Financely Weekly Finance Brief | August 22, 2026

Guarantees Are Becoming Capital, Grid Risk Is Repricing Renewables, and Trade Finance Demand Is Rising

This week's most important financing developments point toward the same conclusion: capital is available, but lenders increasingly want contractual protection around the risks they cannot comfortably underwrite.

The clearest example came from AI infrastructure. Nvidia agreed to provide up to USD 105 billion of support around OpenAI's long-term lease of a huge Ohio data center campus. The financing is expected to combine equity with project finance loans and potentially bonds. Nvidia's guarantee is specifically designed to support portions of lease and power obligations and a minimum residual value rather than simply guaranteeing the entire construction cost.

In India, renewable developers are confronting the opposite problem. The generation assets exist and power demand exists, but inadequate transmission has prevented some projects from delivering electricity to the grid. The government is now considering long-tenor, low-cost loans for affected projects after developers reportedly incurred roughly USD 470 million of losses.

Trade finance tells another version of the same story. Bank earnings show strong demand for trade lending, but major institutions are also becoming more selective about capital-intensive exposures. Development institutions are stepping into the resulting gaps through guarantees, cash advances and risk-sharing facilities.

Acquisition finance is becoming more selective as well. A possible Silver Lake buyout of Workday could require around USD 18 billion of debt and is shaping up as a major test of whether lenders still consider recurring software revenue sufficient protection when AI can disrupt the borrower's underlying product economics.

Nvidia's USD 105 Billion Guarantee Shows Where Credit Enhancement Is Going

The largest project-finance story this week may also be one of the clearest demonstrations of modern credit enhancement.

OpenAI has entered into a 20-year lease for a data center being developed by SoftBank-owned SB Energy in Pike County, Ohio. The campus can eventually reach approximately 8 GW, with the first 800 MW expected online in 2028. Nvidia is investing USD 1.5 billion into SB Energy and has agreed to provide guarantees of up to USD 105 billion.

The important part is the structure of the guarantee.

Nvidia is not simply writing a blank check for construction. The support covers portions of OpenAI's lease and power obligations and provides protection around a minimum value for the site. If OpenAI defaults, Nvidia's exposure relates to the difference between the guaranteed value and what the asset owner can recover through reletting or sale.

That support can materially improve the credit case for lenders. Project debt can then be sized against a combination of the real estate and infrastructure value, long-term lease economics and an additional corporate backstop from one of the world's largest technology companies.

The planned financing is expected to begin with equity and then move into debt that could include project finance loans and bonds.

This is what effective credit enhancement looks like. The guarantor does not need to assume every risk. It needs to support the obligation creating the lender's principal uncertainty.

Financely applies the same principle when structuring credit enhancement for infrastructure transactions. A guarantee is useful when it changes the lender's downside case, increases debt capacity, extends tenor or brings a previously unavailable capital provider into the transaction.

Broadcom Is Exploring Another Massive SPV Financing

Nvidia was not the only technology company moving toward structured infrastructure finance this week.

Broadcom is reportedly discussing more than USD 60 billion of new debt connected with AI chip infrastructure for Anthropic and other companies. The structure under discussion could contain a roughly USD 30 billion junior tranche and a senior secured tranche of approximately USD 60 billion to USD 70 billion, potentially taking the overall financing to as much as USD 100 billion.

Broadcom would reportedly guarantee part of the senior secured debt, while Blackstone and Apollo are among the institutions discussing participation. The debt would be issued through a special purpose vehicle.

This matters beyond AI.

The financing resembles the structures infrastructure sponsors have used for years: isolate the asset or contracted activity in an SPV, introduce real sponsor or strategic capital, place senior debt against the strongest identifiable cash flow, and use subordinated capital or credit enhancement around risks the senior lender does not want.

AI Debt Is Starting to Meet the Limits of Traditional Bond Portfolios

The next issue for AI infrastructure may no longer be whether debt is available. It may be where that debt can sit.

AI-related bond issuance by hyperscalers has reportedly reached about USD 220 billion in 2026 through early August, compared with only USD 12.5 billion during the comparable period last year.

Investors are beginning to require larger concessions to absorb the volume. Technology-company spreads have widened to roughly 89 basis points over Treasuries, around nine basis points wider than the investment-grade market overall.

This is not necessarily a deterioration in the underlying credit quality of companies such as Amazon or Alphabet. It is partly a supply problem. Insurance companies, pension funds and other large investors have portfolio concentration limits. They cannot continuously absorb tens of billions of additional obligations from the same handful of issuers regardless of credit quality.

