The asset that was never safe


· 13 min read
This article is part of In conversation about sustainable finance & emission reduction systems, a new series by Diego Balverde. You’re reading volume 25 of the Collateral Crisis series. Here is volume 24
For decades, financial markets constructed an apparently stable hierarchy between safe assets and risky assets. Sovereign debt issued by selected economies, investment-grade bonds, prime real estate, regulated infrastructure, bank deposits, selected corporate guarantees and a large part of the assets used as collateral entered models in which risk could be quantified through familiar probabilities, historical volatility, ratings, liquidity assumptions and recovery values.
That architecture worked while several underlying relationships remained relatively stable: controlled inflation, negative equity-bond correlation, sufficient depth in sovereign markets, predictable refinancing conditions and an assumption that major shocks would normally occur separately rather than simultaneously. The current environment has weakened several of those relationships at the same time.
The IMF’s April 2026 Global Financial Stability Report describes elevated global financial stability risks following the Middle East conflict, alongside higher sovereign yields, inflation pressure, high public debt, refinancing risks, leverage among nonbank intermediaries, stretched valuations and multiple amplification channels capable of turning market stress into broader instability. The IMF also notes that more frequent supply shocks have weakened the traditional hedging relationship between equities and bonds, increasing the possibility of simultaneous selloffs. The asset expected to protect the portfolio can therefore fall alongside the assets it was supposed to protect against.
This is the starting point of the third cycle of The Collateral Crisis. The first twelve volumes showed how deterioration could move from the asset into collateral, banking and credit; the next twelve constructed a new economy around verified assets, efficiency, infrastructure performance, tokenisation, capital release, private credit and resilience. The next question is what happens when markets begin repricing the assets they still describe as safe.
The issue is no longer only which assets are deteriorating. It is how much of global financial safety depended on conditions that no longer exist. An asset is not safe because its contract says so, because its historical volatility was low or because regulation treats it favourably. It is safe while it retains liquidity, repayment capacity, stable cash generation, collateral acceptance and sufficient buyers exactly when everyone else also needs to sell. When any of those conditions disappears, safety can turn into discount.
A sovereign bond can carry extremely low default probability and still generate significant losses when duration is long and yields rise rapidly. Prime real estate can retain tenants while being repriced through higher capitalisation rates, lower residual value or unexpected CAPEX. Regulated infrastructure can produce apparently stable revenue while facing higher funding costs, regulatory delays or connection constraints. A corporate loan can remain current while the productive asset behind it loses competitiveness. A deposit can be legally repayable and become a source of institutional stress when depositors can withdraw digitally faster than the bank’s assets can become cash. Safety therefore does not sit entirely inside the asset. It depends on the interaction between time, liquidity, markets, funding and collective behaviour.
The Middle East conflict in 2026 demonstrated the point again. The IMF observed equity declines and higher sovereign yields following escalation as markets repriced inflation expectations and monetary policy because of rising energy prices. The ESRB similarly reported in March that systemic risks in the European Union had increased, pointing to sharp oil and gas movements, bond-market repricing and additional vulnerabilities related to critical infrastructure and cyber risk. Markets remained broadly orderly, but the lesson is not that financial architecture has become invulnerable. It is that one shock can simultaneously change inflation, rates, energy, bond valuation, equities and liquidity, weakening diversification relationships that many portfolios treated as permanent.
Safety therefore has to be understood as conditional. An asset must be safe against something: default, inflation, immediate liquidity needs, energy interruption, currency volatility or a particular investment horizon. A long-duration bond can be extraordinarily safe against insolvency and highly vulnerable to interest rates; cash can be safe against nominal volatility while losing purchasing power to inflation; real estate can protect against some monetary conditions while becoming deeply illiquid during forced sales. The word “safe” concealed for years the fact that different forms of safety are often mutually incompatible. The next market will have to stop classifying assets through absolute labels and begin analysing which risk each asset absorbs and which risk it introduces in exchange.
This principle is already entering Eurosystem infrastructure. The ECB began applying climate factors to selected corporate bonds used as collateral in June 2026 and decided in July to extend the methodology to certain credit claims against non-financial corporations. The logic is exactly the one required for this new cycle: even after conventional haircuts have been applied, a technological, regulatory, behavioural or macroeconomic shock can unexpectedly reduce collateral value precisely when the central bank may need to liquidate it. Recognised collateral value therefore no longer depends entirely on what happened historically. It increasingly incorporates uncertainty regarding what can happen over the remaining life of the asset.
For much of modern credit analysis, the central question was whether the borrower would repay. In the emerging environment, an asset can generate major losses without any default taking place. Repricing can occur because the risk-free rate changes, term premiums rise, liquidity deteriorates, inflation reduces the real value of future cash flows, capital becomes more expensive, an industry loses competitiveness or an asset previously assumed to be easy to sell suddenly finds few buyers. This is a fundamental change for collateral management because a bank can own legally enforceable and economically solvent collateral whose financing capacity still declines because market valuation changes abruptly.
