I’m trying to understand what is happening at the WIFIA Loan Program. The mysterious issue is not the near-complete cessation of loans under Trump 2. The reason for that is obvious: OMB’s pause on federal grants and loans in February combined with continuing federal chaos under this Administration.

Much harder to explain is the steady decline in Program loan volume since the end of 2021 through January 2025, despite a continuing rise in US water sector capex. During this period, WIFIA had a supportive Biden Administration (including OMB), plenty of funding and was as efficiently run as ever. Yet the Program’s annual executed loan volume fell from a calendar year peak in 2021 of over $5.5 billion to under $2 billion in 2024.

^{Data sources: EPA WIFIA website, St Louis Federal Reserve FRED}

It’s a dramatic decline, and I think understanding the reasons has important policy implications. Here are my initial thoughts about four possible causes.

Possible Reason 1: Apparent decline is statistical anomaly

Perhaps the premise of ‘decline’ is unwarranted? The monthly volume chart appears to show rising and falling trends, but the data set is small (only 137 loans in total) and the time scale is (in project finance terms) quite limited. It is not impossible, or even improbable, that the apparent trends are simply random noise and some coincidences.

This is certainly valid in statistical terms. If this were an academic exercise, it could be stated that it’s too early to draw conclusions. However, in a real-world context of near-term challenges in US water infrastructure and short-term political impact on policy direction, I think the decline is real enough to cause concern:

• Contrast to growing demand in US water capex: The important reference point is not the growth (or lack thereof) of a government program, but outcomes relative to the problem it exists to solve. In those terms, WIFIA’s decline is even more dramatic, from a ‘market share’ of financing about 17% of US water capex in 2021 to under 4% in 2024. Even if 2021 peak volume remained steady at $5.5 billion (which would be considered a solid ‘success’ in self-referential program terms), the decline in ‘market share’ would still be pronounced:

^{Data sources: EPA WIFIA website, St Louis Federal Reserve FRED}

• Pending loan stack might not indicate strong demand: WIFIA has 80 ‘pending loans’ in the transaction pipeline. That might look like a backlog due to strong demand, but I’m not so sure. There are 42 loans in the ‘paused’ or ‘invited but not yet applied’ categories and only 38 in the application stage. That’s not a great mix. Time in pending list is not specified. In my experience of project finance, borrowers pursue several parallel tracks and keep their options open until the last minute. They’ll push hard to progress deals which have especially beneficial terms and effectively slow walk the others. It is possible that most of these ‘pending loans’ are secondary alternatives for project borrowers, and that few will close. But even if all 38 applications proceeded to close in 2026 at WIFIA average loan size, that total volume would amount to only about $6.6 billion — not nearly enough to restore WIFIA’s peak 2021 share of US water capex.

• Stakeholders and policymakers have agency: In situations where you can’t affect the outcome, waiting for more statistical validity is justified before drawing a conclusion. But WIFIA’s declining volume can be seen more like a business challenge where executives must exercise judgement and act as decisively as possible on behalf of shareholders. If negative trends automatically reverse themselves, fine. But if the executives do nothing while waiting for further information and the trends turn out to be a fundamental downward trajectory — well, that’s bad management 101. So too for WIFIA stakeholders and policymakers: A 63% decline in loan volume over three years is a sufficient red flag to act.

• Perception matters: Regardless of statistical validity, the numbers and graphics depicting WIFIA loan volume since 2021 aren’t great. In political conditions where ideology can drive otherwise purely technocratic policy decisions, perception is the reality. We’re in that situation now and likely will remain so for at least three more years.

Possible Reason 2: Rising long-term rates

Long-term interest rates started rising off historic lows in 2020, but the pace and trajectory solidified by the end of 2021. Perhaps project sponsors increasingly turned to other, non-debt, sources of capitalization, e.g., internally generated equity, reallocation of revenues, or grants?

