Talk:Rating Method Sovereign Wikirating Index (SWI)

From Wikirating, the transparent world rating organization.
Jump to navigation Jump to search

Feedback (from 18 November 2010)

  • The established rating agencies also have the problem of out dated facts and statistics for their reports and ratings. Best example: Ireland, which current rating AA- is based on the last review of Eurostat from 2009. The current rating does not reflect the reality. As "workaround", the rating agencies are working with "warnings", until the next review and publication of Eurostat statistics.
  • As a result, the current rating agencies HAVE to work with unofficial (not yet official) facts, data and statistics (on a quartely basis). The experts working for the rating agencies have access to such data, it is difficult to trace and to identify the sources, but IT HAS TO BE in order to publish updated warnings and ratings. According to the tester, the current methods used by the rating agencies might be not transparent enough, but he says that, like democracy, it is the "less worst" system.
  • The frontier between quantitative and qualitative criterias is almost impossible to define. The reason for that is that rating agencies need pieces of information that are not or late reviewed on a first hand by official institutions (Eurostat, IMF, World Bank, Central Banks reports), therefore they have to perform their own investigations and analysis, which are, according to the tester, very accurate in 95% of the cases (example of non accurate ratings/analysis: Island and Greece, where the agencies have not foreseen the bankrupt of Islandic banks and the dramatic debt of Greece).
  • Rating methods requires many facts and statistics. According to the tester, the SWI is too "light" for being taken seriously. The most missing ones from the SWI are:
    • private debt (houshold, companies, banks)
    • currency reserves
    • level of liquidity/public
    • private saving (households, companies, banks)

--Tester19 (talk) 18:22, 21 September 2011 (CET)

Ideas (from 4 December 2010)

  • Possible Criterias to be considered within the SWI:
  1. Central government foreign-currency debt
  2. Central government domestically issued debt
  3. Fiscal and current account deficit
  4. Excessive deficit procedure (EDP)
  5. Ageing population on public debt
  6. Volatility of government revenue
  7. Household and private-sector indebtedness
  8. Construction and real estate risks
  9. Foreign currency deposits
  10. Domestic deposits
  11. Wholesale funds

--Erwan (talk) 18:23, 21 September 2011 (CET)

