Costly investments: the Savings and Investments Union (Part 2)

The Savings and Investments Union (SIU) initiative, spearheaded by the European Commission, aims, in particular, to increase the return on household savings and, at the same time, to reduce the cost of financing for businesses. Attentive readers will point out that this is contradictory, as a return on the former results in a cost for the latter. Actually, this is not the case, provided that the focus is on the frictions that slow down and increase the cost of channelling savings into the different productive investments. Naturally, these frictions generate income for intermediaries, who, in turn, contribute to GDP. However, insofar as they reduce both the supply of and demand for capital, they result in a social loss: the intermediaries’ profit is lower than the sum of the losses incurred by savers (households) and borrowers (non-financial corporations). 

In my previous post, I examined the benefits of the SIU in terms of savings. Let’s now turn our attention to investments. As everyone knows, external funding for businesses (i.e. excluding self-financing) can be divided into two main categories: borrowing and raising equity capital. Let’s look at them one by one. 

Costly borrowing 

In continental Europe, non-financial corporations (NFCs) borrow mainly from banks. According to ECB data, in June 2026, the outstanding value of bank loans to non-financial corporations in the euro area was nearly three times that of corporate bonds. Small and medium-sized enterprises (SMEs) are particularly dependent on bank loans, as they rarely have the size, credit rating or financing volume required to absorb the fixed costs of bond issuance. 

Bank loans have many advantages. A long-term relationship between a bank and a business provides information that is difficult for markets to obtain and makes credit less pro-cyclical: it increases less when the economy is doing well and contracts less when the economy takes a downturn. This reduced procyclicality is further reinforced by the banks’ resilience to liquidity shocks, owing in particular to deposit insurance and access to central bank liquidity (see Freixas and Rochet, 2008). 

Market debt (negotiable debt securities, in particular commercial paper and bonds) is therefore less of a substitute for bank credit than a supplement to it. By broadening the investor base, market financing allows for better risk-sharing and reduces companies’ dependence on the banking sector. There is a significant potential for developing this form of financing within the EU. Contrary to popular belief, the European bond market is just as integrated as that of the United States (see Chiţu et al. 2026). However, this potential remains largely untapped today. In 2024, debt securities accounted for just 10% of debt financing by European NFCs (11% of EU GDP), compared with 28% in the United States (29% of US GDP) (see Gossé et al., 2026). 

The aim of the SIU is not to replace loans with bonds, but to better harness their complementarity. Securitisation is a good example of this. By pooling loans (particularly loans to SMEs) and then issuing securities backed by these loans, banks can transfer part of the risk and free up funding capacity. Provided that high standards of transparency and risk management are maintained, securitisation can thus increase the lending capacity of the system as a whole. Accordingly, the package proposed by the European Commission in June 2025 aims to simplify the European securitisation framework without compromising its robustness. 

The European project also seeks to enable more companies to gain direct access to the commercial paper and bond markets by reducing issuance costs, in particular infrastructure and supervisory costs. These costs have a proportionally greater impact on smaller issuance volumes and tend to exclude SMEs. There are two levers that can be used to reduce these costs: on the one hand, modernising market infrastructure through distributed ledger technology (DLT) and the tokenisation of financial assets, and, on the other, better integrating market supervision. 

Tokenisation refers to the representation of a financial security in the form of a token recorded within an infrastructure that utilises DLT. This technology makes asset division easier. It also enables certain operations to be automated using programmable contracts and speeds up settlement. Lowering barriers to entry and creating a secondary market could help to broaden the investor base and improve liquidity (see Carapella et al., 2023). The benefits of tokenised bond issues for issuing companies may be substantial: according to Leung et al., (2023), who analysed Asian and European corporate bonds between 2018 and 2023 (primarily in 2021 and 2022), issuing costs (relative to face value) are on average 0.22 percentage point lower, and interest rates would also be lower, by 0.78 percentage point, compared with conventional bonds with identical characteristics. Although it covers only 23 issuers, and is likely to suffer from selection bias, this study suggests the potential for very substantial gains from the perspective of the SIU. 