The IMF identifies precisely this combination inside sovereign markets. High public debt and greater reliance on shorter-term issuance increase rollover risk, while an inflation shock can push yields higher and revive the sovereign-bank nexus. Sovereign bonds remain central to liquidity buffers, collateral systems, reserves and benchmarks for pricing almost every other financial asset, but the more systemic their role becomes, the wider the consequences of material repricing. Higher sovereign yields do not affect governments alone. They change discount rates, bank valuations, mortgage pricing, corporate borrowing, project finance, insurance portfolios and infrastructure values. The conventional safe asset acts like financial gravity: when it moves, the price of almost everything around it moves as well.
The dynamic becomes even more important through nonbank financial institutions. Hedge funds, leveraged ETFs, investment funds and other intermediaries can rely on leverage, derivatives and secured financing; a correction in assets assumed to be liquid can trigger margin calls, forced sales and additional downward price pressure. No default is required to begin the crisis. A sufficiently large price movement can be enough. When that movement creates cash obligations, selling stops being a valuation decision and becomes a contractual necessity. The system shifts from fundamental repricing into mechanical repricing: an asset falls because one institution must sell, the sale creates another decline and the new decline forces the next institution to sell.
The opportunity is to recognise this distinction before portfolios and capital structures are built. Models based only on expected credit loss can understate the importance of forced-sale loss. For industrial and infrastructure companies, the analysis should not stop at whether debt can be serviced under a base case and conventional stress scenario. It should also ask what happens when refinancing arrives during a period of higher sovereign yields, lower collateral values, tighter bank loan-to-value ratios and higher required returns from private markets. An excellent operating asset can still become financially weak when its debt structure assumes permanently benign refinancing conditions.
There is a second illusion this cycle needs to dismantle: the assumption that physical assets are inherently safe because they possess material utility. Ports, factories, buildings, batteries, water networks and energy infrastructure have real value because they physically exist and provide services, but their financial value depends on those services remaining economically competitive. A technically excellent factory can lose value if its energy becomes too expensive; a strategically located port can suffer when the trade route feeding it changes; prime property can require substantial unanticipated CAPEX; BESS can lose economic capacity through degradation; a data centre can own servers and buildings while lacking enough electricity to monetise them; an industrial plant can preserve every machine while losing production because one critical component is unavailable.
A physical asset must therefore be analysed as a system of dependencies rather than simply a collection of tangible goods. Balance sheets record machinery, land, buildings and contracts; DOIX needs to map what allows those components to produce: energy, availability, maintenance, routes, water, critical inputs, software, suppliers, contracts, utilisation and cash flow. Physical security without operating security does not necessarily produce financial security.
The same principle changes how existing infrastructure should be valued. Electricity grids can be strategically indispensable while requiring major investment because congestion prevents new connections. Ports can own valuable land while leaving capacity idle through poor operations. Water systems can contain substantial fixed assets and still lose valuable resources before they become revenue. In each case tangible assets exist, but part of their economic value remains trapped. The previous market often associated tangibility with safety. The next market will increasingly distinguish between tangible assets and productive assets.
This also creates significant positive repricing opportunities. Not every repricing is a decline. When inefficient infrastructure can increase utilisation, reduce energy use, decrease downtime or recover losses using CAPEX that is small relative to the value recovered, asset quality can improve materially. This is a crucial difference from the traditional collateral-risk narrative. Assets do not have to accept passively the haircut markets assign them. They can change some of the variables that justify that haircut.
The next financial market does not need more assets described as safe. It needs assets capable of demonstrating why they deserve that description. This requires a shift from presumed safety toward earned safety. An asset earns safety when it reduces dependencies capable of destroying cash flow, maintains sufficient liquidity, structures debt around its genuine economic life, preserves alternatives, produces reliable information and demonstrates the ability to continue operating when external conditions change. Safety stops being a marketing characteristic and becomes an operating and financial outcome.
BalGreen and DOIX should evolve in this direction. DOIX can establish an Asset Safety Baseline that measures not only normal performance but sources of fragility: energy sensitivity, refinancing requirements, supplier concentration, critical-input exposure, downtime, liquidity buffers, contractual concentration, residual CAPEX, geopolitical dependencies and collateral mobility. BalGreen can determine which vulnerabilities are economically reducible and which must instead be accepted, hedged or transferred; BalGreen Capital can examine whether the financing structure unnecessarily amplifies those vulnerabilities and design debt consistent with real asset risk. DOIX then verifies how much sensitivity has been reduced following intervention.
