Abstract: The Global Financial Crisis reversed the historical norm in bond markets where financial institutions’ debt, supported by regulation, liquidity access, and implicit state backing, had typically traded at tighter spreads than non-financial corporate debt. Following the collapse of Lehman and the subsequent sovereign-bank “doom loop” of the eurozone crisis, investor perceptions shifted sharply, and bank spreads widened structurally despite significant recapitalization efforts. While unconventional monetary policy helped stabilize the sector, banks faced ongoing headwinds from flat yield curves, low returns, and the introduction of loss-absorbing capital requirements. Since 2022, a mix of rate hikes, organic capital generation, reduced sovereign risk, and international diversification has materially improved fundamentals, narrowing risk premia in instruments such as credit default swaps (CDSs). Even so, financials still trade at a modest premium, less a reflection of sector weakness than of the banking sector’s structural complexity and diversity. As tracked by the iTraxx Senior index, a key gauge of CDS spreads across European issuers, this divergence remains a central feature of the post-crisis credit landscape.

Introduction
Prior to the financial crisis, the bonds issued by financial corporations tended to offer bondholders a lower return (or yield) than the bonds issued by non-financial corporations. That lower yield was mainly attributable to the fact that the issuers belonged to a regulated sector, with privileged access to liquidity and in which bankruptcies in developed economies were rare on account of implicit government support.

Since the financial crisis, however, we have witnessed a radical and structural change (within which there have been a few episodes of pronounced stress) marked by investors demanding a premium to hold bank bonds (above the yield offered by corporate bonds) to compensate for the perception that the banks are more fragile issuers than the non-financial corporations.

The collapse of Lehman Brothers in September 2008 shone the spotlight on banking regulations. The worries included losses on mortgage portfolios, counterparty risk in the interbank market and the absence of an orderly bank resolution framework. The regulators reacted with preliminary designed for strengthening the Basel framework, increasing requirements around tier-1 capital (CET1), capital conservation and countercyclical buffering. Between 2009 and 2011, the European banks issued record volumes of senior and subordinated debt to reinforce their capital at considerably higher rates than they enjoyed prior to the crisis. Perceived systemic risk had increased sharply and the banks were no longer viewed as too safe and/or too big to fail.

The premium for holding their bonds peaked at the height of the sovereign debt crisis in 2012. At the time, the banks were holding large portfolios of government bonds, close to 8% of total assets in 2012, with these holdings topping the 10% mark in countries such as Spain, Italy, Portugal and Greece. Those portfolios allowed the banks to accrue high returns and afforded them high-quality collateral, while the peripheral sovereign issuers needed the banks to fill in for foreign investors, which had reduced their holdings sharply. When the Italian or Spanish risk premiums shot up, the value of those bonds fell, raising the spectre of capital erosion at the banks and generating what the European Central Bank coined the sovereign-bank nexus, or “doom loop”. By the time the various interventions by the ECB (through its Securities Markets Programme (SMP) and, later, its Outright Monetary Transactions (OMT)) and the first Targeted Longer-Term Refinancing Operations (TLTROs) eased tensions (from 2012), the bank issuer risk premium had become a structural market characteristic.

In early 2015, in a bid to avoid deflation and spur economic growth, the ECB rolled out its asset purchase programme (APP), as even negative rates had proven insufficient in this regard. Among the various programmes, the Public Sector Purchase Programme, or PSPP, stood out. The rollout of this public debt buyback instrument flattened the yield curve (with the short end in negative terrain) and reduced yields considerably.

The battery of unconventional monetary policy measures was vital to creating an economic and financial environment conducive to allowing the banks to issue instruments at attractive rates in order to recapitalise and digest their toxic assets, paving the way for a gradual reduction in non-performance. However, that environment of negative rates and flat yield curves meant that the banks were unable to lift their ROEs back to pre-financial crisis levels. For the eurozone as a whole, the banks’ ROE averaged 6.3% between 2015 and 2019, clearly below their cost of capital.

Scant returns explains why the banks traded for so long at price-to-book ratios of less than one and also why, despite a reduction in the bank risk premium on the back of their recapitalisation, the market continued to question their business model, incapable of generating sufficient returns, keeping the risk premium above 20bp. Moreover, the introduction of new regulatory requirements, such as the minimum requirement for own funds and eligible liabilities (MREL) in Europe, expanded the hierarchy of liabilities, adding a new loss-absorbing category, senior nonpreferred, which, by virtue of being bail-in eligible in a potential bank resolution, needed to offer additional compensation.

The COVID-19 pandemic injected fresh stress into the financial sector. In addition, since the financial crisis, in any event that implies a shock for the financial markets, the banks tend to be especially penalised during the early moments (also opening up very attractive opportunities for investors during periods of stress). In March 2020, the Itraxx credit spread shot up briefly. However, the packages of public aid and the temporary suspension of dividends allowed the banks to preserve capital, while the unprecedented fiscal stimulus package translated into a much smaller economic impact than initially forecast.

