.webp)
When working with statistics, problems may arise when the collected data is distorted during processing, misinterpreted (link in Russian) or even simply does not stay in the memory (link in Russian), pushed out by a colorful story. However, in all these cases there are no issues with the data itself, that is, what should be measured and how. But what if something more intangible needs describing? Monetary conditions are one of intuitively obvious, but difficult to measure, concepts.
It is clear that the financial sector has an impact on the real sector and that high interest rates on loans correspond to monetary tightening, low rates to monetary easing. But as soon as it comes to specifics, discrepancies appear.
The impact of the financial sector on what exactly are we talking about? The impact on the access to credit (which is most often meant by businessmen talking about monetary conditions) or the impact on the overall demand in the economy (which macroeconomists mean)? The distinction is fundamental - for example, a budget deficit may significantly increase overall demand and money supply, but it can also adversely affect the access to credit.
A no less important question is the impact of what on the real sector we are taking about. The entire financial sector or only that part of it which is affected by the actions of the central bank (the latter is meant when monetary conditions are considered as a feature of monetary policy)?
Any of the answers to these questions has a right to exist, and for each of the approaches to answering these questions there are conditions in which precisely this approach would be required. Examples of different approaches, including those used by the Bank of Russia, are presented in the recently released analytical note (link in Russian) ‘Indicators of tightness of monetary conditions’. It is not a ‘route description’, but a ’guide’ to the statistics of monetary conditions, and it does not impose a specific set of indicators on the reader, but rather offers recommendations applicable for running any set.
Such different conditions (and their indicators)
Tight (or loose) monetary conditions are terms used in modern international practice almost as widely as ‘user friendliness’ or ‘favorable climate conditions’. In the majority of texts these terms are used as self-evident, with no definition provided. However, when it is discussed whether monetary conditions are in fact tight, loose, or neutral, it becomes clear that the definitions that various authors put in this concept can differ significantly. Continuing the analogy - the majority will agree that the climatic conditions in Anadyr are tougher than in Kaluga, but it is much more difficult to make a judgement about the ratio of the severity of climatic conditions in Kursk, Voronezh and Saratov, with which all the residents of these cities would agree.
The variety of meanings that are attached to the concept of monetary conditions determines the abundance of statistical indicators that can be used to assess their tightness and looseness. Unfortunately, none of these indicators can confidently characterise monetary conditions, and each has its own pitfalls.
Interest rates? Yes, they characterise the affordability of credit, but, firstly, rates characterise only affordability assessed by its price. The cautious lending policies of banks, in which they carefully select borrowers often limit the access to loans more significantly than high interest rates. And, secondly, average market rates are subject to changes due to fluctuations in the structure of the credit market (which can lead to the Simpson’s paradox (link in Russian) when interpreting such data).
Volume of loans issued? Yes, a sharp increase in lending to the economy can serve as a sign of loose monetary conditions. But this indicator is one-sided: weak lending activity does not necessarily mean that the monetary conditions are tight. There are situations when, despite attractive credit conditions, the demand for loans remains low (for example, after an economic crisis, when uncertainty about the future overcomes the effect of loose monetary policy).
This list could be expanded, but almost all indicators have obvious or hidden flaws. Relying on one indicator means exposure to potential error or inaccuracy. Therefore, in international practice the main solution was the use of an array of multiple heterogeneous indicators to characterise monetary conditions. The principle ‘together people do not go crazy, together they only catch the flu’ is applicable not only to people, but also to statistical indicators - and an error in one indicator when the rest are correct will be insignificant.
How to measure monetary conditions (and how not to)
There are two main groups of indicators commonly used to characterise monetary conditions: 1) price, 2) volume.
- Price indicators of financial markets (interest rates, bond yields, exchange rates, share prices and credit derivatives, real estate prices) and their derivatives (spreads, volatility indicators, etc.) characterise the affordability of financial resources, the mood of creditors and investors, the risks associated with borrowing.
