European savings and business expansion: the risk-bearing connection

A large stock of financial savings does not guarantee that a young firm can finance an uncertain expansion. The saver may require liquidity and capital protection, the firm may need patient loss-bearing capital, and the intermediary may face costs in assessing a business outside its familiar market. The missing connection is not necessarily an absence of money. It can be a mismatch among the terms on which money is supplied, the risks that must be carried and the opportunities capable of producing a return.

That distinction is central to Europe's savings and investment debate. A deposit, a listed share, a venture fund commitment and a company's purchase of equipment are different transactions. Moving between them can improve the financing system, but the amount moved is not automatically additional productive investment. The connection must be traced through intermediation, issuance, capital allocation and actual expenditure.

This report examines that connection using the European Commission's Savings and Investments Union strategy, ECB analysis of integration and private markets, the EIB's firm survey and ESMA's cost evidence. Its geographical boundaries are explicit: some sources cover the European Union, while the ECB's integration indicators concern the euro area. EU totals are not relabeled as euro area totals. Germany and France participate in that wider setting, but no unsupported national ranking is constructed.

The principal finding is that a successful investment-financing system needs several complementary capacities. It must provide households with suitable investments, bring diversified institutional risk capital to firms, reduce avoidable cross-border costs and offer credible routes through successive financing stages. It must also distinguish financially constrained projects from projects limited by demand or operating conditions. Raising fund assets alone does not demonstrate that these connections have improved.

The Commission's March 2025 communication places approximately ten trillion euros of EU retail savings in bank deposits and cites additional annual investment needs of 750–800 billion euros by 2030. The latter is an estimated need, not an appropriation, a financing commitment or an observed spending shortfall in every year. The strategy proposes a better connection between savings and productive investment. European Commission, COM(2025)124.

The ECB's 2026 financial-integration report presents a differentiated assessment. It finds broad integration progress since late 2022 while identifying weaker equity integration, continuing legal and tax barriers, and constraints in the venture-capital investor landscape. Its analysis also emphasizes exit opportunities and institutional participation. Those findings concern particular markets and mechanisms; they do not support a statement that every form of European financial integration has deteriorated. ECB, Financial Integration and Structure, 2026.

Firm evidence adds an important qualification. EIB's 2025 survey reports that 86 percent of EU firms were investing, compared with 87 percent in the preceding survey. Its measure of externally finance-constrained firms fell from 6.7 to 6.1 percent. That category includes rejection, insufficient finance, excessive cost and discouraged applications. It is a survey definition, not the share of all investment opportunities with positive social returns. EIB Investment Survey 2025, EU overview.

The evidence therefore rejects an undifferentiated claim that European firms generally cannot obtain any finance. It remains consistent with serious constraints at particular stages, sectors or locations. A system can finance established firms adequately while failing to supply the size and type of equity needed for rapid expansion. Aggregating those cases into a single financing gap obscures the intervention required.

A bank deposit is a household asset and a bank liability. It can support a financial intermediary whose assets include loans and securities. Calling the entire deposit stock idle money ignores this intermediation and can misstate what happens when households purchase a different asset. The economic question is whether the system allocates risk and financing effectively, not whether every deposit has been physically withdrawn.

Suppose a household buys a newly issued fund share for ten using a bank deposit. The fund then holds a deposit or purchases an asset from another counterparty. Deposits can move among holders while the household's portfolio changes. The initial transfer does not by itself reduce the banking system's deposits by ten or create ten of corporate capital expenditure. The subsequent transaction and banking settlement determine the balance-sheet changes.

If the fund buys a company's newly issued equity and the company uses the proceeds for machinery, the chain can finance productive assets. If the fund buys an existing share from another investor, the company receives no new cash from that trade. Secondary-market activity can nevertheless improve liquidity, valuation and future issuance conditions. It should be assessed through those channels rather than counted as immediate primary finance.

A similar distinction applies when investors buy foreign assets. Portfolio diversification can improve household risk management and expected returns. Domestic investment can still be financed by other investors. The net economic effect depends on the entire financing pattern and relative opportunities. Treating every foreign purchase as an equal loss of domestic productive capacity assumes a counterfactual that the transaction alone cannot establish.

The Commission's deposit figure is consequently a starting observation about household portfolios. It is not an amount available without risk or institutional cost for direct transfer to innovative firms. Some balances meet transaction needs, emergencies or near-term commitments. A credible policy assessment should identify the portion suited to longer holding periods and the instruments capable of meeting those savers' requirements.

