The market is in for another rollercoaster ride this year on the policy front. Last year set records as developers rushed to start construction of projects ahead of a year-end deadline to avoid new FEOC limits on the amount of Chinese equipment that can be used in projects. Short-term interest rates are expected to fall after a new Federal Reserve chairman takes office in May. Demand for capital will remain elevated due to the data center build out and the need for more generating and transmission capacity to meet rising demand for electricity.

Four leading tax equity investors and lenders talked about what to expect this year for tax equity, tax credit sales and debt during a live podcast in mid-January. The panelists are Jack Cargas, head of originations on the tax equity desk at Bank of America, Rubiao Song, managing director and head of energy investments for JPMorgan, Ralph Cho, co-CEO of Apterra Infrastructure Capital, the private credit arm of Apollo Asset Management, and Beth Waters, managing director for project finance in the Americas for Japanese bank MUFG.

Tax Equity

MR. MARTIN: We always talk first about tax equity volume and compare it to past years, but comparisons have become harder lately because of the proliferation in deal structures and market fragmentation.

Tax credits account for 30% to 70% of the cost of US renewable energy projects. There are at least five strategies for monetizing them. Partnership flips are one. That is where a bank or other tax equity investor makes an investment that it expects to be repaid partly in tax benefits and partly in cash. There are hybrid deals where the tax equity partnership plans to sell most of the tax credits to another company. There are preferred equity partnerships with cash investors where the partnership sells all the tax credits. There are straight tax credit sales. Finally, there are sale-leasebacks, which were an earlier form of tax equity that is making a comeback.

Rubiao Song, break it down for us. What were tax equity and tax credit sales volumes in 2025, and how did they compare to 2024?

MR. SONG: As you said, deal volume has become a lot harder to track, but I think we have good visibility into the traditional technology sectors. We counted $35 billion in transaction volume in 2025 in solar, wind and batteries. That includes traditional tax equity, hybrid structures, preferred equity partnerships and direct tax credit sales.

We count another $10 to $15 billion in 2025 in sales of tax credits on other technologies, like 45X credits for manufacturers of solar, wind and storage equipment, 45U credits for generating nuclear electricity and 45Z credits for producing clean transportation fuels.

That puts the total market last year at $45 to $50 billion.

MR. MARTIN: Do you have a sense for how large an increase that was over 2024?

MR. SONG: We think that is about a 10% increase.

MR. MARTIN: Jack Cargas, what mix of solar, wind and storage projects are you seeing?

MR. CARGAS: The breakdown last year was about a third wind and two thirds solar and solar-plus-storage.

Another interesting question is how much of the tax equity market is traditional tax equity where investors plan to keep the tax credits on their own books and how much is hybrid tax equity, which is a tax equity partnership with the option to sell, or a plan from the start to sell, all or part of the tax credits.

Traditional tax equity structures where there is no intended sale of tax credits were as much as 30% of the tax equity market in 2024. We think they were a smaller percentage in 2025.

MR. MARTIN: Perhaps the surprising thing was a third of projects were wind given the efforts by the Trump administration to make it harder to build such projects.

Also, I suspect most people think virtually all tax equity partnerships today are hybrids.

The tax credit sale market seems interested in all types of tax credits, but not all tax credits lend themselves to tax equity structures. What tax credits besides the traditional credits on power projects do you think are getting the most traction in the tax equity market?

MR. CARGAS: Credits that are structurally similar to investment tax credits and production tax credits on solar, wind and storage work with tax equity partnerships. Examples are 45V credits for producing clean hydrogen and 45Q credits for carbon capture.

MR. MARTIN: Did you see any 45V deals for making clean hydrogen in 2025?

MR. CARGAS: We are interested in them, but we did not see any transactions.

MR. MARTIN: Rubiao Song, same answer on 45V?

MR. SONG: No deals last year.

Adding to what Jack said about the breakdown between wind and solar, we counted 7,000 megawatts of wind projects last year and 63,000 megawatts of solar and storage projects. Wind projects are still drawing interest from tax equity investors who like to maintain some diversification in their portfolios.

Year Ahead

MR. MARTIN: How was 2025 for you, and what do you expect this year?

MR. SONG: Last year was a very busy year. We closed on more than $7 billion in transactions. That momentum is continuing into 2026. We see a very strong pipeline for 2026 and beyond.