The strength of this group of indicators is that they objectively and directly characterise the terms of transactions and that these terms are consistent both for different periods of time and for different countries and regions.
The weak point is that price conditions are only part of the deal. There are also non-price conditions. Moreover, a change in non-price conditions may distort the assessment of the price conditions. For example, if banks relax lending terms, then formerly reliable borrowers can obtain loans at reduced rates, and those who were considered less reliable can also get credit, but at higher rates. As a result, the average market interest rate rises, which is interpreted as a deterioration in the access to funding, although in fact the exact opposite happened (one of the manifestations of the above mentioned Simpson’s paradox).
- Volumetric indicators of financial markets characterise the growth of credit and bond portfolios, money supply, the turnover of the primary market of shares and bonds, showing the amount of funds flowing into the economy through various channels.
The strength of this set of indicators is that they reflect the influence of both price and non-price conditions, that is, all the factors that affect the flow of funds into the economy.
The shortcomings of the indicators in this group are an extension of their merits. They actually characterise all the factors that affect the flow of funds from the financial sector to the real one, including factors that have nothing to do with the situation in the financial sector. For example, the demand for credit may decrease due to deterioration in international market conditions or expectations of higher tax rates even with an easing of monetary conditions.
The diversity of indicators for assessing monetary conditions is due to the fact that none of them can correctly capture all aspects of the impact of the financial sector on the real one. The analysis of the problems generated by the pitfalls of financial indicators contained in the note may be interesting not only in the context of analysis monetary conditions. The same problems can be faced by an analyst who studies the situation in the credit market, and a sociologist who analyses saving behaviour among the population.
The most obvious source of inaccuracies in financial statistics is a change in market structure. It is most noticeable for markets where there are two or three large segments, the rates of which differ significantly (mortgage and consumer loans, loans to small and large businesses, preferential and non-preferential loans) with significantly varying interest rates. In such markets, a change in the share of one of the segments can have a greater impact on the average market interest rate than a change in the rates themselves. Therefore, it makes sense to look not only at average market interest rates, but also at the rates of individual market segments.
Another pitfall of financial statistics is related to changes in the structure of operations. Along with the largest segments of the domestic financial market, attracting the most attention of researchers, there are other smaller segments, such as the bond market, which is an alternative to the credit market for enterprises and banks, and the deposit market for households. In Russia, this market is not so large - its volume is multiple times lower than the volume of the credit or deposit markets, and researchers often focus on the latter two. However, in times of instability, banks begin to prefer more liquid (albeit less profitable) bonds over loans, and an analyst who do not take into account the flow between market segments may overestimate the decline in lending activity.
Among non-experts, nominal rates or gains are more common, and this approach is quite applicable to the analysis of short-term (and to a certain extent medium-term) dynamics. Ignoring the changes in inflation is redeemed by the clarity of the indicators. However, in long-term analysis (and especially in very long-term (link in Russian)) or in cross-country comparisons, differences in inflation that can significantly shift the level of nominal rates cannot be ignored. Obviously, in summer of 2015 when inflation in Russia hit 15%, the loan rate of 10% per annum was very attractive, but two years later, when inflation dropped to 3–4%, the same rate became significantly less attractive. Therefore, in a long-term comparison of monetary conditions, real interest rates, or inflation-adjusted rates, are always used.
Moreover, it is necessary to adjust for inflation not only interest rates, but also volumetric indicators - even medium inflation can turn the price of a new apartment into the cost of a used car in just a decade. Therefore, an important element of long-term analysis is the assessment of real interest rates and real lending activity.
Another, almost philosophical, question is related to the assessment of real interest rates. What is more important: the past - firm, objective, but already over, or the future, which, while present in vague expectations, affects today’s decisions? Which inflation is relevant for the analysis of interest rates: expected (and influencing decisions); or actual (and influencing the results of decisions already made)? When raising a loan or placing a deposit, the client of a bank does not yet know what inflation will be in the future - they decide to make a transaction, guided by their expectations of future prices. An ex-ante approach is therefore used to reflect the impact of the financial sector on the real sector, adjusting interest rates to market participants’ expectations. A similar situation arises with the analysis of price volatility - indicators of historical and expected volatility indicators characterise different facets of the economy.
