Copper Price Forecasting (2013 Update)

In a 2012 post we looked at the copper price forecasts of Cochilco since 2005. To do so we looked at the "Informe Trmestral del Mercado de Cobre" which is published on a quarterly basis.

The following graph shows the development of the forecasted copper price (or rather the deviation from the ex-post price) as a function of the number of days before the end of the period for which the forecast was made.

Separately we show an update of the graph we already showed in last year´s post comparing the deviation of the forecasted price made 18 months before the end of the forecasted period with the spot price at the forecasting date.

2012 was a relatively good year with a relatively small forecasting error, undoubtedely assited by relatively low volatility of copper price and the absence of large price movements

Metal and Ore Exports (as a percentage of exports) 1988 to 2010

The Worldbank has nifty page allowing to plot numerous trade related data points:

The World Integrated Trade Solution (WITS) is a software developed by the world Bank, in close collaboration and consultation with various International Organizations including United Nations Conference on Trade and Development (UNCTAD), International Trade Center (ITC), United Nations Statistical Division (UNSD) and World Trade Organization (WTO). WITS gives you access to major international trade, tariffs and non-tariff data compilations:

• The UN COMTRADE database maintained by the UNSD: Exports and imports by detailed commodity and partner country
• The TRAINS maintained by the UNCTAD: Imports, Tariffs, Para-Tariffs & Non-Tariff Measures at national tariff level
• The IDB and CTS databases maintained by the WTO: MFN Applied, Preferential & Bound Tariffs at national tariff level

WITS is a data consultation and extraction software with simulation capabilities. WITS is a free software. However, access to databases themselves can be fee-charging or limited depending on your status. WITS is a system that is still evolving and we will be adding more features. In subsequent releases of WITS we plan to provide additional features including coupling with ITC's MACMAP system.

We used WITS to plot the share of metals and ore exports as a percentage of total exports for the ten largest copper exporters. Zambia followed by Chile and Peru lead this metric through the whole period (1988 to 2010).

Earliest National Day Celebration Events

In case you ever wanted to know which countries celebrate the earliest events in their national holiday, here is a list for those countries commemorating events which occurred before 1800 (source mostly Wikipedia).

Year	Country	Celebrated Event
660 BC	Japan	Coronation Day First Emperor Jimmu
301	San Marino	Independence from the Roman Empire
461	Ireland	Death of St. Patrick
1291	Switzerland	Alliance against the Holy Roman Empire
1492	Spain	Columbus Discovers America
1523	Sweden	Election of Gustav Vasa as King of Sweden
1580	Portugal	Death of National Poet Luís de Camões
1776	United States	Declaration of Independence from Great Britain
1788	Australia	Founding of Sydney
1789	France	Storming of the Bastille
1791	Poland	Constitution of the Polish–Lithuanian Commonwealth

The list does not include sub-national territories where Scotland (70), England (303), Wales (589), Faroe Islands (1030), Minorca (1287), Catalonia (1714) and Sardinia (1794) also celebrate events before 1800.

Apologies if I have missed any event in the above list.

Chilean Lottery (Polla) Numbers for 2011

We looked at the financial numbers for 2011 and 2010 for Polla Chilena, one of the two official lotteries in Chile.

The basic numbers look as follows (in thousand CLP):

* Lottery Revenues can be categorized in the various games as follows (most of the revenues are from traditional lotto):

** From the Payouts there is a 2% fee going to the Handling Agent and a 15% tax on lottery winnings.

*** The Various Beneficiaries are detailed as follows

This gives the following overview on where the revenues go (handling agent fee added to administration cost, all taxes and state beneficiary summed up):

The following comments can be made when comparing these metrics to other lotteries:

• low winner payout (typically in the 40-50% range)
• high administration cost (normally less than 10%)
• high take by the state (normally less than 20%, if that)
• very low distribution to good causes (even when including the amount given to the "Instituto Nacional del Deporte", normally above 30%)

Reader comments as always welcome.

Chile´s Gini Coefficient (calculated from tax records)

Chile tax authority (Servicio de Impuestos Internos, SII) provides an annual statistics of the number of tax payers and taxable income by tax bracket (here).