MR. MARTIN: "Beyond" meaning 2027?

MR. SONG: Yes. We are already looking at 2027 financings. These projects take a long time to build. Tax equity is committing 18 months in advance.

MR. MARTIN: Jack Cargas, how was 2025 for you, and what do you expect in 2026?

MR. CARGAS: We have three different groups at Bank of America focused on the renewable energy finance market. The one I am in is the renewable energy finance group, which is basically tax equity for utility-scale projects and residential solar. We also have a global infrastructure and sustainable finance group, which is a market leader in the development and execution of bespoke structures across the world and across these platforms. Then we have a highly capable tax credit transfer desk, which we set up directly after the passage of the Inflation Reduction Act, which is also a market leader in that particular specialty.

The reason I mentioned them all is because it was a busy year for all three of our groups. All three groups saw a rush to start construction and close financings at the end of the year. Some market participants, including some of our own, reported that 2025 was their busiest year ever in terms of projects and dollars financed.

Some sponsors and third-party service providers, such as the engineering firms, law firms and financial advisors, were severely stretched. We expect more of the same in 2026, including another rush to start construction by July 4 in order to lock in a four-year construction time frame.

MR. MARTIN: Rubiao Song, most deals in the tax equity market last year involved "legacy" tax credits on projects that were under construction by the end of 2024. Projects in which "technology-neutral" tax credits will be claimed were just starting to come to market late in the year.

Will the market roll forward into technology-neutral financings with the same momentum or have we now turned into a road with a lot of speed bumps on it?

MR. SONG: Speaking from 20+ years of experience in this industry, renewable energy financings have always been a bumpy road. Whether the technology-neutral credit financings can be less bumpy depends on what the Treasury says in the initial guidance on FEOC that is expected any day now.

MR. MARTIN: FEOC stands for "foreign entity of concern." We will come back to that.

Tax equity has accounted historically for about 35% of the capital stack, plus or minus 5%, in investment tax credit deals and 65%, plus or minus 10%, in production tax credit deals. What are the percentages today?

MR. SONG: Tax equity in ITC deals tends today to be around 45% of the capital stack. Most projects today qualify for at least one bonus tax credit and some qualify for two, bringing the ITC on the projects to 40% or 50%. Nearly 80% of the projects we are evaluating today choose an ITC over PTCs.

For PTC projects, we are seeing tax equity as a percentage of the capital stack dropping because of spiraling project costs due to inflation and tariffs. The tax equity percentage of the capital stack in PTC projects is converging with the percentage in ITC projects.

Construction Start

MR. MARTIN: There was a rush to start construction of all types of projects by the end of 2025 to avoid new FEOC limits on the amount of Chinese equipment that can be used in projects. There will be another rush to start construction of wind and solar projects by July 4 this year to lock in four years to build. Otherwise, wind and solar projects must be finished by the end of 2027 to qualify for federal tax credits.

Jack, most developers have been relying on work on main power transformers to start construction. Do you have any benchmarks for what you need to see to treat such work as the start of construction?

MR. CARGAS: Most market participants would say that building the main power transformer qualifies as "physical work of a significant nature," which is the term of art. Whatever work is done at a factory must be on specialized, custom-engineered, project-specific equipment. It cannot be work on standard stock inventory.

The physical fabrication of the transformer must have commenced after an enforceable binding written contract is in place. Those are some of the benchmarks.

MR. MARTIN: Do you have any standard for how much work you want completed before the deadline? We see a wide range of fact patterns, from a conservator tank, to two radiators, to a conservator tank and all the radiators to that plus the transformer core. There are different dollar amounts and different numbers of labor hours.

Where do you draw lines?

MR. CARGAS: As you know, there are no bright lines in the law, and so we don't draw bright lines. It is a facts-and-circumstances test. In general, the more work completed, the better.

MR. MARTIN: The law firms are under pressure to bless factory work on such things as medium-voltage transformers, inverters, inverter skids and trackers as a start of construction. Are you treating work on any such items currently as a start of construction?

MR. CARGAS: We have not had many such requests. The same rules about non-inventory items that I mentioned apply. Some of the items can be viewed as stock or inventory items and, therefore, difficult to use as qualifying assets.