The dream of the ‘one-handed economist’ (and its statistical implementation)
The most popular method of using several complementary indicators of monetary conditions is the formation of a consolidated indicators based on them (MCI or FCI – the Monetary/Financial Conditions Index). The popularity of MCIs and FCIs is associated with their simplicity and clarity, which allows you not to think about the variety of processes taking place in the financial sector, and to give a simple and clear answer to the complex question about the current state of monetary conditions – are they loose, tight, or neutral.
According to popular legend, one of the presidents of the United States (in different versions of the story Truman, Eisenhower, and Roosevelt appear) tired of the tendency of his economic advisers to answer any question ‘on the one hand/on the other hand’, exclaimed: ‘Give me a one-handed economist!’ To some extent, MCIs and FCIs indices represent the realization of the dream of a ‘one-handed economist’ by reducing the complexity of analysis required.
In cases where simplicity of perception is prioritized over completeness and accuracy, these indices can be a convenient and efficient analytical tool. However, the flip side of such simplicity is the approximation and over-simplification of true economic reality. The average annual temperature in the the Gobi Desert is the same as in Yekaterinburg, but hardly anyone will agree with the conclusion that the climatic conditions in these regions are equally harsh (or equally mild). Users of monetary conditions indices face similar problems in interpreting the results.
An alternative to these indices is the use of sets (dashboards) of complementary indicators that characterise different aspects of monetary conditions (see box). This approach is most widely used in the analysis of financial stability (for example, by the IMF, as well as by the Bank of Russia) and the European Central Bank applies it to its analysis of the tightness of monetary conditions.
The use of dashboards makes it possible to provide a more in-depth and systematic analysis of the situation in the financial industry, taking into account the uneven development of its individual segments. The flip side of this tool is high complexity, requiring a lot of labour-intensive analysis, and exceptional analytical skill. If, when using an MCI or FCI indices, the estimates of monetary conditions depend on the subjectivity of the index developer (a notable example of which can be the study of various indices of monetary conditions for the UK economy), then, when using a set of indicators, there is a risk of subjectivity from the analyst operating on this set: the same values of indicators can lead to different conclusions.
In the practice of the Monetary Policy Department of the Bank of Russia, both indices and dashboards of indicators are used to analyse the tightness of monetary conditions, but the potential for applying MCIs and FCIs is limited. Over recent decades, the Russian economy has repeatedly faced structural shifts in both the real and financial sectors, which limits the reliability of the weighting factors used in the indices. The Monetary Policy Department of the Bank of Russia has developed a methodology for calculating the gap in monetary conditions, which is integrated into its quarterly forecast model: at a conceptual level, this methodology is close to the traditional MCI indices, but the use of the mathematical apparatus of the quarterly forecast model allows it to take into account structural changes in the economy.
As for the indicator dashboards (some of them are presented in the Bank of Russia’s monthly analytical commentary ‘Monetary Conditions and Monetary Policy Transmission Mechanism’), then, depending on the purpose of the economic analysis (assessment of monetary risks of inflation, factors of formation of aggregate demand, risks of overheating in the financial sector, etc.), various indicators and methods of their assessment can be used (for example, the ECB uses in its practice a large array of indicators).
Most primary statistical indicators have their own specifics, and the interpretation of statistics is always subjective to some extent. And analysts (whether they are researching monetary conditions or other aspects of the economy) have to analyse not so much statistics as the economic processes behind the changes in the statistical indicators. Otherwise, they risk adding themselves to the pool of authors whose calibrated statistics lose out to the less rigorous, but convincing narratives mentioned at the beginning of this article.