This is generally implausible because of the scale and steady rise in US water capex, the fact that inflation was often higher than interest rates, and uncertainty about where the rise would end (i.e., rising rates would be a reason to lock in current rates). More specifically:

• Higher rates did not affect similar finance volume: Another major source of water infrastructure finance, tax-exempt municipal bonds, did not experience a significant decline — volume dipped slightly in 2022 (perhaps Covid-related), held steady in 2023 and surged in 2024, with 2025 on track to be a record year.

• WIFIA Loans and muni bonds are comparable: It should be noted that WIFIA loans and muni bonds are in fact similar in many fundamental ways from the borrower’s perspective. Both are long-term, fixed rate, amortizing, secured debt that can be done in large scale for investment-grade project borrowers. Both are federally subsidized. Most importantly, the vast majority of WIFIA’s borrowers are highly rated public water agencies that generally issue in the tax-exempt muni market as well. There are critical differences (as discussed below), but the two types of debt are close enough to assume that demand dynamics for US water capex finance will apply to both.

• WIFIA loans and muni bond trends dramatically different: In the chart below, I would have liked to include water revenue muni bond issuance, but I couldn’t find a suitable data series (will keep looking). Still, I think the three types shown — overall muni, muni utility, and muni revenue bonds — ought to incorporate and accurately reflect water infrastructure financing. In any case, the comparison to WIFIA loan volume is not close:

^{Data sources: EPA WIFIA website, UST website, Bond Buyer 2016-2025 Data Summary}

Possible Reason 3: Huge IIJA funding for SRFs

The IIJA passed at the end of 2021 and provided a huge chunk of new funding, about $44 billion, for US water infrastructure through the EPA’s SRFs. Of this amount, about $20 billion had been contractually obligated to state agencies by mid-2025, according to Bluefield Research.

Could 2021 IIJA funding have depressed WIFIA loan demand, either through direct displacement by CWSRF and DWSRF loans or by various other grants? The timing is an exact fit, and the scale is right. Even if the funding took a while to roll out, the prospect of much cheaper SRF loans or free grants might have had a dampening effect on WIFIA loans.

^{Source: Bluefield Research “Mapping the Progress of IIJA Funding for Water Infrastructure” WFM August 25. 2025}

My provisional answer at this point is ‘probably’, but I don’t have enough information to confirm or be more precise about the impact. Still, a few observations can be made:

• IIJA funding impact on smaller projects: SRF loans and IIJA grants might be expected to impact smaller projects in terms of WIFIA loan displacement. It varies widely, but SRFs mainly lend to smaller, less creditworthy local projects. A grant for a larger, investment-grade project would of course be welcome, but it would still only account for a small part of total capitalization in most cases. If that pattern applied to IIJA-funded sources post-2021, you would expect to see the average size of WIFIA loans increase even while total volume declined. Yet, the data shows the opposite:

^{Data source: EPA WIFIA website}

• No significant displacement of muni bond finance volume: If increased SRF lending and other IIJA funding were displacing other sources of finance for US water capex, you would expect to see some impact on the muni bond market. Perhaps that was another factor in the slight decline and then flatline of muni bond volume 2022-2023 noted above? But there was an upsurge in 2024-2025 in bond volume, dramatically different than WIFIA volume’s continued decline.

• Unlike SRF loans and grants, WIFIA loans are exclusively senior debt: Since SRFs generally can lend on a subordinated or non-senior basis, and of course grants can generally be used as part of equity capitalization, it might be expected that some WIFIA loan displacement by projects that no longer needed debt due to IIJA funding would be balanced by other projects that were enabled to proceed with senior debt capitalization, including WIFIA loans. SRF and WIFIA loan terms are often quite complementary and have worked well together in project capital stacks. This point, as well as the one above about muni bond volume, suggests that if IIJA funding contributed to the decline of WIFIA loan volume, the reasons are likely more complex than direct dollar-for-dollar displacement.

