Europe Investment Report 2020 — The Fiscal Taboo Breaks

Europe's 2020 had two stories running on different timescales.

The short one was the pandemic response: an output contraction, an emergency monetary programme, and employment support schemes that were among the most extensive deployed anywhere.

The long one, and the one that will matter in a decade, was that the union agreed to issue debt jointly to fund a recovery programme, with the proceeds distributed partly as grants rather than loans.

The amount was not the point. The point was that it happened at all.

For the whole of the preceding decade, joint issuance had been treated as constitutionally and politically foreclosed. The objection was not primarily economic — it was that debt mutualisation transfers risk between member states in a way several electorates and at least one constitutional court had been understood to prohibit. Every previous crisis had been addressed by mechanisms explicitly designed to avoid it.

2020 demonstrated the constraint was a choice rather than a rule. That is the same reclassification the China Market Report 2015 describes for currency management and the Europe Investment Report 2016 describes for integration's reversibility — and it works the same way. An option that has never been exercised is priced at zero. Once exercised, it is a probability.

The market consequence was immediate and specific: redenomination risk compressed. A union that will act jointly in a crisis is a union whose members are less likely to leave it, and sovereign spreads narrowed accordingly.

What a precedent does to a price

The mechanism by which a first occurrence repriced assets is worth setting out, because it is the same one operating throughout this archive and it is consistently underestimated.

Before a first occurrence, a possible action is assessed on argument. Advocates say it could happen; opponents say the legal, political or institutional barriers make it impossible. In the absence of evidence, markets generally price the status quo, because the status quo has an unbroken record.

A first occurrence replaces argument with evidence. It does not prove the action will recur — circumstances were exceptional, and the design may have been deliberately structured as non-repeating. But it establishes that the barriers are surmountable, which is what was previously in dispute.

Three things reprice simultaneously:

• The probability of recurrence moves from approximately zero to something positive, and the size of the move depends on how exceptional the circumstances appeared.
• The tail scenarios narrow. A union capable of joint action in a crisis has a lower probability of disorderly outcomes, and the assets exposed to those outcomes reprice.
• The option value of the mechanism accrues to every member, including those who did not need it this time.

An action that has never been taken is priced at impossible. The first instance does not have to be repeated to be expensive — it only has to demonstrate that the barrier was political rather than absolute.

The symmetry with 2016 is worth noting explicitly. That year demonstrated integration was reversible, which widened tail risk and repriced assets adversely. 2020 demonstrated joint fiscal action was possible, which narrowed tail risk and repriced them favourably. The same mechanism, running in opposite directions, four years apart — which is a useful reminder that reclassification is not inherently bad news.

Why redenomination risk compressed

The specific channel from the fiscal decision to sovereign spreads is worth tracing, because "confidence improved" is a description rather than an explanation.

What redenomination risk is. The market's assessment of the probability that a member's debt ceases to be denominated in euros — which requires that member leaving the currency.

Why a member might leave. Historically, the scenario ran: a member faces an asymmetric shock, cannot devalue because it shares a currency, cannot borrow enough to absorb it because markets doubt its capacity, faces prolonged adjustment through wages and employment, and eventually finds exit less costly than continuation.

Where joint issuance intervenes. It breaks the second link. A member facing an asymmetric shock can access funding through a joint mechanism rather than solely through its own market access. The adjustment does not have to fall entirely on domestic wages and employment, which is what made the exit scenario plausible.

The consequences for spreads:

• The probability of the exit scenario falls, so the compensation demanded for it falls.
• The effect is largest where the risk was largest — the members whose spreads embedded the most redenomination probability.
• The compression is structural rather than cyclical to the extent the precedent is durable, which is the open question.

The honest caveat is that a single instance does not establish a permanent facility. The programme was designed with a defined size and horizon, and its repetition in a future crisis is not guaranteed. Markets priced a probability, not a certainty — and how much of the compression persists depends on whether the mechanism is used again.

Employment support and why the damage was less permanent

Europe's labour market response differed from several other advanced economies in a way that determined how permanent the damage was.

The mechanism. Rather than replacing income for people who had lost jobs, several European schemes subsidised employers to retain workers on reduced hours. The employment relationship was preserved; the hours and the wage bill were reduced.

Why preservation matters more than replacement:

• A preserved relationship restarts faster. The employer knows the worker's capability; the worker knows the role. Rehiring involves search, selection and training that a retained relationship does not.
• Firm-specific human capital survives. A worker's value to a particular employer includes accumulated knowledge of that firm's processes, customers and systems. That is destroyed by separation and preserved by retention.
• Long unemployment spells have persistent effects. Skills atrophy, and there is a well-documented penalty to earnings and employability that grows with the duration of a spell. Preventing the spell prevents the penalty.
• Business failures are largely irreversible. A business that closes rarely reopens, and its accumulated capability is lost.

The measurable consequence was that European unemployment rose far less than the output contraction implied, while hours worked fell sharply — the adjustment ran through hours rather than through headcount.

The trade-off is real and worth stating. Retention schemes preserve relationships including those in businesses that would not have survived anyway, which slows reallocation of labour toward more productive uses. The choice is between a faster recovery with preserved capability and a slower one with more reallocation, and reasonable economists differ on which is preferable over a longer horizon.

