Mapping heat for assessing vulnerabilities in a financial system

Mapping Heat for Assessing Vulnerabilities in a Financial System

In this post, I write about one of the seminal papers of Professor David Aikman who has a well-known expertise in the field of macro-prudential policy and financial stability. Specifically, I make a brief description of the methodology in Aikman and others (2017) with respect to the heat map for the U.S. financial system.

Justification

The aim of a heat map in macroeconomic is twofold. First, it visually characterizes the vulnerabilities of the aggregated financial system. Second, it identifies early warning indicators for macro-prudential policies. To this purpose, the authors group forty-six indicators of financial and balance-sheet conditions from the U.S economy in three categories: i. valuation pressures, ii. corporate borrowing; and, iii. financial-sector health. Some premises that the authors adopt along with the paper and the indicators used in the paper is reviewed in the following sections.

Premises

In contrast to the existing literature regarding early warning indicators, the authors focus on identifying the build-up of a set of vulnerabilities that may amplify shocks rather than predicting the timing of financial crises. Their work relies mainly on finding indicators that have both reliable information content and timely warning signals of financial stability. 

A heat map also complements the literature that measures adverse spillovers or interconnectedness across financial institutions from balance-sheet and financial market information. For instance, the Conditional Capital Shortfall Measure of Systemic Risk – SRISK measure is included as input along with summary statistic coming from Granger-casuality among financial series.

Categorizing and aggregating indicators

A common approach to condense the information would be using principal component analysis. However, this method poses drawbacks in terms of the easiness of explaining the motivation for the weighted components in a policy context. Moreover, consistency with prior views regarding the importance of the economic mechanism may not be ensured.

Considering these limitations, data of overall vulnerability is grouped into three broad categories. These encompass valuation pressures and risk appetite, nonfinancial sector imbalances, and financial sector vulnerability. The authors also divide them into components to avoid over-weighting data which are easily available. Valuation pressures, for instance, include components such as housing, commercial state, business debt, equity markets, and price volatility.

After grouping the array of forty-six indicators into fourteen components, the authors define a component index. This is built as an average of the standardized indicators for each component. For example, the housing component index includes three indicators in its calculation: price-to-rent ratio (relative to a 10-year moving average 7), the change in lending standards from the Senior Loan Officer Opinion Survey (SLOOS), and the Fair Isaac Corporation (FICO) scores for mortgages sold to Government- Sponsored Enterprises (GSEs). The purpose is then to examine the degree to which each component captures the different aspects of vulnerability in financial stability. 

In the next post, it will be discussed the algorithm and some statistical issues for obtaining probability density functions for component and aggregate indexes.

References

Aikman, David & Kiley, Michael & Lee, Seung Jung & Palumbo, Michael G. & Warusawitharana, Missaka, 2017. “Mapping heat in the U.S. financial system,” Journal of Banking & Finance, Elsevier, vol. 81(C), pages 36-64.