An early research business, a firm proving commercial demand, an established exporter expanding production and a mature utility have different financing requirements. The expected cash flow, collateral, information and tolerance for failure differ. A banking loan that works for the utility may be unsuitable for a business whose main asset is an unproven process or specialized knowledge.

The relevant comparison is the instrument's allocation of risk. Debt generally promises specified payments subject to contractual conditions. Equity absorbs losses and participates in gains. A guarantee transfers defined losses to another balance sheet. A public grant pays for an activity without the same repayment promise. These are different ways of financing an uncertain project; equal headline amounts do not imply equal economic support or fiscal exposure.

The stages can overlap and do not prescribe one instrument for every firm. A young company with contracted revenue may borrow successfully, while an established business entering a risky technology may require equity. The matrix identifies questions to test rather than a rigid funding ladder.

The continuity between stages is especially important. Financing a prototype without a credible route to commercialization can leave a technically successful firm unable to expand. Financing rapid growth without sufficient working capital can create a liquidity failure despite strong demand. An investment program should assess the next stage and its funding requirements rather than report each initial commitment as a completed success.

Consider an explicitly hypothetical project requiring one hundred today. In two years it pays two hundred with probability sixty percent and ten with probability forty percent. There are no intermediate distributions or additional costs. Its expected payoff is one hundred twenty-four. At an assumed ten-percent annual discount rate, the present value is approximately one hundred two point five. This valuation depends entirely on the stated probabilities and discount rate; neither is an empirical estimate.

Now suppose a lender offers the full one hundred in return for a maximum repayment of one hundred ten in two years. It receives one hundred ten in success and only ten in failure. Its expected recovery is seventy, calculated as 0.60 × 110 + 0.40 × 10. The lender cannot capture the upside beyond its contractual repayment. The same project can therefore have a positive stipulated equity value and an unattractive unprotected full-debt recovery.

This example does not show that banks should make the loan or that equity investors will fund every uncertain venture. The assumed project value is only slightly above cost and could disappear with different probabilities, execution costs or a risk-adjusted discount rate. It demonstrates why replacing risk-bearing capital with more ordinary lending can fail to solve the relevant constraint.

Collateral or a guarantee changes the lender's recovery, but the additional protection has a cost and an owner. If the public sector absorbs losses, the risk has moved rather than vanished. If an entrepreneur pledges personal assets, the financing can impose substantial household exposure. A comparison of instruments should therefore include who bears the downside and receives the upside, not only the interest rate.

The controlled example identifies a risk-allocation problem before choosing an intervention. If a lender cannot recover enough in the failure state, lowering administrative costs alone may leave the loan unviable. If an investor can bear losses but lacks reliable information, better disclosure may matter more. Evaluation should identify the binding condition and examine whether the chosen instrument changes it on viable terms.

A venture fund's commitments are not identical to cash available for an immediate new round. Management expenses, reserves for existing firms, investment pacing and contractual conditions affect deployable resources. Several funds can each support small early rounds yet lack capacity for a large later round. A market can consequently have many transactions while still presenting a financing discontinuity at expansion.

For a hypothetical fund with commitments of fifty, assume fifteen is reserved for follow-on investment and five for lifetime expenses. Thirty remains for initial investments. If each initial investment is ten, the fund can finance three such positions under the assumptions. The headline fifty does not support five independent ten-unit initial rounds in addition to those other commitments.

Now consider a larger fund with commitments of two hundred, follow-on reserves of sixty and expenses of ten. Initial deployable resources are one hundred thirty, supporting thirteen ten-unit positions. The different expense proportion is assumed explicitly; it is not a claim about actual fund economies. The example shows how both scale and allocation affect the set of rounds that can be financed.

A larger fund is not automatically a better investor. Concentration, governance, sector expertise and incentives affect selection. Pooling more resources can enable later rounds but also create pressure to deploy larger amounts in a limited opportunity set. The relevant outcome is disciplined capital allocation and successful financing continuity, rather than fund size as an objective in itself.

Institutional participation should be assessed similarly. Long-lived liabilities can make some institutions natural holders of long-term assets, but obligations, valuation needs, governance and risk budgets differ. An allocation rule that ignores those constraints can transfer illiquidity to beneficiaries rather than create sustainable patient capital. Evidence on the actual investor base and commitment terms is needed before a larger institutional allocation is presented as costless.