Is there a room for a kind of LTSWI (Long Term Sovereign Wikirating Index) based on the ability of a state/corporation/product to be self sufficient in energy and human resources? And why not also considering in the rating the impact of the state/corporation/product on the environment, as the "carbon debt"(or equivalent) is becoming more and more critical in an investment choice for long term. In fact if we consume in a way the planet can absorb it, we can continue the business forever. Otherwise if we kill the planet, well, there's no business to make about it! And why not also considering in the rating the impact of the state/corporation/product on the individuals, as the "Gross National Happiness"(or equivalent) is becoming more and more critical as good investment choice for long term. In fact if people are happy while they go to work, well they work. Otherwise they go on strike, or suicide, and we waste some proportion of the investment!
--Markbamboul (talk) 17:44, 31 January 2012 (CET)
Hi Markamboul, You are absolutely correct. We are currently reviewing several non-financial factors that we want add in the credit rating models, such as social environment, environmental behavior, intellectual capital... Your "happiness indicator" would be as well a candidate. But we need quantifiable data and data sources to feed it into Wikirating. Do you know if these indicators have been quantified and if yes where we can find them? Thanks
--Erwan (talk) 07:02, 1 February 2012 (CET)
and there is also this paper from the World Bank, very interesting: Provided by a Wikirating user
--Erwan (talk) 07:05, 1 February 2012 (CET)
Actually such kind of index have been quantified...but for Buthan only as far as I know( )(and the participation rate is about 1% of the population), and that concept is hardly exportable as such a subjective concept as happiness is highly variable on the the culture. But still it seems some inspiration can be taken from it. Otherwise there are plently of other such quantifiable indices that can be found by surfing on wikipedia... Nevertheless if you don't find something reliable or transparent enough, then create it! That's why your website can be THE good place, first to define in an international/transcultural approach what "happiness" IS in its wildest and most general sense; then regarding these criterion (or this set of macroscopical variables that will araise from it) a questionnaire in the essence of the one used here ( could be proposed to people by polling, in the way you did previously for credit rating, from which an index on the satisfaction of each criterion can be extracted. (If the questionnaire contains some questions regarding subpopulation belonging (nationality(ies)/age/level of studies/...etc) that can be cross correlated with official "reliable" data about the population distribution, then the number of answers by subpopulation can be weigthed by their real proportion within the population until a statistically significant number of participants is reached for each subpopulation. Just like a regular survey!)
--Markbamboul (talk) 15:48, 1 February 2012 (CET)
Hi Erwan, to come back on some non-financial factors that could be included in the indices weights, there is a recent paper[1] depicting "an index to assess the health and benefits of the global ocean". Specifically, the tables S30/S31 (in supplementary materials[2]) summarizes the index according to different weightings for 172 different Exclusive Economic Zones (EEZ). All data are there; just need to collect them! (maybe after asking the authors ;)) BTW I think this is the right place to promote such research and raise public awareness and politics on the need to fund this kind of projects.
--Markbamboul (talk) 13:01, 21 August 2012 (CET)
I think using Foreign-exchange reserves and Total Market Capitalization would make the numbers more accurate. For example, you see Brazil: it has almost US$400 bi in Foreign-exchange reserves (the 6th largest among all countries, 3 times larges than USA's) and huge number is not being taken into account.
--EduardoAlba (talk) 18:38, 1 February 2012 (CET)
That is correct - do you have a reliable source where the foreign-exchange reserves of most countries are listed? The reason is to make this criteria comparable to preferable as many countries as possible. Than: How should this criteria be inserted in the SWI?
--Dorian (talk) 18:44, 1 February 2012 (CET)
Maybe CIA's World Factbook ( It covers 175 countries and I guess their data's source is IMF. A good criteria would be Reserves in % of GDP, and also how much is the countries reserve in % of all world's reserves.
--EduardoAlba (talk) 18:50, 1 February 2012 (CET)
Thank you! The CIA World Factbook is ok, but I am not sure, if they are REALLY reliable... of course, at the moment some data are used anyway for the SWI. How should the "Reserves of Foreign Exchange and Gold" be implemented in the SWI? My proposal is to reduce the "Public Debt" to maybe 40% and to add this new criteria with a weight of 10% - just an idea.
--Dorian (talk) 08:13, 2 February 2012 (CET)
Sure, a 10% weight for Reserves would work fine!
--EduardoAlba (talk) 13:29, 2 February 2012 (CET)
There is also another good indicator of risk: the Stock of Domestic Credit. A good criteria would be the Stock of Credit in % of GDP. CIA' source: (
--EduardoAlba (talk) 18:59, 1 February 2012 (CET)
Another indicator that could be used is a "Purchasing Power Parity" related. It would show some distortions caused by manipulated exchange rates...
--EduardoAlba (talk) 20:06, 1 February 2012 (CET)
Hi, the data used for the SWI are already based on purchasing power parity. Cheers
--Dorian (talk) 08:18, 2 February 2012 (CET)

Usefull articles for the SWI method (from December 2010)

  • Glossary:Excessive deficit procedure (EDP) [2]
  • Key Issues for analysing domestic debt sustainability [3]
  • A Positive Theory of Foreign Currency Debt [4]
  • Internal Rating Based approach[5]
  • Advanced Internal Rating Based approach [6]
  • Probability of default (PD) [7]
  • Loss given default (LGD) [8]
  • Exposure at default (EAD) [9]

--Erwan (talk) 18:24, 21 September 2011 (CET)

Ideas (from December 2010)