MR. MARTIN: Many developers need more than four years to finish projects, particularly after the Trump administration froze federal approvals for wind and solar projects. How open are you to proof of "continuous efforts" to buy more time? To be clear, a showing that the developer made continuous efforts to advance the project buys more time only for solar and wind projects that were under construction by September 1, 2025. Solar and wind projects that started after that must show continuous actual construction to be allowed more time.

MR. CARGAS: We are open to it.

Supply Imbalance

MR. MARTIN: Rubiao Song, people listen to this call hoping to get a sense of what to assume in financial models about the cost of capital. What is your advice about what to put in models for the cost of tax equity this year?

MR. SONG: There is no single number. There are many variables. The first question a developer must answer is whether to choose an ITC or PTCs on the project. Different sets of investors are interested in ITCs versus PTCs.

There is a supply-demand imbalance in the ITC monetization market versus PTC market. There are more projects with ITCs seeking tax equity. Roughly 80% of projects we are seeing today are ITC projects. Investors must have a lot of current-year tax capacity to absorb the full tax credit in one year.

That is one reason for the huge growth in the hybrid tax equity or preferred equity partnership market where 90% or 99% of the tax credits are sold in the transfer market.

MR. MARTIN: This is putting more strain on tax capacity.

If people have a good sense of what tax equity cost last year, do you think there will be much change in cost this year?

MR. SONG: I think there will be little change on the PTC monetization front, as long as the long-term interest rate remains stable. We see good supply-demand balance in the PTC monetization market.

But there is a noticeable imbalance in the ITC market that will probably get worse this year and put downward pressure on the price tax credit buyers are willing to pay for ITCs.

MR. MARTIN: Jack Cargas, Sunnova, a major residential rooftop solar developer, filed last year for bankruptcy. Has that changed how you look at residential solar deals?

MR. CARGAS: The residential solar space has had its challenges in recent years. Those challenges have given us an opportunity to stress test our structure. We have been in that market since 2013. We have learned a few things recently, but we remain in the market today and will continue to do business in that market.

This is project finance, and these transactions are structured around the projects. There is no substitute for a top sponsor.

When there are difficulties, if the project financing has been properly structured, it can survive even after a sponsor bankruptcy.

MR. MARTIN: I noted Jefferies this morning came out with a "buy" recommendation for Sunrun stock. It is expecting Sunrun to have a strong year.

Rubiao Song, there has been a push to repower existing wind farms. What are your hot buttons in repowerings?

MR. SONG: We have financed a significant number of repowered wind projects over the last 10 years. We focus due diligence on the technical engineering, the technology being used in the retrofit, the remaining useful life of the project and the business case for doing a repowering. We look for a nontax reason to repower, such as an increase in electricity output or a reduction in the O&M costs.

And then the critical question is whether the work is extensive enough for the sponsor to be considered to have built a new wind farm under the so-called 80/20 test. We like to see a solid appraisal confirming satisfaction of the 80/20 test.

FEOC

MR. MARTIN: Jack Cargas, new FEOC rules took effect on January 1 this year for projects on which technology-neutral tax credits will be claimed. They limit the amount of Chinese equipment that can be used in such projects, and they bar equipment supply and O&M contracts with Chinese-backed companies that give such companies effective control over key aspects of such projects. Congress provided a list of 13 contract clauses that it believes are forms of effective control.

This seems like an area where tax equity investors will just push the risk off on developers. Is it more complicated than that?

MR. CARGAS: It probably is more complicated than that. It is true that there is a risk allocation question, but we never want to invest into an indemnity claim.

We would all benefit from guidance from the Treasury and the IRS about FEOC. We know that Congress and the regulators are well aware of that, not least because the industry organizations have been making that point in Washington. We are also conscious of the fact that guidance is important not only to the traditional renewables industries, but also to other technologies, such as advanced domestic manufacturing, battery storage, geothermal, hydropower and nuclear power.

It may be simple to say the risk can be pushed to the sponsor, but it will be much better for our sponsors if we can all have a clearer view of what that risk is.

PAM Accounting

MR. MARTIN: Some tax equity investors have moved to PAM accounting -- PAM stands for "proportional amortization method" -- instead of the more traditional HLBV or hypothetical liquidation at book value accounting for tracking earnings from tax equity investments. JPMorgan is using PAM accounting if I am not mistaken. What changes should a developer expect when there is such a shift?

MR. SONG: The Financial Accounting Standards Board authorized the use of PAM accounting by energy credit investors probably two years ago. We are seeing many tax equity investors adopt this method. It is simpler than HLBV to implement.