Possible Reason 4: Relative Cost of a WIFIA Loan vs. Tax-Exempt Bond

Perhaps project borrowers were actively choosing between two types of senior project financing, WIFIA loans and muni bonds, throughout the Program’s operational period 2018-2024? WIFIA volume trends, both rise and fall, would then reflect the criteria the borrowers were using, as applied to changing conditions.

As noted above, WIFIA loans and tax-exempt muni bonds are structurally quite similar. They are not perfect substitutes, however, and even relatively small differences mean that the optimal choice of either as senior project debt will depend a lot on the specific character of the project and the situation of the borrower.

Nevertheless, there are two observable factors in the cost of each financing alternative which apply to most if not all projects deciding between WIFIA loans and muni bonds:

• Cost of interest: A WIFIA loan’s long-term is based on then-current US Treasury yield corresponding to the weighted average life of the loan (usually about 20 years) while a tax-exempt muni bond’s equivalent long-term fixed rate is based off the market-clearing rate relative to the Treasury curve but reflecting the value of the tax-exemption and a credit spread. Roughly, for a typical project financing, the WIFIA loan has 20Y UST rate and a muni bond series has (in effect) an overall rate close to the 20Y muni market index. The relationship between the 20Y UST and the 20Y muni index depends on a lot of things and is not at all constant. The difference can be close or positive (muni index exceeds UST) but is typically negative (muni index about 80-90% of UST) because of the tax-exemption.

• Cost of locking long-term financing rate during construction: Infrastructure projects generally have a years-long construction period. To avoid financial risk, borrowers will generally lock in the project’s long-term, fixed-rate financing at construction commencement by borrowing the entire amount, which is then escrowed and available for construction draws. The funds in escrow will be invested in short-term debt, which typically (but not always) has a lower rate than the long-term rate of the financing, thereby incurring ‘negative arbitrage’. Muni bonds issued for infrastructure projects will have this negative arbitrage cost. In contrast, a WIFIA long-term loan’s rate is ‘locked’ at loan closing and construction advances — even over years — can be drawn at this rate. There is no negative arbitrage.

Again, projects are idiosyncratic, but the interest rates that determine the cost of interest and negative arbitrage are observable. The chart below overlays the 20Y UST, 20Y muni index rates, and the 1Y UST (as an approximation of the short-term construction escrow investment rate) on WIFIA monthly loan volume:

^{Data sources: EPA WIFIA website, UST website, MSRB EMMA website}

As a very general observation, it looks like WIFIA volume is roughly correlated with (1) the closeness of the 20Y UST and the 20Y muni index rates, and (2) the positive difference between the 20Y muni index and 1Y UST rates. From 2018 to the beginning of 2020, UST and muni index rates were close, and there was some gap between the muni index and the 1Y UST, both factors favoring WIFIA loans. In 2020 the factors favoring WIFIA were even more pronounced, becoming less so but still favorable 2021 to early 2022. From that point through 2024, however, the factors became markedly unfavorable — the muni index was much below WIFIA’s UST rates and short-term investment rates were higher than long-term muni rates.

A more precise estimate of relative WIFIA loan value can be modelled using the full set of data that’s available (e.g., full rate curves, more precise amortization and loan tenor schedules, estimated credit spreads for typical water projects, etc.). The comparison is boiled down to an estimate of the net present value (NPV) difference between the cost of (1) using a WIFIA loan in 49%/51% combination (the statutory maximum) with a muni bond, and (2) a 100% muni bond financing for the project. The model is not simple, but the methodology is straightforward — muni issuers and advisors use precisely the same NPV analysis to evaluate bond advance refunding, for example. Public water agencies, especially those with relatively big projects, will be very familiar with using it.