  • Mention, that SWI "describes" economic power quite well - also for Oman or Chile.
  • Mention that credit rating agency methods are different from SWI - at the moment there is no analysis research
  • Expand SWI in such way, that political risk is also considered as a criteria
  • Following criteria can be considered, too:
  1. Central government foreign-currency debt [1]
  2. Central government domestically issued debt [2]
  3. Fiscal and current account deficit
  4. Excessive deficit procedure (EDP) [3]
  5. Ageing population on public debt [4]
  6. Volatility of government revenue
  7. Household and private-sector indebtedness
  8. Construction and real estate risks
  9. Foreign currency deposits
  10. Domestic deposits
  11. Wholesale funds

--Dorian (talk) 18:27, 21 September 2011 (CET)

Explicite Debt vs Implicite Debt

Hello, the Wikirating Project is just something the world needs. One country judging the whole rest of the world, making continuosly heavy mistakes, favorizing one country only, herewith manipulating the whole World-Economy, ist really not the right thing. Moody, S&P, Fitch starting to rate even the European Union, make themself simply not trustworthy.

I sniffed a little through Wikirating and I miss something I believe is really important: to judge the future you neet to implement the IMPLICITE debts. As an example (in % of the BIP:

             Explicite Debt       Implicite Debt
Spain        65                   485
Italy        115                  35
France       85                   255
UK           100                  613     
USA          100                  660

These numbers are just approximative. What I wanted to show is the importance regarding the + and the - sign, read "economic future", in simple words: Who ist better off or deserves a AAA rating, Italy or USA ?? (Thats just looking at Ex- and implicite debts only, I am aware of) Further: do you implement election cabaret campaigns and "real" actions, such as in Greece, compared to almost "inactivity" in USA, to fight economic and financial problems. Thank you Siegfried Schwab Miami FL 33033 USA
--Siegfried Schwab (talk) 20:29, 12 December 2011 (CET)

Hi Sigfried, thank you very much for your comment - great input! I believe you are absolutely right about this, but where can be found reliably informations about implicite debts? Do you have a good source? Regards
--Dorian (talk) 20:36, 12 December 2011 (CET)"

Registered voters

I think it might be a lot better to include a named registry, where people can register and then record their votes. Then it is transparent when ratings were changed, one can see trends and one can also see how good or bad analysts might be. Just like with equity analysts it is said how many analysts cover a firm.

One then sees the validity of the rating !
--Christian Langkamp (talk) 20:17, 2 January 2012 (CET)

Hi Christian, your argument is absolutely correct. At the beginning, when Wikirating started 2 months ago, only registered users were able to vote. But unfortunately only very few people wanted to poll like this. So the polling possibility was opened to everybody. But we can change this of course again - back to the roots. A question will be presented asap to the community - and let you decide, whether to revoke the anonymous voting again.
--Dorian (talk) 00:17, 3 January 2012 (CET)

Public Debt overweighted

From my point of view, the public debt is overweighted with 50% in the SWI, even when multiplied with the HDI and other future looking factors. I believe, that the SWI should be more "cash flow"-driven which basically would show, how likely it is, that the country is able to generate the funds to pay the interest and pay back the debt. I think, that the GDP should be weighted more.

Additionally I would include a figure which shows the situation of a country within a larger complex (e.g. Euro) and on how the relation between potential company risks within a country could implement the entire financial system of a country or "region". In Island for example it was the problem, that the "economy", basically existing purely from turism and banks, was too large for the possibilities of the government and could not be saved by the existing funds, whereas in other European countries the single countries, e.g. like Germany, were able to provide securities for companies in trouble and even save or help to save other countries due to an overall strong economy. So the risk to loose an investment in Spanish Bonds is probably lower than loosing an investment in a fairly strong economy like Japan. If worse comes to worse, Japan would be at its own, whereas France would probably get help from other European countries.
--Swschertel (talk) 22:14, 2 January 2012 (CET)

Hi Swschertel, your arguments are reasonable. What do you propose for the public debt weighting? You can download the SWI Excel sheet and see (in "real time") what happens if one changes the weightings between the criteria (see tab "parameters") - the results of the ratings are displayed in the tab "results". Note: cells with blue text are dynamically used. Regarding the additional figure: do you have a good idea/concept how this could be included?
--Dorian (talk) 23:46, 2 January 2012 (CET)