Developers probably will not notice much change.

One way to qualify for PAM accounting is to limit the cash distributions to the investor. The cash distributions to the investor must be less than 10% of the sum of the tax credits, other tax benefits and cash in present-value terms.

The other potential change is also in favor of developers. The investor cannot be considered to have significant influence over project operations.

MR. MARTIN: So it is a net positive for developers. Potentially less cash going to the tax equity investor. Less say for the tax equity investor over operation of the project.

Let's conclude the tax equity segment with this question. What other new trends are you seeing as we enter 2026?

MR. CARGAS: We are seeing more variations in transaction structures. There used to be something close to a vanilla tax equity structure. To use an ice cream analogy, it feels like there are now 31 different flavors. Thirty one is an exaggeration, but the point is there are many bespoke features being added to these transactions. The plethora of tax credit types is also adding complexity.

Another factor driving change is the electrification of everything and the mammoth demand for power from AI and data centers.

We saw a large tech company acquire a mid-sized project developer. That may be a leading indicator of things to come given the voracious demand from the tech sector for electricity.

MR. MARTIN: Google agreed to pay $4.75 billion in late December for Intersect Power. Google probably also has tax capacity to use any tax credits on the Intersect projects.

Rubiao Song, what new trends did we not mention that you are seeing as we move into 2026?

MR. SONG: We mentioned AI for the first time about 30 minutes into the call. We see more load growth and rising prices for electricity. Inflation measured by the consumer price index was 2.7% in 2025 but electricity prices nationwide were up 6.6%. The price increases varied by region.

Rising prices are a tailwind that will support continued development of renewable energy projects.

Finally, there is a pressing need to expand the pool of potential tax credit buyers. We saw robust growth in the number of buyers in the last two years, but more buyers are needed. There is a growing supply-demand imbalance on the ITC side that will only get worse given the number of new projects coming to market.

Debt

MR. MARTIN: Let's move to debt. Ralph Cho, what was the volume of North American project finance bank debt in 2025 compared to 2024?

MR. CHO: Last year set records for lending volume. Any banker who was not busy may be in the wrong profession. According to Refinitiv, 2025 volumes in North America came out just a little under $260 billion spread across approximately 500 deals. That is more than a quarter of a trillion dollars.

Basically every sector we saw was on fire, whether it was LNG, renewables, digital. Every sector contributed significantly to that number. We even saw a $1 billion greenfield combined-cycle gas-fired power project close. We have not seen new construction like that for natural gas for several years.

Volumes in the B loan market, bond market and private credit also were at highs. Global project finance lending also set a record at over $500 billion.

Year-on-year lending volume in the North American project finance market was up 41%.

Human capital was short this year. It was really short. The lending desks at banks could have used support from AI to keep things going.

MR. MARTIN: Lending volume was up 41% in 2025 or over 2024. Is that dollars or number of deals?

MR. CHO: Dollars.

MR. MARTIN: How many active lenders were there in 2025 and was that a change from the year before?

MR. CHO: We saw a host of new lenders come into the market in 2025. Some are banks that I had never heard of. Some are banks that had been in the market, stopped and then restarted their project finance lending desks.

We saw real estate investors coming in. We saw ABS lenders. We also saw a wave of newly formed private credit funds raising capital for deployment.

Not too many new lenders focus on renewables because that sector is already very competitive. We certainly saw new lenders interested in financing LNG and digital assets.

There are more than 100 active infrastructure lenders. Some are more active than others. More important than the number of lenders is that the liquidity in the market is super, super deep. The mega banks are taking large positions, and the retail lenders are all scrounging to find funded loans.

The irony is that even in a year with very high volumes, lenders are still short paper and are looking to buy loans or loan participations. I can tell you that our team was looking to buy funded loans at the end of last year, and we found it very difficult to get any meaningful amount of paper in the secondary market.

MR. MARTIN: Beth Waters, I have a longer question for you.

What effects are the Trump policies -- frequent changes in tariffs, a freeze on federal approvals for wind and solar projects, new FEOC limits taking effect this year on use of Chinese equipment and intellectual property, an expectation that short-term interest rates will fall once a new Fed chairman takes office in May, and mass deportations contributing to a shortage of construction workers -- having on the ability of developers to finance projects?

