In late February, the WIFIA program closed its 100th loan, a $74 million financing for Santa Clara Valley Water District. This is the first part of a master agreement for future WIFIA loans totaling about $2 billion for the rehabilitation and construction of two reservoir dam projects. Since dam finance has been a topic here recently in connection with the new CWIFP program, this loan is worth a closer look to see how three possible CWIFP enhancements – a 75-year term, a 10-year deferral and the Limited Buydown – would be useful in a real-world situation.
As usual for large water public agencies, there’s a lot of public information about the decision making process. On December 13 last year, the SCVWD Board met to consider a final resolution to go forward with this WIFIA loan, in the course of which its potential benefits and costs were reviewed and discussed.
Here’s a link to the video of the meeting — SCVWD Board of Directors Meeting December 13, 2022. The main WIFIA presentation starts at 0:45:10 and finishes at 0:55:03. Discussion among the boardmembers goes to about 1:43:00 at which point the resolution to proceed is passed. WIFIA slide deck pdf here.
Possible Enhancements Could Be Included in Future SCVWD WIFIA Loans
Although much of the WIFIA presentation was in the context of the overall $2 billion WIFIA financing plan for two dams, SCVWD’s specific resolution and subsequent loan closing was limited to a $74 million loan for planning and design costs for one dam project. As many as nine additional WIFIA loans, closing between 2023-2032, are contemplated, but not committed, for the projects. This time frame means that possible enhancements to WIFIA program loans may be directly relevant to SCVWD’s dam program. Active legislative efforts, primarily connected to CWIFP implementation but specifically amending WIFIA law, are continuing. Last year The Water Infrastructure Finance and Innovation Act Amendments of 2022 (HR 8127) was introduced in the 117th Congress and is expected to be reintroduced soon. This bill already includes an amendment to extend WIFIA’s maximum term to 55 years. In addition, the Twenty-First Century Dams Act (HR 4375), also expected to be reintroduced, proposes an increase in WIFIA funding – amendments for enhancement (at least for dam projects) could be included in the same section. Obviously, nothing is certain in this political environment, but enacting minor, technical-seeming amendments to enhance the capabilities of a popular and successful water infrastructure loan program probably has a decent chance, if anything does. Thinking about how the enhancements might be useful to SCVWD is therefore not an academic exercise.
Apart from the financing, it’s clear from the board discussion that real-world plans for the dam projects are also far from finalized. The dam project for which the $74 million WIFIA loan was sought, primarily a seismic retrofit of the Anderson Dam, seems to be more or less greenlighted towards construction. But the other dam project, the Pacheco Reservoir Expansion, is still at an early (and apparently contentious) stage. The fluidity of the situation highlights the value of a WIFIA loan’s overall flexibility, as was pointed out several times in the discussion.
There’s another aspect of this, specific to the enhancements – to the extent that possible WIFIA loan enhancements influence the final design or other real-world aspects of the infrastructure, that would demonstrate that WIFIA’s fundamental policy objective (US water infrastructure improvement, not just subsidization) is achievable. For example, might a 75-year loan term focus attention on, and possibly enable, a plan for the Pacheco Dam that addresses very long-term sustainability, resilience and inter-generational issues? If the enhancements are enacted while SCVWD’s situation is still fluid, their impact will be interesting to see from a policy perspective.
Debt Service Management
A primary theme in the WIFIA presentation was the loans’ usefulness in managing future debt service in the context of SCVWD’s cash flow. In the longer term, this was about being able to customize a WIFIA loan’s amortization schedule and dovetail it with other financing by delayed principal payments. In the front end, the full deferral of debt service (both interest and principal) allows a smoother ramp up of rate revenue. As frequently discussed in prior posts, the ability to offer these features is a real federal government strength compared to private-sector debt markets, especially the tax-exempt public bond market. WIFIA’s US Treasury-based pricing makes delayed amortization cost effective due to the ‘flat-forward’ Treasury curve beyond thirty years. Interest accrued during the deferral is locked at the loan’s original rate. And the loan or any amount of it can be prepaid without penalty. The ‘flexibility’ of a WIFIA loan was often mentioned in the presentation and board discussion – I’d expand that description to include ‘optionality’ as well, a concept that includes potentially quantifiable value (e.g., in a refinancing situation) throughout the loan’s life.
The potential enhancements are simply that – a 75-year term, a 10-year deferral and the Limited Buydown improve the flexibility and optionality of WIFIA’s debt service management features without changing anything fundamental from the borrower’s perspective. They will expand the program’s maximum statutory authority, but don’t prescribe what the borrower can choose, or the program is willing to approve.
Presumably SCVWD’s dam projects have useful lives when completed of at least 75 years. As the WIFIA presenter noted, SCVWD’s credit rating for the WIFIA loan from Fitch was a near-perfect AA+. If the enhancements were enacted, my guess is that WIFIA would approve future SCVWD loans utilizing all three to their fullest extent – risks to the federal taxpayer would remain very low and (as discussed above) program policy objectives would likely be benefitted. Assuming that’s the case, we can consider how useful each enhancement might be in the context of SCVWD’s situation, as described in the WIFIA presentation and board discussion.
The below chart (page 9 of the pdf) summarizes the debt service requirements of about $2 billion of SCVWD’s planned WIFIA loans (blue bars) under current law. You can immediately see the back-end loading of delayed amortization. The 5-year deferral is a bit more subtle — without it, there would have been a sharper spike in debt service in 2028 than the one projected for 2033.