That creates a strong incentive to move some financing away from parent-company balance sheets and into project-level debt, asset-backed facilities, SPVs, equipment financing, leases and other structures where lenders obtain direct exposure to a specific pool of infrastructure and contracted cash flows.

Sponsors looking at large infrastructure transactions can review Financely's project finance debt and capital advisory services for construction, operating-asset and recapitalization mandates.

India's Renewable Problem Is No Longer Generation Capacity. It Is Deliverability.

India's renewable sector provided an unusually clear example this week of why infrastructure dependencies need to be included in lender underwriting.

India now has approximately 162 GW of solar capacity, but transmission expansion has struggled to keep pace. Developers in renewable-heavy states including Rajasthan and Gujarat have reportedly lost approximately 45 billion rupees, or about USD 470 million, since February 2025 because electricity could not be moved through the grid.

Between April and June alone, approximately 8,133 GWh of solar generation was curtailed, equivalent to about 14% of output during the period.

India is now considering low-interest loans with seven- to eight-year tenors for qualifying affected projects.

From a project-finance perspective, the important issue is DSCR. Debt is sized against future generation and revenue. If a technically functioning solar plant can regularly produce electricity that cannot be evacuated through the network, forecast generation is not equivalent to financeable revenue.

Lenders will respond by lowering their recognized generation case, increasing required coverage, reducing leverage or charging more.

Concessional loans can repair part of the damage by reducing annual debt service. They do not resolve the underlying transmission constraint. Future financings will need greater scrutiny of permanent grid infrastructure, curtailment provisions, compensation mechanisms and the contractual allocation of transmission-delay risk.

Public Capital Is Also Moving Into Critical Minerals

The United States announced another significant example of public capital being used to move projects toward commercial viability.

The Department of Energy is awarding USD 500 million of grants to seven domestic lithium, cobalt, recycling and battery-material projects. Lilac Solutions, Jervois and Nth Cycle are each slated to receive USD 100 million.

For a project sponsor, grants of this scale do more than improve headline returns. They reduce the amount of private capital that has to earn a commercial return.

If a USD 500 million processing project receives USD 100 million of committed grant funding, debt and equity investors are effectively being asked to finance a USD 400 million residual requirement rather than the original USD 500 million, subject to the grant's eligibility and disbursement conditions.

That can materially change project leverage, DSCR and sponsor equity requirements. The project still needs credible engineering, supply, offtake and execution. Public support is most effective when it addresses the portion of the economics that prevents an otherwise viable asset from attracting private capital.

Trade Finance Demand Is Strong, but Banks Are Optimizing Their Balance Sheets

Trade finance demand remained strong in the first half of 2026 according to a review of bank earnings published this week.

Several European and Singaporean lenders reported higher trade-loan volumes. The Middle East was more mixed after months of severe disruption to shipping and supply chains.

There is an important caveat for borrowers. Deutsche Bank and Standard Chartered both referred to rationalizing or optimizing capital-intensive exposures within their trade portfolios.

Strong demand therefore does not mean indiscriminate credit expansion.

Banks care about how much regulatory and economic capital a transaction consumes relative to the revenue it produces. Structures supported by receivables, liquid inventory, documentary credits, insurance, guarantees or strong transaction controls can compete more effectively for balance sheet than unsecured working-capital requests with limited repayment visibility.

Financely's structured trade and commodity finance mandates are built around this principle. The financing request has to show the bank where the money goes, what asset or contractual right exists during the exposure period and how the facility self-liquidates.

EBRD Uses Guarantees to Expand Trade Finance in Iraq

One of the better examples of trade-finance credit enhancement this week came from Iraq.

Al Mansour Bank for Investment secured a EUR 65 million trade finance facility from the European Bank for Reconstruction and Development through its Trade Facilitation Programme.

EBRD will provide guarantees covering defined political and commercial payment risks and can also provide cash advances. The facility is intended to increase trade finance available to businesses in agriculture, construction, manufacturing and consumer goods while expanding the Iraqi bank's correspondent relationships.

This is an important model for emerging-market trade finance.

The local bank retains its customer relationships and origination capability. The multilateral institution improves the credit profile of the cross-border obligation. Confirming banks can then participate against a different risk than they would face if they had to take the Iraqi bank and country exposure entirely on an uncovered basis.