An investor values not only the business's future profits but also the feasible ways of realizing its claim. A public listing, sale to another company, secondary sale or continuing distributions can provide different routes. Weak or uncertain exit opportunities can reduce the price investors are willing to pay initially, even if early technical progress is promising.

Consider a company expected to produce a realizable investor payoff of two hundred after five years. Discounted at ten percent annually, that payoff is worth about one hundred twenty-four point two today. If the same payoff cannot be realized until year eight, its present value falls to approximately ninety-three point three. The difference is caused solely by the assumed timing, not by worse operating performance.

These numbers are hypothetical and do not estimate European exit discounts. They show why market depth, disclosure and credible ownership transitions can influence initial finance. A reform improving issuance or settlement should be evaluated through its effect on transaction costs, investor participation and attainable capital, rather than its technical adoption alone.

An exit is also not synonymous with new money entering the company. A sale of existing owners' shares can provide liquidity to those owners without financing the firm's next investment. A transaction may combine primary and secondary elements. Reporting their split helps determine whether the firm obtained expansion capital, earlier investors realized a return, or both occurred.

Nor does an overseas listing establish that the firm's production immediately leaves Europe. Headquarters, legal incorporation, market listing, research, employment and customers can be located in different places. A competitiveness assessment should follow those activities separately. The financial location can matter for future networks and governance, but its operating effects require evidence beyond the exchange selected.

A common currency removes a currency-conversion obstacle within the euro area, but it does not make every investment transaction operationally identical. Due diligence, legal interpretation, tax procedures, investor identification and corporate actions can still involve different processes. The ECB's analysis identifies these frictions as part of the continuing integration problem.

For an investor, a small fixed cost can be significant on a small position. Suppose a hypothetical cross-border investment requires an additional administrative cost of one thousand euros. On an investment of ten thousand, that is ten percent of principal; on one million, it is one-tenth of one percent. The same fixed friction can discourage smaller cross-border allocations disproportionately.

Reducing that cost can broaden participation without changing the project's physical technology. It can also make diversification more accessible. The relevant evaluation should measure actual all-in costs, processing time, failed or abandoned transactions and investor use. A harmonized form or platform can be an enabling step while its economic effect depends on adoption and the remaining legal requirements.

Institutional integration and market concentration should not be conflated. A larger common market can support competition among intermediaries, while a consolidated platform can also have market power. The assessment should consider pricing, access, resilience and the ability to change provider. Scale can lower unit costs without guaranteeing that the saving reaches investors or issuers.

The distinction between broad integration and equity integration is material here. The ECB's 2026 findings allow improvement in some markets alongside persistent weaknesses in others. A single composite index cannot determine whether an innovative business can access the relevant equity investor. The instrument, investor base and transaction process should remain the unit of analysis.

For a household, moving from a deposit to a market instrument changes the distribution of possible returns, access to cash and exposure to loss. The policy objective should be suitable participation, not an assumption that every household should hold the same portfolio. Emergency needs, expected expenditure, pensions and capacity to bear risk differ.

Costs affect both suitability and the capital ultimately available. ESMA's November 2025 study found that distribution accounted for 48 percent of UCITS total costs in its assessment. That percentage concerns the composition of measured costs, not a charge of 48 percent of invested capital. Its distinction is essential when discussing the effect of distribution arrangements. ESMA, total fund-cost findings.

A separate hypothetical compounding exercise illustrates a cost channel. Assume an initial investment of one hundred, a constant gross annual return of five percent and annual costs reducing the net return to either four or three percent. After twenty years the values are approximately two hundred nineteen point one and one hundred eighty point six. The calculation assumes no taxes, withdrawals, volatility or further contributions. It is not a return forecast or product recommendation.

The cost difference matters, but the cheapest instrument is not automatically appropriate for every purpose. Advice, service and access can have value, while incentives can create conflicts. A useful comparison identifies the service supplied, total cost and alternatives. Quoting a product's management fee alone can omit distribution or transaction costs relevant to the investor's experience.

Trust also depends on understandable information and dependable enforcement. A retail participation initiative can attract funds initially yet lose support if risks or access conditions are misunderstood. Evaluation should include realized costs, complaints, persistence of participation and the investor's ability to meet cash needs, rather than only account openings or assets gathered during a favorable market period.

Long-term business investment is often difficult to sell quickly at a reliable price. A fund holding such assets must reconcile that characteristic with the redemption rights offered to investors. Expanding retail access without addressing this mismatch can make the funding base less stable precisely when firms need continuity.