However, technology alone will not be enough to create an integrated market. Without common standards, interoperability between platforms and secure settlement assets, DLT could exacerbate fragmentation rather than reduce it: financial markets would organise themselves into distinct ecosystems, managed by the private sector; national borders would be replaced by technological borders, much like the Apple, Microsoft and Google ecosystems in the field of digital systems. 

That is why the ECB’s initiatives in this area will be crucial. Following the trials carried out in 2024, the “Pontes” project aims, from the third quarter of 2026, to link platforms using DLT to TARGET services in order to enable the settlement of tokenised assets in central bank money – a universal currency for the entire euro area – unlike stablecoins, which are not interoperable. Looking ahead to 2028, the “Appia” project is developing a tokenised, integrated financial ecosystem across Europe. 

These initiatives will, to a certain extent, help to overcome the current fragmentation of European financial infrastructures. For instance, all central securities depositories will have access to shared information on a distributed ledger. However, fragmentation also stems from a heterogeneous supervisory landscape, with a total of 52 national supervisory authorities (Carmassi et al., 2026). In the context of cross-border activities, this fragmentation increases legal and compliance costs. A single supervisory framework would reduce the complexity of supervision for both supervisors and supervised entities. The Market Integration and Supervision Package (MISP), presented by the European Commission in 2025, is part of this approach. It proposes that the European Securities and Markets Authority (ESMA) be given responsibility for supervising certain significant market infrastructures, in order to simplify cross-border activities. 

Equity union 

The chart below compares the financing structure of French and euro area companies with that of companies in the United Kingdom and the United States. As we have seen, European companies rely more on bank loans than on market financing, whereas the opposite is true in the United States. But what is particularly striking is the proportion of equity financing, which is much higher in the United States than in Europe. If we compare the aggregate amount of equity capital not with the aggregate liabilities of non-financial corporations, but with GDP, the difference is even more striking: 230% of GDP in the United States in the first quarter of 2026, compared with 78% in the euro area (93% excluding Germany, where the statistics are not entirely consistent with those of the euro area) and 81% in France. Corporate debt in the United States, on the other hand, accounts for a smaller proportion than in the euro area. 

Based on the Modigliani–Miller theorem (1958), it might be assumed that the structure of a firm’s liabilities is irrelevant to its success and value. The only thing that matters is the return on assets. However, the conditions for validating this theorem are never met. In particular, markets are not frictionless; there is a risk of default; information is not symmetrical; and profits are subject to tax. A well-capitalised firm therefore borrows at lower cost and is better able to withstand financial shocks than a firm financed mainly through debt (see Holmstrom and Tirole, 1993). It also has greater capability to innovate because it is not required to pay its shareholders on as regular and predictable a basis as its creditors. 

Chart: share of equity and debt in the liabilities of non-financial corporations 

(in percent, Q1 2026)

Both listed and unlisted shares are equally important, as these two forms of equity financing are complementary. Venture capital funds the early stages of a start-up, whilst growth capital supports its expansion. An initial public offering (IPO) then allows companies to raise larger sums of money, as well as offering early investors a way to exit successful projects by realising a capital gain, which is then reinvested in new businesses. 

As we know, Europeans invest relatively little in equities, and when they do (or when their life insurer or pension fund does so on their behalf), they tend to favour companies in their own country and US companies. Not that they necessarily want to invest in the United States. But the fact is that the United States is home to the deepest and most dynamic equity markets. According to Lappe and Pinkus (2025), the significant proportion of US listed shares held by European insurers and pension funds (39% of their portfolio in 2023) reflects stock market performance, the depth and liquidity of the US market, and the diversification strategy of institutional investors, rather than a bias against European equities. 

