Although European banks have largely profited from the rapid rise in interest rates, higher interest rates also caused a decline in the fair value of some of the financial assets that they hold on their balance sheets. In a few cases, this may represent a source of latent risk. Overall, S&P Global Ratings doesn't expect unrealized losses to represent a major risk (and therefore a major ratings driver) for European banks, but it will monitor any bank it considers vulnerable to a sudden drop in the fair value of their securities.

The vast majority of fair-valued financial assets are held for trading and any fair value gains or losses are recognized directly, in the income statement. While there is potential for unrealized losses or gains to build up where fair-valued financial assets are not subject to this treatment, this only applies to about 4% of the financial assets held by banks in the EU and European Economic Area (EEA).

Furthermore, in Europe, both large and small banks routinely reflect unrealized losses that affect their debt securities in their regulatory capital ratios, thus making them visible. In the first nine months of 2022, some securities saw a significant loss of value. The resulting increase in unrealized losses created a drag of around 40 basis points (bps) in the regulatory capital ratios of European banks. While this was not a negligible amount, it was largely offset by other sources of capital accretion and presented no impediment to the very solid capitalization levels at European banks.

As for debt securities reported at amortized cost, we view as remote any scenario in which their fair value losses (or gains) are realized, as they are generally held to maturity rather than sold.

We are also mindful of the possibility that banks could see an indirect, but still negative, effect as the fair value of securities falls. Indeed, this could cause counterparty risk to materialize. Many nonbank financial intermediaries (NBFIs) are significant holders of securities; therefore, if fair value losses were to accumulate and be suddenly crystallized, they could face heavy losses and even liquidity issues.

Frequently Asked Questions

How do European banks account for their financial assets, and how do changes in the fair value of those assets affect their income and equity?

Most rated European banking groups report their consolidated financial information under International Financial Reporting Standards (IFRS), although Credit Suisse reports under U.S. generally accepted accounting principles (GAAP).

Under IFRS 9, banks use two tests to classify their financial assets:

• The business model test considers the business model the bank uses to manage the asset; and
• The contractual cash flows test considers the form of the asset's contractual cash flows, and specifically whether they are solely payments of principal and interest (SPPI).

Chart 1 shows how these two tests are applied to financial assets that are debt securities--a category that includes customer loans, government bonds, corporate bonds, collateralized debt obligations, and certificates of deposit. This defines the relevant accounting classification and how fair value gains and losses are treated.

Under IFRS 9, after the initial accounting classification, a bank may reclassify debt securities only when it changes its business model for managing those securities, and may not apply the reclassification selectively; it would have to reclassify all affected securities. The standard also emphasizes that such changes are expected to be very infrequent. Thus, European banks operating under IFRS have limited ability to reclassify debt securities, in contrast with U.S. banks operating under U.S. GAAP, which views reclassification as more permissible. That said, under IFRS, if FVOCI debt securities are reclassified to amortized cost, any unrealized fair value losses in the OCI would be immediately reversed and adjusted against the new amortized value of the securities. This has benefits for accounting equity and regulatory capital. Under U.S. GAAP, the unrealized loss would instead remain part of the accumulated OCI and would be amortized over the life of the securities.

In practice, European banks willing to rebalance their portfolios could choose to reinvest the proceeds of maturing debt securities previously accounted at fair-valued into bonds held at amortized costs, a trend that has become more prevalent in the current rising rate environment.

What type of financial asset is most commonly held by European banks? How has this evolved over recent years?

In the EU and EEA, most financial assets are classified at amortized cost--as of September 2022, such assets accounted for 78% total financial assets, although there was some variation by country (see chart 2). The remaining 22% were reported at fair value, either through profit and loss (FVTPL; 18% of total financial assets) or through other comprehensive income (FVOCI; 4%). In absolute terms, FVOCI financial assets amounted to €1.1 trillion as of June 2022, of which 57% were government bonds. FVOCI financial assets also include derivatives held for cash flow hedging purposes.

Chart 2

Given that only the 4% of financial assets are held at FVOCI and therefore can generate unrealized losses (or gains), the effect of rising interest rates leading to potential fair value losses for such assets is not generally a material credit consideration for European banks. A few small jurisdictions--such as Malta, Slovenia, Romania, or Poland--have a higher share of FVOCI financial assets, at 10%-15%, but we do not regard this as material.

The relative share of fair-valued financial assets held by EU/EAA banks dropped to 22% in June 2022 from 24% in December 2019. During this period, central bank refinancing operations had made liquidity abundant. We conclude that European banks preferred to park their excess liquidity with central banks, rather than invest in securities.

