The short statement seemed worth doing as a succinct summary of where WIFIA et al should be going and what might be required to get there. Another opportunity to practice the narrative, as it were.
Of course, I don’t really expect a statement or anything else about WIFIA to be considered or added to the meeting agenda by EFAB. Despite the Board’s name, the mechanics of federal loan programs, however impactful on EFAB’s stated policy priorities, just aren’t what they do. Instead, I assume the meeting will be consumed by the usual au courant topics and worthy initiatives. Many of which will soon be swept away by the hard reality of economic uncertainty and possible decline. The latter are the real issues impeding infrastructure renewal and improvement, with significant consequences for real-world environmental and social outcomes. The value of federal infrastructure loan programs designed to mitigate some of these challenges will likely be recognized soon enough by people that need to make decisions. The others can catch up later.
Yesterday the 20Y UST rate was about 4.75%. Where will it be in September 2024? Probably about the same or even higher, given recent Fed indications. But who knows? We live in interesting times and a lot could happen in a year. Known unknowns are bad enough — massive geopolitical change, serious economic uncertainty and a ferocious presidential election, to mention a few. Add black swans to that mix and it’s entirely possible that Fed policy will make a radical U-turn at some point in the next few years and long-term rates will be much lower, at least for a temporary period.
A Quandary for Long-Term Infrastructure Financing
If you’re planning on financing a public water infrastructure project with long-term debt anytime soon, uncertainty about falling rates poses a quandary. Yes, a standard tax-exempt bond series with a weighted-average-life (WAL) of about 20-25 years will have a sub-UST overall rate (around 4.25%) and the cost of the construction escrow will be negligible. Once issued, you’re protected from further rate rises. But you’re also locked out of tax-exempt refinancing by advance refunding restrictions until the bonds can be called, usually about 5-10 years out [1].
A WIFIA loan for 49% of project cost will have a higher fixed UST-based rate (about 4.75% for 20-25-year WAL). But it has other benefits. The rate is locked for construction draws so an escrow is unnecessary, the full loan term is about 40 years, and the loan has no commitment cancellation, prepayment or make-whole penalties. Obviously, a WIFIA loan is not subject to advance refunding rules and can be refinanced anytime.
On the surface, the option of refinancing to take advantage of near-term opportunities might seem to offset the loan’s slightly higher rate. But a federal loan will have crosscutters like Davis-Bacon that increase project cost. The increased costs are not necessarily major, but it would be better (at least optically) if they were amortized for a few years before sinking them. And there are more direct political angles. In contrast to standard muni bond decisions, a WIFIA loan will need to be championed internally. Externally, federal & local pols may show up for a loan closing photo-op and speechify about how great the loan is. You may hesitate before unwinding those stories too soon. As a practical matter, a WIFIA loan’s first refinancing option might be a few years after closing, not that different in principle from the first par call on some muni bonds.
WIFIA’s Downward Rate Reset
This is where WIFIA’s reset comes in. In 2018 and early 2019, WIFIA executed a few big loan commitments with highly rated water agencies. The commitments remained undrawn because short-term tax-exempt debt for construction draws was cheaper. In effect, the agencies were using their commitments as fixed-rate options for permanent financing. By late 2019 and 2020, these options were increasingly out-of-the-money as interest rates fell to historic lows. In a purely private-sector transaction, such options would simply be written off. As noted above, things are more complicated when a federal loan commitment is involved. But the political complications work both ways — with the threat of cancelling the big, much photo-op’d and press-released loan commitments, the agencies could pressure WIFIA into re-setting the option strike price (i.e., the loan’s rate) to market. It took a few months of internal wrangling, but WIFIA duly delivered the first loan reset in late summer 2020. With the precedent established, resets for six more out-of-the-money loan commitments were completed in late 2020 and early 2021. As interest rates began their steady rise to more normal levels and beyond, previously executed commitments were increasingly less underwater and new commitments were in-the-money, so further resets were unnecessary. [2]
A Bureaucratic Accommodation
It’s important to note that in essence WIFIA’s decision to do resets was a matter of unusual bureaucratic accommodation, not economic reallocation. If you have an out-of-the-money interest rate option but want one with a lower strike, it’s easy enough to sell the old option and buy a new one, though of course you’ll incur a net loss. For a WIFIA ‘option’ (the undrawn loan commitment), the borrower can cancel its existing one and re-apply for a new one executed at current UST rates. By clarifying that the purpose of the cancellation and re-application is solely to obtain a better rate, there shouldn’t be any sunk project costs or political friction. And since both options are ‘free’ anyway, there’s no loss. Well, other than transaction costs, right?