MS. WATERS: For a start, the banks are not open for financing offshore wind.

The new rules on onshore wind and solar on federal land are definitively having an impact.

But for every other aspect of the business, the banks are there. As Ralph said, there is a tremendous amount of liquidity. We are seeing a lot of due diligence underway on potential financings. Diligence is especially important to evaluate tariff and FEOC risks.

The federal funds rate went down three times last year by a total of 75 basis points. However, the 20-year swap rate, which is what is used for project finance, was unchanged at year end at 4.10%. Since then, it has moved up to 4.25%. The expectation is that it will fall by about 25 basis points this year and settle at 4%, but it is important to recognize that reductions in the federal funds rate are not translating into similar reductions in long-term borrowing rates.

Pre-NTP Loans

MR. MARTIN: Ralph Cho, the hottest new product in recent years has been borrowing-base loans where a developer can draw on a revolving loan facility to cover late-stage development costs. What metrics apply to such facilities?

MR. CHO: We have been talking on this call for the past few years about pre-NTP loans. I would divide them into three main categories.

There are mainstream pre-NTP facilities. The amount loaned is a function of the debt service coverage ratio. The loan is made at the holding company level. Lenders for the most part maintain collateral over the whole portfolio, meaning operating, construction and development assets. You basically take the cash flows from the operating and construction assets and size the loan to a debt service coverage ratio. The borrower uses the loan to pay costs to advance its development assets.

We have syndicated billions of dollars in these types of loans and have kept hundreds of millions of dollars on our own balance sheet and priced them somewhere around 350 to 450 basis points over SOFR.

The second type is a borrowing-base loan. That is a loan against a pool of late-stage development assets. The projects have power purchase agreements and interconnection. Everything is lined up, but the developer may not quite be at the point to start construction. The lender will assign a present value or enterprise value to the construction and development assets and advance 50% or 60% of that value as a loan. They will take the operating assets, do a similar calculation and advance 75% or 80% of value against them.

These types of loans are priced a little wider at anywhere from 450 to 600 basis points over SOFR. The spread varies based on the size of the developer and its track record.

The third type of loan is a riskier pre-NTP facility. We do not see many of these any more. There were specialty borrowers that would take a super early look at a development pipeline and assign some kind value to it and advance against that. The challenge with this type of loan is that if the developer is unable to advance the assets, the lender would have to figure out how to liquidate the portfolio. These loans priced at equity-replacement rates of 600 to 800 basis points over SOFR.

Debt Metrics

MR. MARTIN: Beth Waters, focusing mainly on electric generating and storage facilities, what percentage of project costs can be borrowed during construction?

MS. WATERS: If you assume a good clean deal, you could get 90%. The leverage drops below 90% the more risks there are with the project. The advance rate depends on the specifics of the project.

MR. MARTIN: Is 90% the advance rate for the combined construction and bridge loans?

MS. WATERS: Yes. The sum of the two cannot exceed 90%.

MR. MARTIN: There are both covered and uncovered -- I think you call them naked -- tax credit bridge loans. What is the difference, and what metrics apply to them?

MS. WATERS: Covered, contracted or committed tax credit bridge loans are loans against the expected future proceeds from selling the tax credits on a project where someone has committed legally to buy the tax credits. In that circumstance, we would give a 98% advance rate of the projected sales proceeds, and pricing should be around 150 basis points over SOFR.

When the bridge loan is uncovered, uncontracted or naked, the advance rate is lower. For an uncovered investment tax credit bridge loan without a tax credit buyer already lined up, the advance rate is 75% with the tax credits assumed to be worth 90¢ on the dollar. That nets you about 67.5% of the tax credit face amount versus 98% of the agreed purchase price if you have a committed buyer. The pricing is 225 basis points over SOFR.

Some lenders take different approaches. Some use a 70% rather than 75% advance rate for uncovered loans.

Most of the loans are against ITC sales proceeds. We do not see as much lending against future sales proceeds for uncontracted PTCs, although we are open to such loans.

A loan may start as an uncovered bridge loan and once a tax credit buyer is found, the advance rate increases to 98% of the agreed purchase price and the spread drops to 150 basis points.

MR. CHO: I agree with those advance rates, but we have seen some construction lenders go as high as 95% advance rates given the competition among lenders. That said, the larger the deal, the less likely lenders will be willing to be super aggressive.