How are unrealized gains and losses on financial assets treated in European banks' regulatory capital ratios?

In Europe, fair value changes on FVOCI debt securities are incorporated in the regulatory CET1 capital for all banks, via the OCI accounting entry. This differs from the U.S., where regulators only hold the largest banks to this higher standard (see "Credit FAQ: How Unrealized Losses On Securities Affect U.S. Bank Ratings," published on Dec. 13, 2022).

That said, at the start of the COVID-19 pandemic, EU legislators introduced a temporary quick-fix to capital rules that created an exception to the standard. The prudential filter, which expired in December 2022, gave banks the option of temporarily excluding fair value gains and losses on government bond holdings from other comprehensive income (and therefore from regulatory capital). The aim was to prevent any increase in volatility on government bond prices from creating potential volatility in regulatory capital ratios. A review of disclosures from several major European banks suggests that they did not make use of the filter.

Further, EU capital rules include several other prudential filters, which are permanently available to banks and also aimed at smoothing out the impact of selected market movements on regulatory capital. For instance, banks are allowed to neutralize in their regulatory capital the unrealized gains or losses from cash flow hedges or from their own liabilities, accounted at fair value.

Did the decrease in OCI significantly depress European banks' shareholder equity or regulatory capital in 2022? What is S&P Global Ratings' expectation for 2023?

Although decreases in OCI created a nonnegligible drag on banks' regulatory capital ratios in 2022, they were not the major reason why we saw a relative decline in these ratios. Going forward, we don't expect unrealized losses to represent a major risk to European banks' regulatory capital levels.

Regulatory capital ratios tightened over the first nine months of 2022: the average CET 1 ratio for EU/EEA banks dropped to 15.0% in September 2022 from 15.8% in December 2021. As chart 3 shows, the fall in CET1 ratios affected most European banking systems. However, a closer look at the drivers underlying this trend shows that the decrease in OCI played only a small part in the shift (see chart 4).

Chart 3

Chart 4

Overall, the key reason why CET1 ratios decreased was inflation in risk-weighted assets over the period, largely driven by high loan growth. We estimate that OCI decreased by about €32 billion over the first nine months of 2022, which represents a drag of about 37 bps on CET1 ratios. We consider this decrease in OCI to be relatively small, given that many financial assets saw a significant negative repricing; to some extent, this may reflect the effectiveness of the interest rate hedging strategies at some banks. Importantly, the decrease in OCI was more than offset by the positive effect of increased retained earnings during the period, which itself stemmed from improved profitability at European banks and their relative prudence with regard to shareholder distributions.

Although it is hard to predict future OCI movements, because they depend on market valuations, we anticipate that the effect of volatility in reported OCIs on capitalization levels at EU/EAA banks is likely to remain relatively minor. Furthermore, market and book values should steadily converge over time where banks continue to hold assets; this will gradually offset the impact on capital of specific instruments. We consider that European banks remain very well capitalized to absorb this volatility, based on an average CET1 ratio of 15.0%.

Beyond system-level aggregates, are there significant differences across European jurisdictions and/or banks?

We did see some outliers in the first six months of 2022, at both the system and bank level.

At the system level, the main outlier was Poland, where the decrease in OCI, before taking into account prudential filters, reduced banks' average CET1 ratio by 230 bps (see chart 5). We also saw a higher-than-average drag in certain relatively small EU jurisdictions, such as Slovenia, Greece, Portugal, Malta, and Latvia. At the bank level, most banks were clustered close to the average, but for a handful of banks, negative OCI movements (again, before the application of prudential filters) diminished their CET1 ratios by more than 200 bps in the first six months of 2022 (see chart 6).

Chart 5

Chart 6

Based on discussions with several banks, we understand that negative fair value movements on structural cash flow hedges accounted for a significant share of these OCI decreases. For instance, for banks carrying a high proportion of floating-rate loans and hedging against a potential decline in interest rates, these cash flow hedges generated an accounting loss in shareholders' equity in the first half of 2022, due to rising interest rates. However, this loss was offset by the increase in the intrinsic value of the hedged portfolios, and European regulators recognize this hedging relationship by entirely and permanently filtering out unrealized gains and losses from cash flow hedges from regulatory capital.

Beyond cash flow hedges, the higher-than-average impact on certain banks likely resulted from their exposure to securities that reported significant fair value losses in the period. We saw this in action in Poland, where banks have a fiscal incentive to hold government bonds. The effect was likely exacerbated by Polish banks' tendency to hold a higher proportion of FVOCI securities--systemwide, these represent 13% of all financial assets, compared with the EU/EEA average of 4%.