Ah, transaction costs — there’s that. The full WIFIA loan application and execution process involves a series of steps and approvals. Repeating it will involve re-submitting (and perhaps updating) a lot of documents and reports, which, though likely to be successful again, takes time and the outcome is not assured. Political and policy priorities may have changed in the competitive selection process for scarce credit subsidy, or there are just more worthy candidates this time around. Or OMB is in a surly mood and won’t approve anything to do with loans. Or the government has shutdown altogether for an indefinite period. And so on. Even if the incremental fees are trivial relative to the lower rate’s value, the process itself is a risky and potentially costly downside.
WIFIA’s accommodation was to short-circuit all that and simply re-execute the loan commitment with new UST rates. The important point is that they didn’t need to do it. There’s nothing in WIFIA’s statutes or typical federal credit procedures that anticipate this kind of situation because they’re built on the assumption that hard-pressed project finance borrowers will be only too happy with a UST rate and start drawing the loan right away. The default bureaucratic response in such cases is simply to shrug. Yet the WIFIA program, to its credit, saw the reverse of the medal in the absence of guidance — a re-execution wasn’t explicitly prohibited — and pro-actively got approval to do it. No doubt there was a political assist at various levels. Even so, I don’t know how technically difficult the expedited re-execution was, or what questions might have been asked (e.g., why do the borrowers need a WIFIA loan if they’ve got such good alternatives?), but a reset got done and then repeated six times to date. And the program wasn’t shy about advertising the outcomes in the usual terms. [3]
How Established is the Precedent Going Forward?
Obviously, there is still no WIFIA statutory provision that requires a reset even to be considered, much less granted. The program’s statement above from 2021 suggests that at least one reset will be “available”. But what is unexpectedly given by a bureaucratic system can be taken away just as opaquely. And what exactly is meant by the qualifier, “in limited circumstances”? Do they include your loan commitment being out-of-the-money by 25 bps. relative to the Aa3/AA- tax-exempt bond alternative the local underwriter keeps talking about over a 5-star lunch? Or your loan commitment is underwater by 2.50%, local loan sharks are offering a better rate, and you’re desperate to make your innovative environmental project at least somewhat economically feasible? They don’t say.
The exact nature of the reset precedent is relevant with respect to future applicants. Assuming that the reset remains “available”, there are two ways to see it in the context of the applicant’s intentions in applying to the program. The first sounds like WIFIA’s own statement:
The WIFIA loan was applied for as a primary and essential component of the project’s financing. If rates fall significantly, the program can and will help you out with a one-time rate reset if the loan commitment hasn’t yet been drawn, due to construction timetable or other real-world factors.
The other is quite different:
The WIFIA loan was applied for as an optional component of the project’s financing. Tax-exempt bonds are expected to be the primary source. The WIFIA loan commitment is an interest rate option and will remain undrawn until it’s optimally in-the-money for permanent financing. If rates fall significantly, or even materially, relative to bond alternatives, the program must reset the loan rate once when requested, as equitable treatment established by precedent.
The former is in line with WIFIA’s stated overall policy purpose and rationale, where the reset is provided as a helpful benefit to keep the project going in the face of unexpected developments. The latter is not. There the WIFIA loan is being used as a financial option, not a financing source, and the reset is simply a feature to improve its value. Of course, if the applicant is a public-sector agency (almost all are), then either way the benefit ends up reducing total project cost to local ratepayers, so there’s still a public purpose angle. Rather, the problem with the pure financial option usage is the explicit lack of any real or imagined additionality with respect to the actual infrastructure project, which gets built in any case. Not quite the narrative policymakers intended.