The chart below overlays NPV results on the above rate patterns. The results are consistent with the general observation: NPV benefit was solidly positive for the WIFIA/Bond combination from 2018-2021 with a very positive spike in 2020, but thereafter declining, eventually into deep negative numbers, from 2022 onwards:

^{Data sources: InRecap model, UST website, MSRB EMMA website}

However, NPV is just one factor in the WIFIA vs. muni bond decision. Several valuable features of a WIFIA loan don’t show up in the calculation, and their exact benefit is both idiosyncratic to the project and difficult to quantify. For examples, WIFIA’s rate lock also can be used in some ways as an ‘exotic’ interest rate option, WIFIA’s loan term can extend to about 40 years, the loan is immediately prepayable at par, a debt service deferral is usually available, the Program represents some diversification of capital sources, etc.

For now, one way to incorporate these non-NPV benefits into the comparison is to simply (and, admittedly, somewhat arbitrarily) add a fixed number to the NPV result — I’ll use 3% pending further information and analysis. The chart below overlays the resulting ‘WIFIA Benefit Index’ on monthly volume:

^{Data sources: InRecap model, EPA WIFIA website, UST website, MSRB EMMA website}

• A good fit: Apparent trends in the WIFIA Benefit Index appear to track actual trends in monthly volume fairly well. This is intuitive — when the Index is very positive, WIFIA not only gets more applications but more applicants already in the pipeline will press harder to close. When less positive or even negative, fewer will apply or press to close. Fewer, but not ‘none’ — again, the value of WIFIA loan is idiosyncratic to the project, so there are likely always be some in absolute need of a unique feature the Program offers. Projects in need of WIFIA’s unique, non-NPV related features are likely to be smaller — perhaps that explains the declining average WIFIA loan size described above?

• Comparable analysis for IIJA SRF impact: In theory, a comparable analysis examining the relative value of WIFIA loans to SRF loans in each state could be done. That would provide an even clearer picture of the possible impact of IIJA funding, e.g., if the funding allowed more SRFs to offer better terms or had more loan availability than pre-2022, WIFIA loan volume would have been hit with two negative factors (better muni bond pricing, better SRF loan terms) in the period 2022-2024. However, a lot of state-level data would be required for a complete analysis. Perhaps anecdotal or spot-check based evidence would be sufficient to provide some guidance for now?

• Hypothetical 2025 no pause/no Trump volume: The chart also adds a hypothetical volume for 2025 as if the OMB pause (or Trump 2) had not occurred. This is essentially a guess based on the Index’s slight upturn. Since September 2025, however, the Index has dipped deeper into negative territory, at about negative 4% as of 11/13/25. The hypothetical, no pause/no Trump, volume for all of 2025 would probably have been about $1.7 billion — that is, a continued decline though at a slower pace than in 2022-2024.

Summary

• WIFIA loan volume decline is significant and persistent enough, especially in the context of growing US water infrastructure needs, for stakeholders and policymakers to act.

• Overall rise of long-term interest rates over the period does not provide an explanation for the decline — and therefore falling long-term rates at some point in the future are unlikely to restore WIFIA loan demand.

• Some impact from IIJA funding for SRFs on WIFIA loan volume is likely, but its scale or nature is unclear. More investigation is warranted — stakeholders and policymakers for both sources of federal finance should want to see a more coordinated approach between the two EPA programs.

• The relative value of WIFIA loans to tax-exempt municipal water revenue bonds appear to explain much of the trends in WIFIA loan volume for its entire operational history, 2018-2025. This may seem to be a somewhat abstract or technical conclusion, but if changing value relative to near-substitutable muni bonds is indeed the predominant factor in WIFIA loan volume, serious policy questions and considerations for the Program arise: What is the purpose of WIFIA? Is the Program really enabling more US water infrastructure — or is it acting as an intermittently active ‘interest rate arbitrage’ adjunct to another federally subsidized source of water infrastructure finance? If the latter, how can this be justified as intentional policy, especially under the current Administration? And so on.

Again, these are initial thoughts. But I intend to continue keeping a close watch on WIFIA developments. Like much else in the federal landscape, major changes for the Program under the Trump 2 Administration are possible, perhaps likely. US water infrastructure sector stakeholders and policymakers should be aware that the next few years will pose both risks and opportunities — serious thinking is warranted.