To sum up, some banks have shown more volatility in OCI last year, and these deserve a closer monitoring to understand the main underlying drivers of OCI volatility, and the related impact of prudential filters on the final reported regulatory capital. However, we take comfort from the fact that most banks with very high OCI decreases last year still reported elevated capital ratios (see chart 7).

Chart 7

Although U.K. banks are not included in the data sets disclosed by the European Banking Authority, we see a similar story with regard to unrealized losses in OCI. That is, fair value losses on bonds have not materially affected U.K. banks' regulatory capital ratios because:

• Their treasury portfolios are dominated by central bank reserves, and therefore held at amortized cost;
• Asset swaps mitigate directional risk on bonds; and
• Their structural hedges mostly use derivatives, and they also benefit from prudential filters against unrealized gains and losses on cash flow hedges.

Internationally active banks, particularly HSBC, were more affected, but are increasingly mitigating fair value risk.

How does S&P Global Ratings treat unrealized gains and losses on financial assets in its analysis of bank capital adequacy?

In our primary measure of a bank's capital, our risk-adjusted capital (RAC) ratio, we neutralize the impact on shareholders' equity of unrealized losses and gains from debt securities and cash flow hedges. That is, we do not count such unrealized gains or losses as part of our total adjusted capital, which is the numerator of the RAC ratio. By contrast, we do not neutralize the impact of unrealized losses and gains from equity securities.

In our view, if total adjusted capital included unrealized gains and losses, the RAC ratio would become more volatile, which could distort the true picture of a bank's capital adequacy if we do not expect the bank to realize those gains and losses. Nevertheless, we still consider the magnitude of any unrealized losses or gains, and the probability that losses could be realized. If we thought that the RAC ratio was not an accurate reflection of the strength or weakness of a bank's capital, we could adjust our capital assessment of the bank; we consider unrealized losses as part of that analysis. Specifically, our criteria allow for an adjustment to the capital assessment after consideration of the relative strength or weakness demonstrated by other capital metrics.

How does S&P Global Ratings view a drop in the fair value of amortized cost securities?

In our opinion, European banks are unlikely to realize market value losses on amortized cost securities because such losses do not reduce a bank's shareholder equity and regulatory capital ratios, and IFRS restricts the reclassification of financial assets. In practice, it would take an acute liquidity crisis to force the materialization of unrealized losses on amortized cost securities. Even then, we expect that banks would first attempt to pledge the assets to obtain cash, and only sell these assets (and therefore realize the market value loss) as a last resort. As chart 8 shows, European banks have elevated liquidity buffers and therefore this is not presently a noteworthy source of risk.

Chart 8

Beyond any direct impact on their equity, could the drop in the fair value of FVOCI financial assets have a broader negative impact on European banks' creditworthiness?

Some banks could see an impact through their exposure to counterparties that have themselves suffered from the decline in the fair value of financial assets. For example, nonbank financial institutions (NBFIs) often hold significant portfolios of debt securities and derivatives and some operate with elevated leverage. These NBFIs would be hit by fair value declines, for example, through additional margin calls that may strain their liquidity. In the extreme, a disorderly repricing could trigger broader market dislocations. In the absence of public intervention, this could create turmoil in the financial system, spelling trouble for banks and potentially causing counterparty default risk to materialize.

An example of this kind of situation occurred at the end of September 2022, when U.K. pension funds were engulfed by a crisis triggered by their "liability-driven investments." The sudden fall in the value of U.K. government bonds created a massive liquidity squeeze for these pension funds. Although rising long-term yields are positive for their funded status, U.K. pension funds had entered into interest rate derivative transactions. When the rise in yields caused the value of these derivatives to plunge, margin calls were triggered. Some pension funds were forced to sell U.K. government bonds to meet these calls, which exacerbated the rise in long-term yields and so created a need to post additional collateral. The Bank of England halted this negative spiral by announcing measures to backstop the market by acting as a buyer of last resort for the U.K. bond market, on a temporary basis. This allowed U.K. pension funds to reduce their leverage and rebuild their liquidity buffers.

We understand that U.K. banks had limited direct exposure to pension funds, or at least were not caught on the same side of the interest rate derivative transactions. That said, the episode served as a useful reminder that, when central banks implement measures to support monetary tightening, they can cause an unexpected crystallization of risks in the financial system and expose previously overlooked vulnerabilities.

Many NBFIs have seen a rapid increase in size in recent years--given that they use various forms of leverage and financial conditions are rapidly tightening, we could see similar episodes occur in future. The identification of unexpected sources of financial stress remains a downside risk for European banks.