The choice of unlisted shares is also greater in the United States. Therefore, US venture capital funds are significantly larger than their European counterparts, with total invested capital amounting to EUR 930 billion in the United States compared with less than EUR 150 billion in the EU (for venture capital funds that have participated in at least one transaction during the period 2015–25, see Banu et al., 2026). Europeans wishing to invest in venture capital will therefore naturally look to the United States for their investments. 

In essence, European investors buy whatever they can find on the market, but the market is based on a flawed equilibrium: European companies find it easier to raise equity capital in the United States, and European investors are therefore pushed to look across the Atlantic to access a dynamic equity market. 

Still, raising capital through listed shares is not necessarily more expensive in Europe than in the United States: although, on average between 2004 and 2025, the risk premium on EU-listed shares (the additional return required compared with risk-free assets) was around 1.6 percentage points higher than that observed in the United States, this difference virtually disappears when share buy-backs are included in the calculation. This suggests that there is latent demand for listed European shares, which is likely to materialise once companies have easier access to this form of financing and financial intermediaries operate within a less fragmented legal, tax and technological framework. 

The economic stakes are substantial: according to simulations by Arnold et al. (2025), the integration of capital markets could boost the EU’s GDP by 1.2% over ten years, thanks in particular to an increase in venture capital investment, which would foster business creation, investment and total factor productivity. For their part, Venditti et al. (2025) estimate a reduction in the cost of capital of around 0.5 percentage point, which would increase investment by around 1% of GDP and raise GDP to 1.5% above the baseline scenario after ten years. This is particularly important for the unlisted segment, which is key to innovation. Beetsma et al. (2026) show, for example, that when a pension fund acquires a stake in an unlisted Danish company, this results in a 3-5% increase in its productivity in the years following the investment (up to five years). The effect increases with the size and duration of stake. 

How can listed and unlisted shares be positioned at the centre of the European financial system? Four successive reports – Noyer (2024), Letta (2024), Draghi (2024) and Kukies-Noyer (2026) – present, from different perspectives, a convergent assessment of the fragmentation of European capital markets and the shortfall in equity financing. In particular, the Noyer report (2024) highlights the institutional barriers to the development of deep capital markets, notably the fragmentation of the supervision of settlement infrastructures. He also stresses the importance of the investor base and advocates the development of long-term European savings products and the revival of the securitisation market. The Kukies-Noyer report (2026) focuses more specifically on the funding gap faced by innovative companies, particularly those in the growth phase, which it attributes to a lack of institutional funding, regulatory barriers and the fragmentation of market infrastructure. 

The “MISP” package, mentioned above, directly addresses the need to reduce fragmentation in the European financial market, notably through the centralised supervision of significant financial market participants – trading venues, clearing houses, central securities depositories and crypto-asset service providers. This European legislation, which is currently under discussion, will also facilitate the tokenisation of the financial market, which could make equity financing more accessible to SMEs (see, for example, Lise – Lightning Stock Exchange – in France). In addition, the “28th regime” (EU Inc.) project aims to reduce the legal fragmentation of European companies by offering them a single framework, which will remain optional. 

Against the backdrop of these efforts to address fragmentation, the European Commission wishes to encourage long-term saving through the development of pension funds to complement pay-as-you-go pension schemes. It therefore recommends the auto-enrolment of employees and the introduction of tracking systems for pension entitlements; and proposes a revision of the IORP II Directive to facilitate the cross-border development and operations of occupational pension funds, and of the PEPP Regulation, to make the pan-European individual pension savings product more attractive. 

Lastly, regarding innovation, the EU Startup and Scaleup Strategy directly addresses the limited capacity of European venture capital funds and the funding gap faced by European scale-ups, notably through the revival of the “fund of funds” European Tech Champions Initiatives (ETCI 2.0) and the creation of the Scaleup Europe Fund. In addition to the funds raised and directly invested (EUR 15 billion of public funds channelled into European funds under the first scheme, and EUR 5 billion invested directly in European technology scale-ups under the second), the strategy is based on leveraging private investors. The ECTI 2.0 initiative could therefore raise up to EUR 80 billion from institutional investors. 