Suppose a hypothetical fund has assets valued at one hundred, including ten in cash and ninety in illiquid company claims. Investors request redemption of twenty during a stressed period. Even before any valuation change, the fund lacks ten of immediately available cash. Selling a claim at a discount can transfer losses among exiting and remaining investors depending on the rules and valuation process.

The example does not imply that every private fund has daily redemption or that a particular European product is unsafe. It identifies the contractual question: who bears the cost when investors seek cash faster than assets can realize value? Notice periods, gates, closed-end structures and cash reserves are different responses. Their use should match the investor and asset rather than be assumed from the fund's label.

Public guarantees do not necessarily solve liquidity mismatch. They can protect a defined credit loss while leaving the timing of cash realization unchanged. A liquidity facility can supply cash while increasing leverage or contingent exposure. Instrument design should separate credit protection, market valuation and liquidity, because each addresses a different condition.

The systemic assessment should also follow connections among banks, insurers, funds and investors. Diversification of financing channels can reduce one concentration while creating new common exposures. Growth in nonbank assets is therefore evidence of structural change, not an automatic improvement in resilience. The required evidence concerns leverage, funding, counterparties and the behavior of claims under stress.

A guarantee can make private funding possible by changing who absorbs losses. Its financial leverage ratio does not measure its economic additionality or expected fiscal cost. The face value supported, maximum payout, expected payout and behavioral response belong on separate lines. An instrument can have a small expected cost while exposing the budget to a concentrated loss in a rare state.

Consider an invented portfolio of one hundred. Assume a ninety-eight-percent probability of a one-unit loss and a two-percent probability of a twenty-unit loss. The expected loss is 1.38. Now suppose a public guarantor covers the first five units of loss. Its expected payment is 1.08: 0.98 × 1 + 0.02 × 5. Private investors' expected remaining loss is 0.30, arising from fifteen units in the stress state. The sum remains 1.38; the guarantee has redistributed the assumed loss rather than eliminated it.

The maximum public payout is five. Describing the instrument as mobilizing twenty times its maximum public coverage uses the ratio 100 / 5. That ratio is an exposure description, not proof that all one hundred of financing is additional. If most investors would have participated without protection, incremental funding is smaller. If the guarantee changes project selection or monitoring, the loss distribution can also change.

Stress-state probability illustrates another limit. Raise it from two to five percent, leaving the two loss amounts fixed. Expected public payment rises to 1.20, while expected private loss rises to 0.75. A relatively modest change in the assumed tail probability materially changes the private exposure left after protection. These are sensitivity calculations, not calibrated European default probabilities.

Evaluation should therefore examine the allocation rule, covered claims, loss priority, fees, monitoring and counterfactual funding. It should retain failed projects and claims paid after the program's investment period. A guarantee can be useful because a public balance sheet can bear a particular risk on acceptable terms, but that judgment requires an explicit account of costs and objectives. The headline private amount does not supply it independently.

Cross-border procedures can matter even when investors ultimately owe the same tax. The ECB identifies withholding-tax refund procedures among the remaining frictions. The economic mechanism includes both administrative effort and the time between a payment being withheld and an eligible refund being received. Final tax incidence and temporary cash loss are different quantities.

For a hypothetical distribution of one hundred, assume thirty is withheld initially, while the properly determined final liability is twenty. The investor expects a refund of ten after two years. At a five-percent annual discount rate, that refund has a present value of approximately 9.07. Relative to receiving it immediately, the financing cost is about 0.93 before any administration fee or uncertainty about recovery.

The example is deliberately not tied to a member state's actual tax rate or treaty entitlement. Those require current transaction-specific rules. It shows why reducing processing delays can improve an investor's net outcome without changing the final statutory liability. The appropriate evaluation records refund time, rejection, documentation costs and actual participation, rather than counting a procedural reform as an automatic reduction in every investor's tax bill.

Finance can be abundant while firms decline to expand because expected sales are uncertain, energy or labor costs are high, or operating inputs are unavailable. The EIB survey's distinction between investment activity, expectations and external financial constraints is valuable precisely because these margins differ. A financial reform should not be credited with removing a nonfinancial constraint it does not address.

An established manufacturer may have access to a credit line but postpone a plant because customer demand is unclear. A research company may have strong customer interest but lack equity for validation. The first case calls for understanding demand and uncertainty; the second concerns risk-bearing finance. Their observed investment can be equally low while the appropriate explanation differs.

