Eric A. Jordahl
Hospital and health system finance leaders can lower debt-financing costs through intelligent use of variable-rate products appropriate to the organization's credit and risk profile.
At a Glance
- Hospital and health system finance leaders should position their organizations to participate in the variable-rate market.
- To this end, one important step is to establish the right baseline variable-rate exposure target for the organization based on its credit and risk profile.
- Leaders also should be thoroughly familiar with the available products and understand the circumstances (pricing, terms, and embedded risk) under which the organization would be willing to deploy them within the overall capital structure.
Because it has an attractive risk profile-including ongoing predictability of cost-fixed-rate debt issued in the municipal market represents the primary form of external capital used by U.S. not-for-profit hospitals and health systems. However, variable-rate debt, with interest rates that change over time, has been used for decades by these same organizations to build balanced capital structures that create the possibility for lower-cost funding.
History supports the expectation that variable-rate instruments will cost less over time than fixed-rate alternatives. The exhibit at the top of page 81 illustrates the general differential between rates. Although the relationship widens and narrows in response to specific market and economic forces, the clear expectation is that variable exposure will produce a lower cost.
Risks and Rewards of Variable-Rate Exposure
The Federal Reserve's injection of a massive amount of liquidity into the global financial markets in the years since the credit crisis has kept interest rates artificially low, yielding an extraordinary benefit to those hospitals with net variable-rate exposure in their debt portfolios. The Fed has indicated its intention to keep rates low through late 2014; unless global developments force the Fed to shift its stance, there is an expectation that variable-rate debt will continue to have a very favorable impact on debt costs.
Like fixed-rate debt, variable-rate debt is accessed through public or private debt instruments. The underlying interest rate "resets" on a periodic basis-either off of a defined index, such as the Securities Industry and Financial Markets Association (SIFMA) index, or on a "free-floating" basis that reflects investor demand for the particular obligation. Variable-rate exposure can be created by the debt instrument itself ("direct" variable-rate debt), or it can be added through a derivative instrument that sits alongside a fixed-rate obligation ("synthetic" variable-rate debt). Synthetic structures have been less attractive of late, but a higher rate environment may increase their appeal, making it important for hospitals to understand their application in effective capital structure management.
The risk-reward relationships vary significantly between fixed-rate and variable-rate debt and between different types of variable-rate obligations. When a hospital or health system issues a fixed-rate bond, it effectively transfers all of the main debt-related risks to the investor. In contrast, issuing variable-rate debt means the hospital is willing to retain certain risks, which singly or in combination may be significant over time or at specific points in time.
Choosing the right mix of fixed and variable-rate debt and the portfolio of variable-rate products that best fits within the hospital's cost and risk objectives is a high-priority, ongoing task for leadership. Leaders should carefully evaluate potential financing programs based on a clear understanding of available products and the associated risk transfer/retention dynamic. Exhibit 2 shows the broad risk categories and the general risk/reward relationships that are at play in the general fixed-versus-variable-rate decision.
Introduced in the early 1980s, variable-rate demand bonds (VRDBs) represent the "anchor" variable-rate debt product for hospital borrowers and, therefore, help to frame the risk-reward equation. Because the SIFMA index is a composite of weekly resets of a number of VRDB programs, this index is the benchmark off which an individual VRDB program is measured. As noted earlier, SIFMA reset rates have historically been well below natural fixed-rate alternatives; In fact, the average SIFMA value since 2009 has been 0.27 percent, whereas the 20-year MMD average for the same period has been 3.84 percent.
In exchange for securing access to this potential interest rate savings, the hospital borrower must assume two main types of risk with VRDBs: interest-rate risk and put risk.
Interest-rate risk. This is the risk that interest rates will increase, potentially to levels well above what could be achieved by selling natural fixed-rate bonds or by synthetically fixing the variable-rate exposure. Due largely to Federal Reserve actions lowering the rate at which banks lend to each other (the federal funds rate), the variable interest- rate environment has been artificially low for an extended time. The target rate has been 0.25 percent since December 2008.