The safe asset paradox 

A safe asset is a financial instrument characterised by a very low probability of default, as well as low liquidity and market risks, which means that it is easy to buy or sell even with limited information (see Gorton, 2017). Sovereign debt securities denominated in “hard currency” and with high credit ratings are the prime example of this. Unlike purely “safe haven” assets (such as gold), whose value is based solely on an increase in demand during times of uncertainty, high-quality government bonds can fulfil both functions: they are used as collateral in financial markets and as a source of liquidity for banks in normal times (safe assets), and even more so in times of crisis (safe-haven assets). 

In comparison with the United States, Europe has a relative lack of safe assets: in April 2026, if we combined the aggregate outstanding sovereign debt of euro area countries rated at least A (on a scale ranging from AAA to D) with the different segments of European supranational debt, we would obtain a total of EUR 9.8 trillion. At the same date, the outstanding US federal debt was almost three times as high (USD 31.2 trillion, or EUR 26.7 trillion). 

It might seem paradoxical to advocate the development of a safe European asset with the aim of stimulating the market for risky assets – equities and corporate debt. However, no one is surprised by the fact that the United States is both the largest market for equities and for Treasury bonds. Both markets (safe assets and risky assets) develop in a complementary manner. In particular: 

• The so-called “risk-free” interest rate on a safe asset provides a common benchmark for the cost of corporate finance, which is simply the risk-free rate plus credit risk premiums and liquidity premiums. By reducing the risk-free rate, a strong demand for euro-denominated safe assets will, in turn, bring down borrowing costs for businesses. This common benchmark will also help to foster convergence in financing costs across companies. 
• Safe assets play a part in the balance sheet management of financial intermediaries: they provide liquidity (particularly in the repo markets) and a hedge against systemic shocks, enabling them to take on more risk in other areas, notably by investing in less liquid assets. 

The subject is not new, but the approach has changed. When it emerged in the wake of the euro area crisis, the initial focus was on the stabilising role of safe assets in breaking the negative feedback loop between sovereign risk and banking risk, and in mitigating the consequences of the fragmentation of sovereign debt markets (see Gossé and Mourjane, 2021). This stabilising role remains important given the vulnerability of fragmented markets to global shocks (see Mosk and de Vette, 2025). However, the debate has resurfaced today, with the emphasis now on the role of the safe asset as a catalyst for private investment, a means of financing European public goods and a vehicle for the international role of the euro (see Villeroy de Galhau, 2025; Rehn, 2026). Two sets of proposals have been put forward: 

• The “stock” approach is based on various financial engineering options aimed at transforming national sovereign debt into “senior” supranational debt, after some financial engineering and/or partial pooling (Brunnermeier et al., 2017; Blanchard and Ubide, 2025). While this approach is politically challenging, it avoids creating new debt while enabling the volume and liquidity of the market to be increased rapidly. 
• The “flow” approach (see Hildebrand, Rey and Schularick, 2025) relies on the funding requirements for public goods (for example, European defence) and is more politically acceptable but entails a relatively slow development of the market. This approach would involve institutionally strengthening the EU’s permanent borrowing capacity (Zettelmeyer and Lappe, 2026), with a possible first step centred on a coalition of the willing to issue common debt (Dorrucci and Rossi, 2026). 

Conclusion 

Beyond its direct impact on the financing of the economy, the SIU project serves as a strategic lever for strengthening the European Union’s economic and financial position in the long term. By channelling more savings towards productive investment and innovation, it will support potential growth and reduce Europe’s technological dependencies, whilst offering savers a better return. The rise in productive investment will also help to absorb some of the euro area’s excess savings and, consequently, its current account surplus, thereby contributing to the rebalancing of the global economy. Lastly, enhancing the attractiveness and depth of European financial markets should draw more investment flows from other regions, promote the international distribution of euro-denominated assets and strengthen the single currency’s role in the international financial system.