For investors the practical consequence was that European labour market data through this period is not comparable to markets that used income replacement instead. Unemployment rates understated the disruption; hours-worked data captured it. Comparing headline unemployment across the two systems produces a wrong conclusion about relative damage.

The venture market's independence, again

European venture funding held up considerably better than the macro environment implied, confirming a pattern the 2016 report identifies.

Why the venture market is partly insulated from the macro cycle:

• Its capital is committed in advance. A fund raised in 2019 deploys through 2020–2023 regardless of conditions. The deployment decision is separate from the fundraising decision by several years.
• Its investments are long-dated. A seed investment is underwritten on an outcome a decade out, which makes a single year's economic conditions largely irrelevant to the thesis.
• Its valuations are not anchored to public comparables at early stage, per the mechanism the US Venture Capital Report 2016 describes — so public market moves do not transmit directly.
• The discount rate moved favourably, per the Global Investment Outlook 2020, which supports long-duration assets specifically.

What did change:

• Sector composition shifted toward categories the conditions favoured — digital health, remote work infrastructure, e-commerce and logistics.
• Process moved remote, with the same consequences the US Venture Capital Report 2020 describes: faster decisions, wider geography, and the question of whether the in-person requirement had been informational or a throughput constraint.
• The distributed geography described in the 2017 report benefited. When every meeting is remote, being outside the largest hub costs less, which advantaged Europe's many mid-sized ecosystems.

The growth-stage gap was unaffected, as it has been by everything else. Imported US capital continued to lead large rounds, and the domestic capacity that would substitute still did not exist — which is the setup the 2023 report describes resolving badly.

Grants versus loans, and why the distinction mattered

A detail of the recovery instrument's design carried more weight than its headline size, and it is worth isolating because it is where the precedent's significance actually sits.

A loan-based mechanism advances funds that must be repaid. The recipient's debt rises. The issuing entity bears credit risk but no ultimate transfer. Every previous European crisis mechanism had been structured this way, precisely to avoid the objection that debt was being mutualised.

A grant-based mechanism transfers resources that are not repaid by the recipient. The obligation to service the jointly-issued debt falls on the union collectively, funded from the common budget to which all members contribute.

The second is fiscal transfer. The first is not. That is the distinction the preceding decade's design had been carefully constructed to preserve, and 2020's inclusion of a grant component crossed it.

Why markets cared about the composition rather than the total:

• A loan does not reduce a recipient's debt burden, so it does not change the sustainability analysis that redenomination risk depends on. It changes the timing of financing, not its ultimate incidence.
• A grant does. A member receiving resources it does not repay has a materially better debt trajectory, which directly reduces the probability of the scenario redenomination risk prices.
• The precedent value differs sharply. A loan facility can be characterised as liquidity support. A grant funded by joint borrowing cannot be characterised as anything other than shared fiscal capacity, however carefully the accompanying language is drafted.

The durability question sits here too. A mechanism designed as one-off, with a defined size and a sunset, is not a permanent facility. But the constitutional and political objection was to the category, not to the amount — and once the category has an instance, the objection to a second instance is weaker than the objection to the first was.

The size of the instrument was a policy decision. The inclusion of grants was a constitutional one, and markets priced the second.

What an allocator could act on

Price precedents as changes to a distribution. An action that has never been taken is priced at impossible. Its first occurrence establishes the barrier was political rather than absolute, which moves every scenario that depended on the barrier holding. This is the same mechanism as 2016's reversibility demonstration, running in the opposite direction.

Trace a spread's components before attributing a move. Euro area sovereign spreads embed credit risk and redenomination risk, which respond to different things. Joint fiscal capacity addresses the second directly and the first only indirectly, and knowing which moved is the difference between a durable repricing and a sentiment one.

Do not compare unemployment rates across different support regimes. Retention schemes and income replacement produce very different unemployment statistics from the same underlying disruption. Hours worked is the comparable series, and it is published free by Eurostat.

Treat venture funding as partly independent of the macro cycle. Committed capital deploys on its own schedule, and early-stage valuations have no public anchor. A macro view is a poor predictor of venture deployment, particularly in the first two years of a downturn.

Watch whether the precedent recurs. The compression in redenomination risk depends on the joint mechanism being repeatable, which a single exceptional-circumstances instance does not establish. The next crisis is the test, and until then the compression rests on an expectation rather than on a facility.

Note that the growth-stage gap survived another shock unchanged. It has now persisted through ecosystem maturation, a political shock and a pandemic. That is strong evidence it is structural, and it means any European growth model should assume the constraint rather than hoping conditions relieve it.

What 2020 established for Europe

• Joint issuance demonstrated the fiscal constraint was political rather than absolute, repricing a decade-old assumption.
• A precedent was shown to change a distribution, narrowing tail scenarios that had depended on the barrier holding — the 2016 mechanism running in reverse.
• Redenomination risk compressed through a specific and traceable channel rather than through generalised confidence.
• Employment relationship preservation produced less permanent damage than income replacement, and made cross-country unemployment comparisons misleading.
• The venture market's structural independence from the macro cycle was confirmed, along with the growth-stage gap's independence from everything.