MS. WATERS: I don't think I have ever done 95%. A majority of lenders really do not like going above 90%.

MR. MARTIN: The debt service coverage ratio determines how much permanent debt a project can support. What are current coverage ratios?

MS. WATERS: They have not changed since last year. I am giving you middle-of-the-road figures. Depending on the sponsor and how much opportunity the bank sees for additional lending to that sponsor, banks may be willing to be more aggressive.

The debt service coverage ratio for solar at P50 output figures is 1.25 to 1.30 times debt service. Wind at P50 output is 1.35 to 1.40 times debt service. Storage -- where there is no P factor -- is 1.15 to 1.20 times debt service. For wind and solar projects, the project must also demonstrate full repayment of the loan via 100% cash sweeps under the P90 output case while the project or portfolio is still contracted.

For projects with merchant revenue, P50 merchant cash flow for solar must cover 1.75 times debt service. For wind, it is 1.80 times and for storage -- again, where there is no P factor -- it is 2.0 times.

These are middle-of-the-fairway numbers.

MR. CHO: I would say you should forget about any meaningful efficient financing and demand from lenders for projects in ERCOT with merchant revenue We tried and looked around. Financing for merchant batteries in ERCOT is essentially on life support.

My guess is you could probably get away with 25% to 30% merchant exposure. Above that, it is pretty dead.

MR. MARTIN: Are you talking about storage or any merchant projects?

MR. CHO: Anything merchant. For solar, wind and batteries, I would limit the merchant revenue to a maximum of 25% to 30%. You can finance the part that is not merchant, but once you have 25% to 30% merchant exposure, the entire financing becomes difficult.

MR. MARTIN: Beth Waters, do you agree with that?

MS. WATERS: Yes, but it is site specific. We look closely at each transaction. I am definitely not going over 40% merchant.

MR. MARTIN: You said merchant revenues at this level are a problem in ERCOT. What about in other parts of the country with organized markets?

MS. WATERS: We would have an issue with a high merchant component in any market, but ERCOT is where we typically see developers looking to finance 100% merchant storage projects. Banks have been burned on such projects in ERCOT, and that is why you are seeing a pull back on how much merchant banks will finance.

MR. CHO: I don’t think the market will accept significant merchant renewable financing in any region in this country.

The only other asset class I would mention is data centers because they are taking up so much volume in the market. The required debt service coverage ratio for those is down to 1.15 times where the data center has contracted revenue from a hyperscaler.

MR. MARTIN: Beth Waters, what are current spreads above SOFR for construction debt versus permanent debt and how do they compare to spreads a year ago?

MS. WATERS: There has not been any real change in spreads for construction loans. For a clean deal, the spread is 150 basis points. Some developers can get spreads as low as 125 to 137.5 basis points, but that is not common. The spreads are higher for projects with long construction periods, projects with merchant revenue, projects in ERCOT, projects with electricity basis risk, projects with offtake agreements that have a unique structural aspects such as some resource adequacy contracts in the California market, or other issues that make the project riskier from a credit perspective.

The middle-of-the-fairway spread on term loans is 162.5 to 187.5 basis points. If there is 10% to 20% merchant revenue, the spread increases by 25 basis points. If merchant revenue is 30% to 40%, the spread increases by another 25 basis points.

MR. CHO: In terms of changes in the last year, the spreads have probably fallen an eighth of a percent or 12.5 basis points on both the low and high ends of the range.

There has definitely been some tightening on spreads because everyone is trying to beat each other up to win business.

Data Centers

MR. MARTIN: The biggest data center developers are expected to spend $500 billion just this year on expansion. Have there been any new developments in the last year in lending to finance new data centers?

MR. CHO: The first one is the risk around having CoreWeave as a tenant. Underwriting that risk has become very tough in the current market. Billions of dollars of deals have been sold to the market with spreads of between 325 and 425 basis points over SOFR. Everyone is worried about an overbuild situation. One path forward is to move this debt to institutional lenders, but the pricing has to move closer to where CoreWeave bonds are trading, and they have been volatile.

A second theme is Oracle. Oracle is an investment-grade entity, but it is having a tough time in the market over fears of a potential downgrade. Banks have sold billions of dollars of this credit into the market at around SOFR plus 250 basis points. There are some large “hung” underwritings today. That is another piece of paper that might have to be placed with institutions. I am hearing some shadow talk of prices as high as SOFR plus 350 basis points, but the reality is you might also need additional structural enhancements to create a path to successful syndication.