75-Year Term
SCVWD’s chart can be recreated to include an approximate model of the WIFIA loans using other information provided in the presentation. Assumptions can then be changed to reflect the enhancements. Here’s what a 75-year loan term would look like:
Unsurprisingly, there’s a big drop in level-payment debt service, more than $120 million annually, as the amortization period goes from 13 to 48 years. But this effect won’t start until 2055 — more than 30 years from now. Does it have practical value for SCVWD in terms of decisions that will be made in the next five years or so?
Depending on other SCVWD’s other capital plans and long-term projections, there might be a more or less direct application for the additional cash flow in the far future. But I think even outside that case, there’s always value in securing a very long-term amortization schedule for very long-lived assets like dams because it increases WIFIA loan optionality. Prepayments can always be made without penalty — SCVWD can opt to follow a 13-year amortization schedule or refinance the loans altogether if lower interest rates can be obtained. But a lot can happen in 30 years. If the need for additional financing for other projects arises, or projected revenues for some unforeseeable reason become significantly lower, $120 million of annual cash flow freed up by a 48-year amortization schedule on the WIFIA dam loans might become an important factor.
10-Year Deferral
If a 10-year deferral feature is included, WIFIA loan interest-only debt service simply starts in 2038 instead of 2033. The spike is still about the same — in fact, it’s very slightly higher due to capitalized interest. But now there’s an additional five years to ramp up rates or other revenue sources to pay it, something which may have relatively near-term value.
In any case, the same observations about increased optionality for the 75-year term apply to the 10-year deferral. Ramping up water rates is probably always a somewhat difficult and multi-factored process — having more time to accomplish it can only help. If things go better than expected, voluntary debt service payments are always an option.
Limited Buydown
When describing the loan’s projected interest rate in the WIFIA presentation, the speaker was careful to note that it was only an estimate and wouldn’t be locked until loan execution. In the current economic and financial environment, interest rate volatility is (and likely will remain) a fact of life. Rates can rise or fall quite significantly and unexpectedly in a relatively short time. In SCVWD’s multi-year planning horizon for the dam projects, big changes in this central factor are almost a certainty.
A Limited Buydown feature — currently included in the TIFIA and CIFIA loan programs but not yet in WIFIA — could help mitigate this uncertainty in SCVWD’s WIFIA financings, at least to some extent. The feature simply allows the program to execute a loan at the Treasury-based rate prevailing on the date the loan application was accepted (usually many months or even years earlier) if that rate is lower than current, with a maximum 1.50% decrease. In effect, the Limited Buydown is an interest rate cap that shields the borrower from rate increases (up to 1.50%) in the period between an accepted application and loan execution (possible draft language for the provision here).
I am fairly certain that each loan under the Master Agreement between SCVWD and WIFIA will require some form of separate execution and its interest rate will be set on the day that happens. The rate for the $74 million loan was locked in the February closing but future loans — comprising the vast majority of the $2 billion program — are presumably still exposed to interest rate volatility.
If WIFIA had a Limited Buydown provision, an interesting question arises — does the Master Agreement constitute an ‘application’ for all the individual loans contemplated to be made under its terms? A WIFIA Master Agreement might be only a conditional agreement with respect to funding loans (per CIFIA law definition anyway), but it certainly represents a high degree of ‘acceptance’ by the program of all the loans contemplated thereunder. Legal interpretations at this level of detail are way above my pay grade, but I’d guess that if the WIFIA statute had had a Limited Buydown provision when the program accepted SCVWD’s application for the financing plan under a Master Agreement, or when the Master Agreement was executed, then a Limited Buydown adjustment would be available, as of that date, for all its individual loans [1].
Let’s assume that’s the case for a simple illustration. Also assume the Limited Buydown rate is the same as estimated in the WIFIA presentation — 3.68% [2] — and that US Treasury rates for future loans under the Master Agreement rise by 1.50% to 5.18% before execution.
Without a Limited Buydown cap, forecast debt service would increase by $50 million annually, as shown in the red bars below. That’s about a 40% increase in WIFIA debt service during the interest-only period, and an 18% increase in forecast debt service overall during the same time.
With the Limited Buydown provision, debt service could remain, subject to program approval [3], as originally forecast. The assumptions for this illustration make the potential monetary value obvious, but the more general point is that the Limited Buydown reduces uncertainty of forecasting financing cost of large capital programs, at least for the WIFIA component.
_____________________________________________________________________________________________
Notes
[1] Some sort of ‘reset’ or re-application of previously executed agreements might be required. Discussed here in the context of the CIFIA loan program — WIFIA Rate Lock Reset Precedents for CIFIA
[2] WIFIA rates are set by using the US Treasury yield applicable to the weighted average life (WAL) of the loan. On 12/2/22, the UST 20Y and UST 30Y yields were 3.79% and 3.56% respectively — presumably the presentation’s 3.68% rate is an interpolation corresponding to a loan life of about 25 years? That makes sense for the $74 million initial loan, but the WAL for the overall $2 billion financing is closer to 40 years, due to the deferral and long interest-only period. Still — even if the whole $2 billion were executed at 12/2/22 rates, the ‘flat forward’ convention for Treasury pricing would result in using the UST 30Y rate of 3.56%. That detail illustrates some of the value of the 75-year term — once WAL is close to 30 years (typical for long-lived infrastructure financing) there’s no extra cost in going much longer.
[3] The Limited Buydown permits, but does not require, the program to reduce the execution rate. That decision might depend on the availability of discretionary appropriation funding. Discussed here, also in the context of the CIFIA program — CIFIA’s Limited Buydown – Mind the Gap. To be really effective, WIFIA will need a much higher level of funding from Congress. Well — why not? Current funding (about $60 million per year) is almost trivial. The program has demonstrated remarkable success and is achieving its objectives — now how about some real funding?