This group of factors, which reflect intended risk weight variability, explains 56% of risk weight variance.

After controlling for risk, we investigate whether different implementation standards and supervisory practices (unintended risk weight heterogeneity) also matters. For this function, we add dummies for every country where the bank granting the loan is headquartered. Within an ideal world with equal implementation standards, these HQ fixed effects will be statistically zero.

When we do that, the explained variance rises and then 57%. However the HQ effects are essential. 3

These statistically significant and economically important effects (consistent with Turk-Ariss 2017 and EBA 2017a) are large in Denmark, Sweden, and Italy, where banks have low risk weights. The contrary holds true in Ireland, UK, Portugal and Austria. For instance, the expected IRB risk weight of a bank headquartered in Italy is 18.3 percentage points less than the IRB risk weight of a Portuguese bank with the same destination, asset class, and macroeconomic environment. Remember that a few of these identified countries have previously taken macroprudential measures targeting bank risk weights (ESRB 2017).

Figure 2 depicts risk weights regarding the same asset class and the same destination (Germany) of most banks headquartered in selected countries. Germany was chosen as a destination just because a large numbers of banks from different HQ countries actively grant loans to Germany, which enables us to draw this comparison. The HQ countries were chosen predicated on their HQ regression dummy coefficient. While Figure 2 illustrates the heterogeneity across HQ countries, the results in a regression we can identify HQ effects after controlling for portfolio-specific NPL ratios.

** Figure 2 ** IRB risk weights for corporate exposures (left) and retail exposures secured by property (right)

Source: EBA transparency exercise and author calculations.

## The consequences of changing HQ countries on a bank’s capital ratios

How would capital ratios change if we change only the united states in which a bank is headquartered and keep every thing else equal? Because of this hypothetical exercise, we choose the largest banks from each country and utilize the estimation results obtained in the last section to calculate capital ratios.

Clearly, these calculations are hypothetical in a number of respects, and are not really a ‘what if’ bank A moved its HQ to country B, because risk models that already are improved are unlikely to improve abruptly. Table 1 shows the hypothetical capital ratios for the biggest banks in every countries. The primary diagonal represents actual capital ratios as at June 2016. The off-diagonal elements often deviate from the actual ratios. The consequences are economically large.4

** Table 1 ** Hypothetical CET1 ratios of selected banks

## Will there be unintended risk weight heterogeneity?

We also look at other styles of unintended risk weight heterogeneity, but find no evidence that large banks are better in a position to outmanoeuvre supervisors by increasing the complexity of their models. We find marginal statistical evidence (p ** Topics: ** Financial markets Financial regulation and banking