Many indicators-most important, direct statements from the chairman of the Federal Reserve-suggest that policymakers expect this low rate environment to continue for the next several years. However, one need only recall the late 1970s and early 1980s to recognize that interest-rate environments are dynamic and can shift in fairly short order as policymakers and market participants respond to changing conditions. In the 1980s, the Federal Reserve drove the federal funds rate to 20 percent in a successful effort to contain double-digit inflationary increases.
Put risk. This is the risk that investors "put" (i.e., sell) their positions back to the borrower for payment at par, which they may need to do for any number of reasons. VRDB programs are built on the premise that investors have "guaranteed" liquidity, typically on a daily or a weekly basis. The first source of funding for a bond put back to the hospital borrower is to sell the tendered bond to a new investor. But if a new buyer cannot be identified, a liquidity backstop is required.
A hospital or health system with particularly strong liquidity might promise to fund any puts out of its own cash resources (i.e., through self-liquidity). This approach has become somewhat more common for stronger credits for at least some portion of their VRDB portfolio. But self-liquidity creates the potential for a substantial overnight claim on cash, so most hospitals either have to or choose to use a commercial bank as their primary source of liquidity in the form of a standby bond purchase agreement (SBPA) or a letter of credit (LOC).
Through such agreements, the bank promises to purchase non-remarketed bonds and to carry them until they can be remarketed or until the hospital repays them on an agreed-upon accelerated schedule. The introduction of a bank moderates the direct put risk, but introduces a range of other risks that need to be understood and planned for. Importantly, the bank agreement mitigates an immediate claim in favor of an accelerated repayment of those variable-rate bonds that are purchased by the bank.
To secure access to lower-cost variable-rate debt through a VRDB, a hospital or health system must be willing to tolerate the risk that interest expense might be materially higher than expected or that there could be an accelerated claim on cash to pay off bonds that have been put and cannot be resold to new investors. Between these two risk buckets, the accelerated repayment is the most problematic. Higher-than-expected interest rates might be disruptive, but being forced to pay off variable-rate debt in three years, for example, might threaten the organization's survival.
Alternatives to VRDBs
Every product developed in the healthcare variable-rate markets since the 1980s has ultimately been about trying to eliminate or at least reduce one or more of the event risks attached to VRDBs.
One set of alternatives to VRDBs emerged with the introduction of derivative products, specifically fixed-receive swaps and total-return swaps. Both of these strategies "synthetically" manufacture variable-rate debt by placing a separate derivative instrument alongside a natural fixed-rate bond.
Fixed-receive swap. This swap converts a hospital borrower's fixed-rate debt into variable-rate debt. A borrower with fixed-rate debt contracts with a swap counterparty to provide that counterparty with variable-rate payments (at a stated index, such as SIFMA) over the life of the swap in exchange for receiving fixed-rate payments from the counterparty. The net funding rate achieved by the hospital is the variable swap index plus the difference between the fixed-rate debt cost and receipts on the fixed-rate swap. See the exhibit below.
The base funding rate is comparable to a VRDB with a bank facility (trading at SIFMA plus the bank charges plus other program fees), but the risk metrics are quite different. Bank-related cash flow or event risks are replaced with swap-based risks; as an example, the risk that collateral might need to be posted on the swap replaces ongoing VRDB put risk. A fixed-receive strategy is not particularly attractive today given very low swap rates versus relatively higher bond coupon rates, but the appeal may shift with a transition to a higher-rate environment.
Total-return swap. This swap also creates direct variable-rate exposure. A borrower sells fixed-rate bonds to a trust that finances the purchase through the sale of variable-rate notes. Under the terms of the total-return swap, the hospital receives its coupon rate on the fixed-rate bonds and pays a variable rate (typically the SIFMA index) plus a fixed spread, which reflects the hospital's credit, the trust's funding costs, and/or other factors.
One risk associated with the classic total-return swap is a potential "make-whole" provision at the unwinding of the program, meaning that a settlement payment might need to be made to close out the position. Further, collateral posting against this potential unwind liability also might be required while the program is in place. Although the structure has improved over time (with advances in trying to minimize the unwind exposure and related collateral posting requirements), the total-return swap continues to have somewhat limited use due to its complexity and perceived risk as well as the limited number of counterparties that continue to offer the product.