This information looks as follows (this information includes all types of taxable incomes):

With this information on hand the following Lorenz curve can be constructed, which looks as follows:

Assuming (not entirely realistically) linear income distribution within each tax bracket, the Gini coefficient can be calculated. The following should be noted:

• calculation is based on individuals and not households
• calculation only incorporates taxable income and not other forms of income (i.e. state transfer, tax deductions, undeclared incomes)

Also the Gini coefficient was calculated for the after tax income and assuming the tax intake was equally distributed among all taxpayers after redistribution. The effect is actually minimal, which should not be surprising given that despite high marginal rates (40%) only 0.28% of all taxpayer (24'400 individuals) were paying 58.73% of all income taxes in 2012. This should be kept in mind if higher marginal tax rates would ever be suggested as a policy instruments for higher equality (keeping in mind that the fairest taxes are those with a large tax base but low tax rates).

Interestingly despite the shortcomings of the calculation (see above) the results are close do the numbers published in other sources (0.521 for 2009 according to the Worldbank):

German Billionaires: Data Reconciliation

We  looked at the billionaire's lists available from Forbes and MM (Manager Magazin behind paywall, the 2010 list is available in Wikipedia) and compared the entries. To convert the numbers in Manager Magazin we used a exchange rate of EURUSD = 1.30.

Most strikingly the Forbes list is much smaller (55 entries with a total of USD 251.3 billion) than the MM list (156 entries with a total of 482.4 billion)

The differences in more detail are as follows:

• Five entries with a total wealth of USD 12.1 billion (according to Forbes) are not recorded in the MM billionaire list:
• Three entries (Schoeller, von Opel, Thurn und Taxis) appear in MM but significantly below the USD 1 billion threshold
• Ludwig Merckle is recorded with with USD 5.3 billion in Forbes, while there is significant valuation uncertainty after the near collapse of the group in 2008/2009, the group has successfully deleveraged und is probably worth above USD 1 billion
• Finally Vladimir Iorich (according to Forbes Russian born, German nationality and Swiss residence) is neither recorded by MM nor by the Bilanz list of billionaires resident in Switzerland)
• The Herz family is recorded with five entries in Forbes (Michael, Wolfgang, Günter, Daniela, Ingeburg), while MM only has the two family entries (one for Günter and Daniela and another one for Ingeburg and her children)
• 109 entries with a total wealth of USD 248.6 billion (according to MM) are not recorded in the Forbes list:
• Two entries with a net worth of more than USD 10 billion (Dieter Schwarz with Lidl and Kaufland, Reimann family with Reckitt Benckiser and Coty)
• Six entries with a net worth between USD 5 and 10 billion (Oetker family with Oetker Group, Braun family with B. Braun Melsungen, Jacobs family with Barry Callebaut and Adecco, Knauf family with Knauf Gips, Peter Thiel with early investments in Paypal, Facebook, Zynga and Linkedin, Liebherr family with Liebherr Group)
• 32 entries with a net worth between USD 2 and 5 billion
• 69 entries with a net worth netween USD 1 and 2 billion
• Finally 47 names are recorded on both lists (Forbes and MM) with a total wealth of USD 233.8 billion according to MM and USD 239.2 billion according to Forbes. Despite total wealth for these entries matching nicely, for individual records the differences are more significant as shown in the graph below:

Obviously putting together billionaire lists is a difficult tasks especially when it comes to valuation of private companies. Nevertheless the overall differences are quite striking. When German billionaires are recorded at 156 instead of 55, the number of German billionaires exceeds those of China or Russia and its per capita number increases from 6.4 to 18.3 (exceeding the US level of 13.2) billionaires per 10 million population.

In a future post we will look the billionaires resident in Switzerland, where Bilanz has 131 entries (USDCHF exchange rate 0.95) compared to Forbes with only nine entries. Obviously one major source of differences for this list can be easily explained in terms of Forbes' nationality criteria versus Bilanz residency criteria.