Unemployment factors

I think using the U3 number, those that are currently collecting unemployment minimizes this segment of the calculation. There are many that want to work; but are working part time, are under employed, been replace with automation and do not have required skill set to qualify for new job requirements. As a result the U3 value is a partial input to the calculation. I suggest that the U6 value or some other calculation may prove to be more accurate. Additionally unemployment factors in the calculation should be consistent between nations so a valid comparison may be made.
--Anonymous user 18:01, 26 January 2012 (CET)

Subnational entity Ratings

Here goes an idea: What if we could create a Subnational entity Credit Ratings index?

This Index would aggregate the data of the administrative divisions (states, provinces, etc) of each country. For example: United States would have its own Sovereign Credit Rating index and also an individualized Credit Rating index for each one of the states.
--EduardoAlba (talk) 19:49, 1 February 2012 (CET)

This "takes the same line" like the discussion initiated at Talk:Belgium... I will open a new thread in the Forum.
--Dorian (talk) 08:22, 2 February 2012 (CET)
I have inserted a thread on the Forum for the new rating method Subnational Entity Wikirating Index (SEWI) - proposed by EduardoAlba.
--Dorian (talk) 15:27, 2 February 2012 (CET)
Ok. Thanks!
--EduardoAlba (talk) 18:37, 2 February 2012 (CET)

Sovereign Rating Distributions

Judging from the chart on the SWI page, it appears that the sovereign ratings follow something close to a normal distribution. This may make them less comparable to established agency ratings than you might wish.

In recent decades, ratings have become associated with default probabilities, although they could also be linked to expected loss (default probability * loss given default). Established rating agencies publish studies which show historic default rates by rating. Moody's and perhaps others have smoothed these rates out so that they rise monotonically as the rating gets worse. You can find these rates at page 9 of the following pdf: For example, rating Baa2 (which is like BBB) has a three year default probability of 0.38%.

Advanced economy sovereigns like Switzerland, Austria, Germany, France, the UK, US, Canada, Singapore and Australia have very low historic default rates. Emerging economies also have relatively low rates. Indeed, there have only been a couple of dozen defaults since 1998 - and defaults were even less frequent in the years before that.

Consequently, sovereign ratings should be concentrated between BBB and AAA if you want them to represent three year default probabilities that is compatible with industry norms.

Further, to improve transparency I suggest that Wikirating consider the following steps (perhaps not now, but when more volunteers are available):

  1. Publish a rating - default probability map so that everyone knows what the ratings are intended to mean
  2. On an annual basis, publish default rates by rating and see how they correspond to the map. You may or may not wish to adjust methodologies based on these findingsm since defaults tend to be cyclical and the low number of issuers in certain sectors (like sovereigns) causes a lot of statistical noise.

--Joffemd (talk) 02:56, 23 February 2012 (CET)

Hear, hear. For example: Australia is rated below AAA which indicates a level of risk unlike that of a nation with no defaults and an healthily serviced public sector debt. This indicates an issue with the model and illustrates the requirement for approximated default rate values as opposed to the floating indexed values. An example distribution can be seen in the bond rating default rates:
--Carter (talk) 14:56, 9 June 2012 (CET)

Equation, both simple and complex

If I have simply overlooked this please direct me to where it is located, otherwise, I think it would be appropriate to provide the equation of the model on this page as it is a little unclear as to exactly how it is calculated in regards to the secondary weighting via HDI etc. Something like the following for the simple version, and the 'full' equation should also be posted (including MAX and other Excel functions used) Score = (D*0.5 + A*0.2 + g*0.1 + i*0.1 + U*0.1)*(HDI*0.6 + C*0.2 + P*0.2) Where D = Public Debt, A = Account Balance, g = GDP Growth (%), i = Real Inflation Rate, U = Unemployment Rate, HDI = Human Development Index, C = Corruption Perception, P = Political Instability

Edit: Just did this.
--Carter (talk) 15:05, 9 June 2012 (CET)