The distinction from a credit rating is fundamental. We are not independently rating whether the issuer will default. We are identifying what needs to go wrong for the asset to stop behaving as the financial model expected. In one factory that may be energy; in a port, routes and throughput; in a data centre, electricity and semiconductor availability; in BESS, degradation and revenue concentration; in a highly leveraged company, refinancing. Once the fragility mechanism has been identified, CAPEX can be directed toward whichever variable protects the greatest amount of cash for each euro invested.
This creates market opportunities as well. Assets penalised because markets cannot distinguish their quality can use verification to reduce information asymmetry; infrastructure portfolios with homogeneous operating data can become more attractive to insurers and pension capital; private credit can finance improvements to existing assets; banks can recycle capital using securitisation and SRT; public guarantees can be concentrated where they genuinely mobilise additional private investment. The objective is not to invent another “safe asset” label. It is to construct a system in which capital can identify more precisely which assets deserve a safety premium and which remain dependent on excessively fragile conditions.
The Asset That Was Never Safe does not argue that low-risk assets do not exist. It makes a more useful argument: no asset is safe against every shock simultaneously, and relevant safety depends on the financial purpose. A central bank needs collateral whose value can be protected under stress; a pension fund needs duration and long-run repayment capacity; a treasury desk needs immediate liquidity; a bank needs financeable assets able to preserve value; a corporation needs infrastructure capable of generating enough cash to service debt. Treating all these objectives as one universal category of safety produces mistakes because each answers a different risk.
The economy of 2026 is making those distinctions visible. The IMF identifies pressure in sovereign markets, high public debt, rising rollover risks, nonbank leverage and a less reliable equity-bond relationship during supply shocks. The ESRB continues to assess European systemic risks as elevated. The ECB is simultaneously adapting its collateral framework to recognise that forward-looking vulnerabilities can affect value even after conventional haircuts are applied. These institutions are not announcing the end of safe assets. They are recognising that the concept of safety must evolve as the sources of risk evolve.
For corporations, the conclusion is concrete. CFOs should not assume that an asset is safe because it has operated for decades, has strong collateral or belongs to a defensive sector. They should know how much cash disappears when rates, energy or insurance rise; how much refinancing is required over the next five years; what proportion of production depends on one critical input; how much residual value depends on maintenance; how long the business can operate without access to new debt markets; and which vulnerabilities can be removed before markets price them. Safety becomes a management discipline.
For investors, the opposite opportunity appears. Some of the strongest returns in the next cycle may come from identifying assets whose fragility is correctable. An asset discounted because it consumes too much, operates poorly or carries a defective capital structure can represent an exceptional opportunity when the source of the discount can be removed at a cost below the value recovered. Investing then becomes more than purchasing cheap risk. It can mean buying an asset with a haircut, eliminating the source of the haircut and capturing the repricing.
For years, financial systems treated selected assets as fixed points around which the rest of the architecture could be constructed. But no price is truly fixed when inflation, interest rates, energy, geopolitics, technology, liquidity and investor behaviour change simultaneously. The IMF demonstrated in 2026 how geopolitical shocks can raise energy prices, change inflation expectations, push yields higher and weaken traditional diversification relationships; the ESRB continues to assess European risk as elevated; and the ECB has begun incorporating future uncertainties directly into its own collateral valuation framework. Financial architecture is acknowledging something markets will have to learn quickly: safety does not mean absence of risk. It means understanding which risk exists, who can absorb it and how much value remains when that risk materialises.
For BalGreen, this creates a decisive evolution. It is no longer enough to demonstrate that an improved asset is efficient. We need to demonstrate why its cash flow has become harder to destroy. DOIX identifies the dependencies supporting and threatening performance; BalGreen removes economically correctable vulnerabilities; DOIX verifies the new resilience; BalGreen Capital aligns debt, liquidity and duration with actual performance; and markets can revalue the asset using evidence that previously did not exist. Where a discount arises from a correctable weakness, the haircut stops being only a loss. It becomes an opportunity for value creation.
The safe asset of the past asked markets for trust. The safe asset of the future will have to demonstrate continuously why it deserves to retain that trust. That is where The Repricing of Collateral begins.
The asset was never safe. The market had simply not yet repriced its fragility.
illuminem Voices is a democratic space presenting the opinions of leading Sustainability Thought Leaders, their views do not necessarily represent those of illuminem.
The world needs sustainability knowledge. At illuminem, no interest group or shareholder can influence our work. Thank you for supporting our mission to make high-quality and independent sustainability information free for all. Every contribution helps. Thank you for donating today.
illuminem

Climate Change · Public Governance
Shuwei Zhang

Coal · Energy
Steven W. Pearce

Climate Change · Adaptation
GOV.UK

Sustainable Investment · Public Governance
The Guardian

Sustainable Finance · Public Governance
Euronews

Public Governance · Climate Change