The intentions of WIFIA’s actual applicants and borrowers to date can objectively be said to fit somewhere between the two alternatives. This includes the seven recipients of past resets, all highly rated and sophisticated water agencies whose intentions would arguably be closer to the latter alternative.
WIFIA applicants obviously don’t advertise any intentions that might deviate from the approved policy narrative, and as far as I can tell, the program doesn’t ask too many questions. If available, and limiting circumstances are more nominal than constraining, the program will likely grant future resets to those of pure and impure intentions alike.
But the intentions of the applicants may matter more with respect to a different question that involves the reset — whether to apply for a WIFIA loan in the first place. For applicants that intend to use the loan as a primary source of financing, the reset is a welcome, but not necessary, feature. They’d apply even if it wasn’t offered again or if the precedents were considered an indicator of, but in no way a commitment to, future actions. Likewise, the ambiguity of “limited circumstances” won’t be a deterrent.
A potential applicant that intends to use the WIFIA loan commitment primarily as an interest rate management tool, however, might have a different perspective. They would want to see the precedents as establishing a de facto commitment to continue offering the feature and the “limited circumstances” as referring only to the fact that rates have fallen after a loan agreement has been executed. More like a real option, in other words. In the current rate environment, as discussed above, a reliable, hard-edged reset option might be an important factor in deciding whether to apply at all.
An Amendment to Attract Applicants?
The first resets were done in a reactive context — a pandemic caused rates to fall dramatically, which made powerful water agencies press for an adjustment to their loan commitments, which in turn prompted the program to re-execute their loans at the lower rate.
This time, the perspective on resets can be forward-looking. Long-term UST rates are high and may remain so, but there are plenty of reasons to think they may fall dramatically in the next few years. And many WIFIA loan commitments executed in the meantime will again be underwater, as in 2020. Potential demand is entirely predictable.
Of course, the WIFIA program and its stakeholders can simply wait for events to unfold. If rates fall, affected borrowers can be expected to request a reset, and the program can decide at the time whether the 2020 precedents apply. As described above, applicants for whom a WIFIA loan is a primary financing source will probably find such a passive approach acceptable enough.
But potential applicants that see WIFIA’s interest rate option features as the primary reason to apply may have a very different view. In current conditions, a reliable and relatively unlimited rate reset might the deciding factor in an application. If the reset feature were made more certain, both in terms of availability and specific limitations, an increased number of WIFIA loan applications would presumably result. This is especially true in light of WIFIA’s less-than-compelling current story relative to tax-exempt debt, as discussed here and here.
It’s easy enough to make the reset a statutory provision like the rate lock by a short amendment. That would give the feature a high level of certainty, enough for measurable option value in any case. The draft language at the top of this post is on the hard-edged side, but other versions could work too.
Bigger Questions
This scenario raises bigger questions that in theory ought to be addressed — does WIFIA want the type of options-focused applicant described above? That plans to build the project anyway and is just using the program as a source of financial subsidization? Where there’s actually no additionality with regard to US water infrastructure? Is this what the program is for?
In practice of course, none of these questions matter because no one will even want to admit that they’re worth asking. Any federal program wants more qualified applicants adding to their track record, and WIFIA is no different. Policymakers have no reason to disturb a narrative that appears to be working. And WIFIA stakeholders — especially options-focused potential borrowers — will be happy enough to see expanded capabilities, and policy details aren’t their concern.
There is one area where the questions matter in a practical way, however — predicting the future growth trajectory of federal infrastructure loan programs. Will their actual additionality be primarily in connection with real-world infrastructure? Or sophisticated and largely off-budget financial subsidies? How things develop for WIFIA’s reset going forward might shed some especially interesting light on that
[1] There is perennial discussion about reversing or softening these restrictions, but in light of current fiscal constraints and bitter spending battles, I’m pretty sure that anyone planning a major infrastructure financing will be assuming that they’ll remain in place for the foreseeable future.