A third development is that data center developers are pooling their assets and borrowing holding company debt against portfolios. We can be very aggressive on this type of loan. Debt service coverage ratios for loans at the data center project company level are 1.15 times. We have gone as tight as 1.05 times against portfolios with contracted revenue from hyperscalers.

Developers want enough runway not to have to refinance quickly after the data centers reach commercial operation. Refinancing is most often into the ABS market, but that market continues to dwindle on capacity.

Finally, we are also seeing an increasing number of opportunities for behind-the-meter power generation at data centers. The data centers need power right away to operate. They are stuck in long interconnection queues that could take years in some cases to get to the front of the line. They are looking for alternatives. We have talked to a number of developers that are trying to come up with creative solutions and put some kind of generation on site. This could be interesting and lead to some deal activity in 2026.

MR. MARTIN: Are you concerned about the ability of generation that is not grid-connected and whose sole source of revenue is a single data center to repay debt service?

MR. CHO: The devil is in the details, but I assume in these cases that the buyer of the power is contractually locked in with the hyperscale tenant under a full passthrough agreement. The risk is the same as the data center being able to collect rent payments from hyperscalers. To answer your question, I would not be concerned in this situation.

New Debt Trends

MR. MARTIN: Let's wrap up with this question. What other noteworthy debt trends are there as we enter 2026?

MS. WATERS: The demand for power from data centers is huge. We are seeing a growing number of financings of generating facilities being built to supply power to data centers. Some involve using reciprocating engines because of the difficulty getting new full-fledged gas turbines to market quickly.

We are also starting to see deals that combine financings of data centers and generating assets into one financing package. Some generators add solar or battery storage to the reciprocating engines because the hyperscalers still like renewables. Sometimes the renewable generation is expected to be the primary supply with the reciprocating engine added initially to get things running with the plan of having it shift eventually for use as backup.

We should continue to see a lot of interest this year in pre-NTP loans. Our bank is not crazy about doing those, but we are helping clients with whom we do business by bringing in other MUFG product teams who can do things like equipment financing or corporate lending hopefully to satisfy their needs on the pre-NTP front.

We are seeing generators having difficulty finding solar panels that do not raise FEOC issues. That is coming up more frequently as we move into 2026.

I expect to see continued delays in construction timelines. Projects are not moving as quickly. Some of this is due to the need for more diligence on tariff and FEOC risk, and sometimes it is because construction is just taking longer.

We are seeing more financings of projects that are selling investment tax credits, but the projects are sold first to a preferred equity partnership to step up the tax basis. We closed recently on a club bank deal where the banks provided back-levered debt in such a structure. We are starting to see more of that.

MR. MARTIN: Ralph Cho, what other new trends are you seeing as we roll into 2026?

MR. CHO: Lots of trends, but I will limit this to four.

The market will be really busy this year with financings of renewable energy projects. Developers are pushing these projects to get done before tax credit cliffs. Many developers have large pipelines of safe-harbored projects that have locked in tax credits, provided the projects are completed by end of 2028, 2029 or 2030. It is best to get them done as quickly as possible. Lenders will want a buffer.

The second trend is natural gas. New gas-fired power plants will be hard to build this year because it is difficult to find turbines, but with valuations of operating gas plants going through the roof, the B loan market has been a hot market for borrowers to tap. Many of these operating plants have been financed in the B loan market. The owners can get higher leverage, better pricing and better refinancing terms. Everyone is doing this. It is rinse, wash and repeat over and over again. I expect to see more of this and maybe even some acquisition financings in the B loan market.

The third new trend is all the greenfield financings of LNG terminals, renewables and data centers are creating an opportunity for the private placement lenders to refinance longer term at COD. We have a life insurance bucket of capital with Athene, so we could participate in that. There is a lot of demand for long-term, fixed-rate debt.

Finally, with all this activity, you need people. Demand for human capital will remain white hot this year. Like others, we have been hiring at every level, and it is hyper-competitive.

MR. MARTIN: Summing up, everyone is expecting another extremely busy year. People were exhausted after last year. The shortage of human capital means a lot of us will be looking for ways to harness AI during the year to fill in some of the gaps. That said, the unpredictability on the policy front in Washington will remain a cloud all year over the market.