Since 2022, as shown in the initial exhibit, using CDSs – the most liquid instrument – as our proxy, spreads have been narrowing consistently. Several factors underpin this improvement in fundamentals:

• The ECB’s hawkish shift in July 2022 marked a regime change. Between September 2019 and September 2023, the deposit facility rate went from -0.5% to 4%. This shift allowed the banks not only to substantially increase the rates charged on new loans and earn more from all floating-rate loans as they were repriced, it also led to remuneration on their liquidity at considerably higher rates than borne on retail funding. This in turn gave their net interest income a significant boost and pushed their ROEs back up towards 10%.
• The banks’ net issuance volumes have been much smaller in recent years, having frontloaded their refinancing effort during the period of ultra-lax monetary policy, and also thanks to renewed organic capital generation.
• As for their funding, the banks have bolstered their structural liquidity thanks to growth in deposits. The average liquidity coverage ratio is currently well above the regulatory threshold, so reducing the need to tap the wholesale funding markets frequently.
• In parallel, the banks are currently less exposed to sovereign risk. Their exposure to domestic sovereign bonds has dropped to just 6% of their assets, according to the EBA as of June 2024, and the average duration of those portfolios has also decreased. This reduced domestic exposure coupled with lower average duration is mitigating the sovereign-bank nexus that hit bank issuer spreads so hard in the past.
• The large traded banks are generating more than 40% of their gross operating income outside of their home markets, up from 25% in 2010 (ECB estimates and annual results). This international expansion, via subsidiaries, online banking platforms and pan-European investment banking businesses, reduces dependence on the domestic economic cycle and, by extension, eases the correlation between the banks’ creditworthiness and sovereign credit ratings. A greater geographic spread of risks also smooths out earnings volatility and fortifies the ability to absorb losses.

Lastly, the jump in the banks’ earnings in 2023 initially sparked sustainability concerns: many investors feared that once the ECB began to cut rates, profitability would deflate. However, three factors are tipping the balance in favour of a more stable earnings path:

• Firstly, credit volumes continue to register growth in the eurozone where there are no signs of fiscal consolidation; in fact, Germany has already announced an ambitious infrastructure and defence investment programme.
• Secondly, a steeper rate curve, shaped by a growing need for long-term public financing, preserves net interest margins even if official rates are being tapered.
• And thirdly, growing numbers of banks are diversifying their earnings streams by getting into the insurance and asset management businesses.

Combined, these factors reinforce the idea that the reversal of the bank risk premium is attributable to both the new regulatory framework and a structural (and not merely cyclical) improvement in sector profitability.

Structural appeal of the bank bond market
Beyond the recent premium compression, the financial issuer bond market stands out for its depth, diversity and dynamism. Approximately 26% of the universe of senior bonds denominated in euros – including both investment grade and high yield issuers – corresponds to financial sector issuers, with more than 1,148 active issues. That contrasts with the next biggest sector, utilities, with around 565 issues. This breadth, coupled with the frequency with which banks of all sizes tap the primary market, translates into high liquidity, fully covered credit curves and multiple tactical entry points. Advantages that are hard to replicate in other sectors where just a few issuers account for the bulk of the index.

However, an analysis of the bond market from a sector perspective (and not through the CDS market, which is limited to the major banks), could give the idea that the bank sector continues to offer an additional premium. To illustrate this idea, we selected a sample of bonds that mature in 3 to 5 years with credit ratings ranging between BBB and A-. [1] In this group, the average spread over the bank sector swap rate stands at 88 basis points, which is comparable to that of sectors that are currently under pressure, such as the automotive industry (93bp) and the real estate sector (90bp), and clearly above the sample average (71bp) and the minimum spread observed in the telecommunications sector (56bp).

The bank sector, however, is not going through anything of the kind. The reason for its higher average spread lies with the market structure itself: a far broader and more heterogeneous issuer base made up of entities of different sizes, from different markets and with different business lines and risk profiles. This diversity contrasts with the concentration that characterises other sectors, dominated by large, consolidated corporations or national players operating in quasi-monopolistic environments.

In short, the spread observed on bank bonds should not be interpreted as a sign of weakness but rather as the manifestation of the structural richness of the financial issuer market which offers unique opportunities for analysis and tactical investment.

Bank bond spreads: Between normalisation and caution
The circumstances underpinning the bank bond risk premium have disappeared. Recapitalisation, the bank resolution framework and international diversification have put the European banks on an even footing – and even at an advantage in liquidity – with the large corporate issuers. With over 95% of the MREL targets already met, the sector offers visibility into issue schedules and volumes, eliminating the threat of a ‘maturity wall’. There are no fundamental reasons for financial issuer bonds to trade structurally at a premium to non-financial corporate credit.

The convergence is reversible, however. It should hold as long as: (i) bank ROEs remain above their cost of capital; (ii) asset performance remains in check, particularly in the CRE and SME segments; (iii) sovereign deleveraging prevents reactivation of the banksovereign loop; and (iv) the ongoing reduction in sovereign bond holdings limits the banks’ sensitivity to country ratings.

It could reverse if a return to low rates compresses margins and rekindles the search for risk/returns; if the energy transition increases the cost of risk in carbon-intensive sectors; or if a geopolitical shock triggers mass issuance of public debt, exerting fresh pressure on bank asset mixes.