An empirical assessment should examine denied finance, offered terms, internal resources, investment plans and stated obstacles together. Self-reported constraints provide useful evidence but can be influenced by expectations or interpretation. Firms not seeking finance may be unconstrained or discouraged; those categories require different treatment. A high share investing does not establish that every firm invests enough, and a low rejection rate does not establish that no viable applicant is deterred.

Selection also affects comparisons between bank-funded and equity-funded companies. Investors may choose more productive or faster-growing firms. Those firms' later performance cannot automatically be attributed to the funding instrument. The ECB's private-market discussion itself recognizes screening and risk differences, reinforcing the need to distinguish a selection effect from a causal productivity effect.

The report's conclusion is therefore complementary rather than exclusive. Better financial connections can remove specific constraints and improve risk allocation. The amount of investment ultimately undertaken depends on viable opportunities and operating capacity. A diagnosis that identifies both finance supply and investment demand is more actionable than assigning every shortfall to a national savings preference.

An evaluation should trace a defined reform to an observable financing change and then to the firm's use of funds. For a cross-border cost reform, the intermediate outcome might be lower processing costs and wider investor participation. For institutional venture participation, it might be larger follow-on rounds on viable terms. For retail distribution, it might be lower total cost with suitable investor outcomes.

The next stage is additionality. A firm receiving new equity might replace a planned loan, acquire another business or retain cash. Those uses can have economic value but differ from additional fixed capital formation. A credible assessment compares the investment path with what would otherwise have happened, including substitution among funding sources.

The unit should remain stable through the process. Record primary funds raised, secondary shares sold, fees, refinancing, acquisitions, cash accumulation and actual expenditure. Then identify the location and asset type of expenditure. A large fund commitment, a large issuance and a large transaction value can refer to the same capital chain; adding them would overstate resources mobilized.

Research designs should exploit documented changes in eligibility, cost or access while testing pre-existing trends and selection. A simple before-and-after increase in investment can reflect the business cycle or a concurrent subsidy. A comparison across countries can reflect industry composition and demand. Where those confounders cannot be separated, the result should be described as an association or a financing-account reconciliation rather than a causal estimate.

A monitoring framework should also preserve failures and unsuccessful applications. Measuring only surviving funded firms overstates performance and hides the cost of experimentation. Appropriate outcomes can include innovations tested, productive assets deployed, viable follow-on funding, investor returns after costs and losses borne. No single success ratio can summarize all those objectives without explicit weights and a stated horizon.

Household access, institutional risk capacity, cross-border processes and firm investment opportunities are connected. Improving one can have a limited effect if another remains binding. A lower-cost retail account may direct funds toward existing liquid securities rather than early-stage enterprises. A larger venture pool may struggle if viable exits and later financing are scarce. A common settlement process may remain underused if legal or tax frictions persist.

The implementation sequence should therefore follow the bottleneck being addressed. Establish the relevant costs and constraints; identify the instrument and investor capable of bearing the risk; provide credible disclosure and market access; and evaluate the resulting primary finance and expenditure. Reforms can proceed together, but their outcomes should remain separately measurable.

Public support should retain a transparent fiscal account. A first-loss guarantee, co-investment and grant expose the budget differently. Their expected costs, maximum exposure and conditions should be recorded alongside private commitments. Reporting only a leveraged headline amount conceals how much risk the public sector absorbs and how much private finance would have arrived without the intervention.

The investor's interest and the domestic investment objective should both remain visible. Geographical restrictions may increase local allocation while reducing diversification or returns. Unrestricted portfolios can provide household benefits while funding activity elsewhere. Choosing a policy requires an explicit assessment of those tradeoffs rather than assuming the two objectives always coincide.

Europe's savings are a substantial financial resource, but their contribution to business expansion depends on the form and terms of intermediation. The official evidence supports both progress in integration and persistent weaknesses in particular equity, risk-capital and cross-border connections. Firm-survey evidence also cautions against treating every investment constraint as a shortage of external finance.

The controlled accounts show why the distinctions matter. A secondary-market purchase is different from primary finance; debt and equity allocate uncertain payoffs differently; fund commitments differ from deployable follow-on resources; and liquid investor claims must be reconciled with illiquid business assets. Each connection can be evaluated with observable transactions rather than inferred from the size of the savings stock.

The strongest standard of success is additional, economically valuable investment financed on sustainable terms, alongside suitable outcomes for savers and transparent allocation of risk. Meeting that standard requires evidence across the full chain from household portfolio choice to operating business assets. It turns the savings debate into a concrete analysis of capital allocation rather than a proposal to redirect an undifferentiated pool of money.