturing's Dismal Decade, http://buchanan.org/blog/manufacturings-dismal-decade-4612,  where he used many facts and figures on the trade deficit to make several points in his column.  I went to the Census Bureau  http://www.census.gov/econ/currentdata/ftd/ to get US trade data from Jan 1992 to Dec 2010 to cover the last decade and more.  I prefer to look at data over longer time frames to establish periods of consistent performance and when there are statistically significant changes.  In this case, I looked at total Imports, Exports and the resulting Trade Deficit (Trade Balance).

In Mr. Buchanan's first 4 paragraphs, I can confirm his percentage changes in Exports and Imports versus 2009, however, he states that the total annual deficit from 2009 to 2010 rose 33% (actually 32.8%...why reduce the precision of the first 2 paragraphs?) and then goes on to say this is "the largest percentage increase in a decade".   Here is another idea:  compare the two years after the recession to the two years before the recession. 


So why didn't Mr Buchanan tell us this: Trade deficit reduced by 37.7% after the recession rather than this is the largest percentage increase in the last decade?   Mr Buchanan, if you are writing about the last decade, please use more than 2 years of data to form your conclusions.  By the way, the two year US deficit has not been as low as it is in 2009/2010 since 9/01!

However, I get nervous using year on year indices to form conclusions which is a common technique in business, government and industry since it is NOT statistically sound.  Since trade numbers are continually increasing over time, a better approach is to use compounding growth rates with statistical limits to ensure that your indices are not calculated using a statistical outlier.  Also notice the graph above, that the last two years do not stay within the statistical boundaries.  So here is a better technique applied to Exports, since the Balance of Trade is a negative number and makes compounding problematic.

 This is best fit compound growth line with statistical limits for the 2009 - 2010 data and yields an annual growth of 16.0% since the recession.

 The two years prior to the recession yield a best fit exponential line of 16.3% annual growth.  The next highest 2 year growth rate over the last 18 years was 12%.  Seems as though we are doing pretty good post recession at 16.0% compounded growth, although from a smaller base number.
In a similar analysis of the import numbers, in the two years prior to the recession, US Imports grew at 10.5% compounded annually.  After the recession the two year growth of Imports was 18.5%, also from a smaller base.  The good news is that growth for 2010 has slowed to 11.7% helping to slow the deficit growth.  Even with these recent growth rates, it will be over 40 years before the deficit can be eliminated, so there is still work to be done in getting our products sold to the rest of the world.

Finally, in Mr. Buchanan's column, he states:
" In that decade, America ran a total of $6.1 trillion in trade deficits, more than our entire economic growth.  To finance those 10 years of deficits, America had to borrow$1,533 billion every day."
First of all, the 10 year deficit was $5.6 trillion, not $6.1 trillion.  In any case, dividing this $6.1 trillion by the 3,650 days in 10 years, the daily amount is $1,671 MILLION  not $1,533 BILLION as is stated in the column!  Pat, your article is about the trade deficit not the Federal deficit!  If we import $5.6 trillion more than we export, this deficit is not paid by the Federal Government by borrowing.  It is $1,542 million dollars that US consumers paid to "foreign" manufacturers more than "foreign" consumers paid US manufacturers, which they might have borrowed on their credit cards!  The Feds are not involved in this deficit, other than collecting tariffs.

Do you believe everything you read??  Check it out before you act on it!  If you have other data you would like analyzed, send it to me.  I am planning on working on the Federal Deficit, Medicare, and Social Security soon, so come back