As outlined in a previous post, a 55-year post-completion loan term might be an especially useful feature for CWIFP’s initial implementation. This post expands on the idea that limiting the scope of an amendment to extend WIFIA’s maximum may be critical to its chances.
Competing for High Quality Applications
The primary reason that the WIFIA program was surprisingly successful out of the gate was that very creditworthy entities submitted high-quality applications for low-risk projects. Applications like that get through the federal transaction processes and oversight bureaucracy with relative efficiency, quickly establishing an experience base of what works and a track record of reliably successful transaction completion. That kind of successful launch is self-reinforcing, internally by continually improving processing efficiency and externally by attracting further high-quality applications.
Contrast WIFIA’s story to that of TIFIA, a program that started with applications from complex and risky project financings. Those kinds of deals are hard enough for the private sector, but rough material indeed to feed into easily jammed federal machinery. And it showed in the program’s results and reputation.
However, there is one problem with high-quality applicants: they have very cost-effective financing alternatives in the tax-exempt bond market. Back in 2017, when WIFIA was starting up, a WIFIA loan with a 35-year post-completion term was competitive with 30-year muni high-grade water revenue muni bonds, primarily because the loan’s rate lock avoided negative arbitrage costs. That’s not the case in current market conditions, as outlined in the prior post.
A Quiet Veto
A 55-year loan term would dramatically improve the numbers and be a game changer for WIFIA and CWIFP loans. Unfortunately, the municipal bond industry knows that, too. Their lobby will have a quiet word with the Decision Makers Who Matter, and the 55-year term will be quietly dropped, as it has been when proposed over the past several years.
Wait…a federally subsidized debt market that has an effective monopoly on US public infrastructure finance due to that subsidy is acting behind-the-scenes to protect said monopoly by curtailing the capability of another federally subsidized source of infrastructure finance, one that incidentally utilizes genuine federal comparative advantages in lending and costs the taxpayers less? Isn’t that…rent-seeking? Yes, I’m afraid it is. And?
A Limited — and Realistic — Approach
Back in the real world, a 55-year term that can apply to all qualifying WIFIA projects (as HR 5664 and past amendments have proposed) will likely be a non-starter because the scope will include many municipal water projects that would otherwise be financed with bonds. But if the scope was limited to specific asset classes that aren’t part of the muni market’s bread-and-butter, then relative indifference, if not leniency, might be shown. The chances of that are improved if it can be shown that WIFIA loans to the asset class are likely to enable projects that wouldn’t have happened otherwise, providing the bond market (or at least its PAB segment) a new possible opportunity to get some of the non-WIFIA 51% share of project capitalization. I’d guess that similar arguments were critical in getting a 75-year term for TIFIA’s P3s and other tough transportation project financings.
It seems to me that dam safety projects are that kind of asset class, at least in terms of relative infrequency and unimportance to the muni market. And that’s the class CWIFP must start with in any case, not just because of the specific funding the program has received for dam safety projects, but due to the draconian strictures of the current FCRA Criteria, which block pretty much everything else that Congress intended CWIFP to do.
A successful start is important for any enterprise, but especially for a government lending program. A lot of long-term aspirations can be built on near-term results. So, with the real-world binding constraints and short-term goals in mind, I think there’s a compelling case to limit the current 55-year term amendment to dam safety projects for which CWIFP has funding. The operative language might look something like the draft at the top of this post.
At the end of a prior post, I noted that the authors of the current Criteria ought to explain why the ‘or assets’ was omitted from the conclusion of the Background section. Or, more exactly in the context of an apparently smoking gun, to provide an alternative explanation to the obvious one, which is that the omission arbitrarily makes otherwise eligible non-federal cost shares in federally involved projects ineligible for WIFIA financing, and that’s the result they wanted.
Well, it’s only fair that the authors be given a chance to explain their side. Below is an email sent yesterday to the OMB contact noted in the Criteria’s Federal Register publication, cc’d to relevant published contacts at GAO and CBO. Naturally, I’m not expecting any reply, but if there is one, I’ll post that too.