90-Year Precipitation and Temperature Data for Santiago de Chile (Quinta Normal Measuring Station)

We have looked at the meteorological/climatological yearbooks (Anuarios Meteorologicos 1920-1996/Anuario Climatologico 1997-2010) available for the period of 1920 to 2010 at the website of the Chilean Metearological Services (Dirección Meteorological de Chile). The data is unfortunately only available in pdf format, so we had to get through each report and retrieve the data manually. The datapoints for 1930 and 1962 were omitted as for the former, the 1931 reports was posted and for the latter data was not available for the Quinta Normal measuring station (we pondered whether to replace with Cerillos measuring station for 1962).

The following shows the time series for precipitation and temperature:

 

The correlation matrix looks as follows. No surprises: temperature is correlated with AMO annd precipitacion with ENSO.

The temperature in Santiago (Quinta Normal) and AMO track each other quite closely. The increase in temperature was roughly 0.008 °C per year during the observation time.

Correlation between precipitacion and ENSO is not so strong, but it is still visible from the following graph that high precipitation years typically go together with negative SOI values and vice versa. 2012 was a neutral year in terms of the ENSO cycle and a below average year in terms of precipitation.

Major Foreign Holders of US Treasury Securities

The US treasury produces a handy list of major foreign holders of US securities which can be found here. As  of September 30, 2012 the total was USD 5.5 trillion up from USD 5.0 trillon on December 31, 2012.

The distribution between the major foreign holders is shown in the following graph (total USD 5.5 trillion):

The countries fall within one or more of the following categories:

• financial centers
• strong export nations
• participant in competitive quantitative easing and currency devaluation

When looking at the change for the first three quarters in 2012 against the absolute value the following picture emerges (Asian countries in green and Latin American countries in yellow).

The increase of Switzerland and Japan are notable while China and the Oil Exporters remained at a stable level.

Voter Turnout Chile 1870 to 2012

The recent municipal election in Chile, saw a dramatically reduced voter turnout. This was mostly triggered by a change from compulsory to voluntary voting.

The following post briefly addresses the history of Chilean voter turnout from 1870 to present (2012). The period from 1870 to 1973 has been covered by Patricio Navia in a paper in the Revista de Ciencia Política (Journal of Political Science) from which the following table is extracted. The data is actually extracted from two books, namely:

• Meller, Patricio. 1996. Un Siglo de Economía Política Chilena. 1980-1990. Santiago: Andrés Bello.
• Cruz-Coke, Eduardo. 1984. Historia electoral de Chile, 1925-1973. Santiago: Editorial Jurídica de Chile.

The period from 1988 to 2001 is covered in a second table basing itself on the INE (for Population data) and two government sites covering the current and previous (since 1989) elections.

We have completed the data from 2001 to 2012. We added the population data back to 1988 and the voting age population data back to 1992 using INE's population projection. For the voting age population we took all the 20+ age cohorts and 40% of the 15-19 age cohort (rounded to values of 50'000). Note that the 2012 population data point (17.402 million) is significantly higher than the census data (16.572 million, see our previous post on the subject) and the 2012 voting age population data point (12.750 million) is significantly lower than the electoral registry data (13.404 million).

The census data and the electoral registry are just horribly inconsistent, as shown is the following side by side table (note that the age cohorts for census data have been extrapolated from the total using the ratio of 0.7333 in terms of voting age population to total population as observed in the INE projection data, for the electoral registry the total population has been extrapolated using the same ratio). Given that some persons should appear in the census, but not in the electoral registry (foreigners with less than five years of residence, prisoners etc), the data of either the census or the electoral registry (or most likely both) seems indeed quite flawed.

The voter turnout (voters as a percentage of voting age population) can be graphically summarized as follows:

Copper, Gold and Silver Price Changes after Quantitative Easing Announcements

Following a post of ZeroHedge ("Spot the Odd One Out") we were curious to see how copper price reacted to the various QE announcements in the last couple of years. The complement the picture we also looked at gold and silver.

Price data we have downlaoded from Wikiposit (front contracts): Copper, Gold and Silver.

According to our count we have five FOMC announcement with QE characteristics:

1. November 25, 2008
2. March 18, 2009
3. November 3, 2010
4. September 13, 2012
5. December 12, 2012

In order to capture the entire price change due to the announcement we have looking at the closing prices at the day before the announcement (t-1) and after the announcement (t+1) and obtained the following results:

Data is here.