Slovakia’s Rating Is the Worst in Decades

I have retained the headline exactly as it appeared in the quotation.

“Kamenický Should Not Be So Pleased with the Agencies’ Report Cards

Finance Minister Ladislav Kamenický never misses an opportunity to boast about good news. On Friday, the DBRS agency left Slovakia’s rating unchanged. Kamenický described this as a ‘positive report card’ for the Slovak economy. Yet when we look at the chart prepared by my colleague Martina Kláseková, the picture is less cheerful. Since Robert Fico’s fourth government took office, all the major rating agencies have downgraded Slovakia. In the case of the ‘Big Three’ – Moody’s, Fitch and S&P – Slovakia currently has its worst ratings in many years.

Slovakia’s rating is the worst in decades. Slovakia’s long-term rating according to four agencies (the higher the position on the chart, the better the assessment; 1 = AAA/Aaa). The dashed lines indicate the dates on which governments took office.”*

I did not know that agencies issue report cards.

Why should citizens believe such unscientific claims? What does “higher on the chart” mean? What exactly is a “positive report card”? It is vague.

This is at a time when science has long since developed scientific methods for evaluating economic performance, such as the Mahalanobis distance, the Ivanović distance – devised by a world-class statistician – or the work of R. A. Fisher, who as early as 1936 published The Use of Multiple Measurements in Taxonomic Problems.

Measuring performance by means of some scale such as 1 = AAA/Aaa is inadequate, superficial and an insult to science. Even Geoff Dougherty’s well-known book Pattern Recognition and Classification: An Introduction deals precisely with the question of how to classify.

Does the Big Three know how to do that?

Who exactly are the “Big Three”? And what is the statistical basis of their classification?

Do they know?

I do not know why the media keep talking about “ratings”. Surely even Zuzka from Maríková understands that. A rating simply means an assessment.

Why should the finance minister be delighted by some scientifically unsupported taxonomic system?

Ivanović’s book Diskriminaciona analiza has 210 pages. I studied under him, passed his examination and own the book.

China has produced female humanoid robots, humanity is travelling into space, yet the Big Three still has to rely on letters.

Surely this is pseudoscience.

Slovakia supposedly has its worst ratings ever!

But these ratings are unscientific.

Discriminant analysis deals with the problem of dividing statistical populations – countries, districts and other units – as well as with determining the membership of a particular observation in a given statistical group. Above all, however, it deals with the classification of statistical populations and groups according to one or several characteristics.

Today, it is used in solving problems of regional development, including the classification of regions according to their level of economic development.

The Big Three does not do this.

Of exceptional importance is also the study of the dynamics of development, as well as questions concerning the allocation of investment among sectors and regions with the objective of maximising the degree of economic development and efficiency.

We can therefore see that many economic problems cannot be addressed precisely without employing the methods of discriminant analysis.

Let us look at this in concrete terms.

The Ivanović distance, or I-distance, is defined as follows:

or, in expanded form:

Does the Big Three employ such a relationship, or anything comparable to it?

For a specific illustration, let us consider a numerical example.

Suppose we have four variables and six districts – six regional units. Let us assume that we know their interrelationships, variability and perhaps even their asymmetry.

The standard deviations are:

σ₁ = 5, σ₂ = 5, σ₃ = 10 and σ₄ = 20.

Let all variables be mutually connected by stochastic relationships as follows:

r₁₂ = 0.5, r₁₃ = 0.5, r₂₃ = 0.5, r₁₄ = 0.5, r₂₄ = 0.5 and r₃₄ = 0.5.

Let the reader try to construct the ranking.

It probably did not work.

Let us suppose that the variables can simply be added across each row and that we disregard asymmetry. First, we construct the ranking solely on the basis of the sum of the values of the variables, shown in the final column.

We obtain the following order:

(1) District II: 450
(2) District I: 440
(3) District III: 410
(4) District IV: 400
(5–6) District V: 370
(5–6) District VI: 370

There is no difference between the fifth and sixth districts.

Surely the reader can sense how simplistic this solution is.

Now let us apply the I-distance.

The hypothetical weakest district has the following values for its variables – the minimum observed values:

80, 90, 80 and 80.

Their total is 330, as shown in the final row of the table.

Let us calculate and compare the economic strength of the districts using the I-distance:

We have obtained numerical values for the degree of economic development of the districts, denoted by D.

Pay particular attention to Districts V and VI. We can now clearly see both the difference between them and the reason for the difference in their level of economic development.

Looking merely at the raw numerical values of the variables, we cannot see this.

If we rank the districts according to the size of the I-distance, we obtain the following order:

(1) District I, D = 13.375
(2) District II, D = 10.000
(3) District III, D = 8.750
(4) District IV, D = 5.625
(5) District V, D = 2.653
(6) District VI, D = 0.500

We can see a substantial difference between the fifth and sixth districts.

From the assessments produced by the Big Three, we would probably not see it.

Subjectivity?

We do not see the causal relationships among the determining elements of the concept being assessed.

Let X1 represent kilometres of motorways per 100 square kilometres, and let X2 represent kilometres of railway lines per 100 square kilometres. There will certainly be a correlation between them, even though these two modes of transport may be both competitors and complements.

If we changed the numerical value of X2 for District VI by +10, we could easily determine how its position changed.

If X3 represents agricultural production and X4 food-industry production, then there will also be a correlation between them – and surely a different one from the relationship between X1 and X2.

Can this be captured by the grades 1 = AAA/Aaa?

And does the Big Three use statistical characteristics such as the Ivanović distance?

It is precisely the dependence among economic indicators that is crucial – their correlation coefficients rᵢⱼ, as well as their σᵢ values.

Conclusion

Classification is the process of assigning objects, phenomena or concepts to groups or categories according to their shared properties or characteristics.

This intellectual and organisational tool helps economic policymakers to understand the complex economic world more effectively, to identify and assess the state of the economy objectively, and to facilitate access to the information required for economic policymaking.

Classification is a demanding problem spanning several scientific disciplines.

In the quotation cited above, perhaps someone had an interest in discrediting the government.

As I have demonstrated with the numerical example, discriminant analysis deals with the problems of dividing multidimensional statistical populations – countries, districts and other units – as well as determining the membership of a particular observation in a given statistical population.

Above all, it deals with the classification of statistical populations and groups according to one or several characteristics.

The variables must be ordered according to their importance – building a motorway is more important than the size of a mayor’s apartment.

Slovak economists have used 72 indicators to classify Slovakia’s districts.

Given the enormous development of econometrics and statistical theory, classification should today be a standard scientific instrument employed by all governing bodies – including the European Union.

Prof. J. Husár

Rohovce, 18 September 2026