Guarantees Could Also Unlock India's International Factoring Market

India's factoring market is growing rapidly, but international factoring remains exceptionally small relative to the country's trade volumes.

India recorded approximately EUR 42 billion of domestic factoring turnover in 2025 but only EUR 1.5 billion internationally, despite annual goods exports of roughly USD 860 billion.

Industry participants meeting in Gift City this month argued that DFI guarantees and greater bank participation could materially increase international factoring capacity.

The Trade and Development Bank provided a useful data point. Around USD 150 million of guarantees issued over approximately five years reportedly facilitated USD 1.8 billion of trade.

That is precisely why guarantees matter. The objective is not for the guarantor to fund the entire commercial transaction. A relatively small amount of risk-bearing capacity can allow banks and factoring companies to finance a substantially larger volume of receivables.

Gift City's recent recognition of qualifying credit insurance and guarantees for capital-relief purposes also shows another mechanism. Credit enhancement has value not only because it reduces expected loss. It can reduce the regulatory capital consumed by the exposure and make the transaction economically viable for a lender.

Hormuz Is Becoming a Working-Capital and Insurance Problem

The International Chamber of Commerce published a significant brief this week on the economic consequences of the Strait of Hormuz disruption.

Before the closure, Persian Gulf producers dependent on the Strait accounted for approximately 40% of global urea trade. Nearly 20% of global oil supply and more than 20% of LNG trade also passed through the route.

The World Bank fertilizer-price index increased 44% between February and April, while its urea benchmark rose 82%. ICC estimates that approximately 3.7 million tonnes of urea demand was destroyed or deferred between April and June as buyers postponed purchasing or became unable to afford normal volumes.

The trade-finance implications are substantial.

Longer routes increase transit time and therefore financing tenor. Higher marine insurance increases landed cost. Sanctions uncertainty can eliminate otherwise workable shipping routes. Expensive fertilizer increases the working capital required for the same physical volume. Importers can then need larger facilities precisely when banks are becoming more cautious about route and counterparty risk.

ICC is calling for measures including a contingent public-private insurance facility for eligible voyages and coordinated financing for importers, distributors and farmers over the next two to three planting cycles. That is effectively a proposal to combine insurance capacity with working-capital credit enhancement so essential trade continues even when the private market alone cannot absorb the shock.

Acquisition Finance Is Starting to Price AI Risk Explicitly

A potential Silver Lake acquisition of Workday could become one of the year's most important tests for leveraged acquisition debt.

Reuters Breakingviews estimates the transaction could sustain around USD 18 billion of debt if lenders continue to treat Workday as a high-quality recurring-revenue software borrower.

The problem is that software no longer receives automatic credit for subscription revenue.

Recent refinancings of software companies have reportedly priced 50 to 150 basis points wider as lenders account for the possibility that AI reduces switching costs, allows customers to build internal alternatives or undermines established application-software businesses.

Workday has an important defense: subscriber gross retention of approximately 97%. Market data cited by Reuters suggests lenders have historically been comfortable with leverage slightly above 5x EBITDA for software companies maintaining retention above roughly 95%.

This tells acquisition sponsors something useful.

Financing an acquisition is increasingly about demonstrating durability rather than presenting historical EBITDA alone. A lender needs to understand customer retention, concentration, recurring revenue, competitive threats, capex requirements and what could make the acquired company's cash flow materially different five years after closing.

The highest leverage will continue to go to businesses where the buyer can show both strong current cash generation and credible protection against structural disruption.

What Borrowers Should Take From This Week

The market is not short of money.

AI companies are discussing financing packages measured in tens of billions. Banks report strong trade-finance demand. Development institutions are increasing guarantee capacity. Governments are providing grants and concessional capital to strategic infrastructure.

The constraint is increasingly the allocation of risk.

A data center needs credible tenant and power support. A renewable project needs reliable transmission and a bankable curtailment regime. A commodity trader needs a transaction structure that protects the lender through the entire purchase, shipment and repayment cycle. An acquisition borrower needs to demonstrate that EBITDA remains durable after the deal closes.

Sponsors who approach financing as a search for "a lender" tend to miss this distinction.

Institutional financing works better when the capital stack is designed around the risks each investor is equipped to hold. Senior lenders take the strongest part of the cash flow. Junior capital absorbs more volatility. Guarantees isolate defined risks. Public capital closes economically justified gaps. Equity remains beneath the entire structure.