In certain environments, swap-based variable-rate strategies can provide an exceptional cash flow opportunity. We suspect that hospitals issuing fixed-rate debt today will be considering fixed-receive swaps once we move to a higher-rate environment. Shifts in the availability, costs, or risks associated with VRDBs or other variable-rate products may also increase the appeal of fixed-receive or total-return swaps. Organizations will need to determine whether the "event" risk profile associated with a swap structure is more or less acceptable compared with VRDBs or other variable alternatives.
Interest-rate swaps have been battered in recent years, but these products should be reviewed and reassessed, particularly if the risk and/or cost of bank-backed VRDBs become unacceptable.
The auction-rate securities interlude. The most significant break from VRDBs came with the introduction of auction-rate securities to the municipal market. The product started slowly, but gained traction as focus shifted to its significantly enhanced risk profile versus the risk profile of VRDBs.
Two risk considerations were of paramount importance.
First, the investor had no guarantee of liquidity. The only way one investor could liquidate an auction-rate holding was if another investor was identified to assume the holding. For the issuing hospital or health system, this meant there was no put risk-and also, no potential for an accelerated claim on cash and no need to introduce a commercial bank to mitigate this risk. Second, the bonds typically were secured by an irrevocable bond insurance policy that covered the full term of the debt from a triple A-rated bond insurance company. Investors had no exposure to the underlying healthcare entity, so the long-term expectation was that the underlying debt obligation would trade in very close correlation with the SIFMA index.
Equally important, the altered risk position provided a platform whereby hospitals and health systems became willing to enter into long-duration fixed-pay interest-rate swaps. The thinking was that the insured auction structure represented a "committed capital" foundation that could support a long-dated swap, especially since the extension of the triple-A insurance to cover the swap released the hospital from any obligation to post collateral, even if the value of the swap fell significantly.
The credit crisis in 2007-08 resulted in the collapse of the municipal bond insurance industry and the municipal auction-rate securities market-and led to debt- and swap-related stresses that hospitals continue to deal with today. The central presumption during the auction-rate era was that a range of key risks had been fully transferred to reliable insurance companies. As the financial position of these companies deteriorated, hospitals found themselves confronted by a fundamentally different (and far riskier) capital structure than what was originally intended.
A Return to VRDBs
Following the auction market collapse, the only potential variable-rate alternative for hospitals again was VRDBs. However, one challenge was that bank ratings were under intense scrutiny and stress, and many banks-particularly the regional firms that had long been important providers of credit and liquidity to smaller hospitals-were no longer acceptable to VRDB investors.
VRDBs are primarily purchased by tax-exempt money market funds, which must comply with strict regulations related to the credit and/or liquidity of what they can purchase. These funds focus primarily on yield rather than risk, meaning that unlike a mutual fund purchaser of long-term fixed-rate paper, they may be less likely to hold onto a position once trouble hits. So after the credit crisis, bank capacity was leaving the VRDB market at precisely the moment demand for bank liquidity was increasing. The capacity that remained was more often than not targeted at higher-grade hospital and health system credits, creating yet another level of stress for smaller and less creditworthy hospitals.
Concerns about the integrity of bank credits have been more or less intense at different points since the credit crisis. Ratings at many banks, including some of the most active providers of credit and liquidity support to the municipal market, have continued to deteriorate, and concerns about the reliability of banks as a long-term partner in a public VRDB program have increased.
A detailed description of current issues with bank-supported VRDBs appears in a web extra to this article.
Direct Bank Placements
Partly in response to these concerns, but also due to the significant amounts of capital currently available for lending, a number of banks started providing a "direct purchase" alternative. Under a baseline direct-purchase structure, the bank essentially makes a tax-exempt loan to the hospital or health system. The loan has a stated length-anywhere from three years to more than 10 years, depending on credit quality. Pricing is established as an index rate-typically a stated percentage of one-month LIBOR (London Interbank Offered Rate)-plus a fixed credit spread that may vary over the term of the loan based on the borrower's credit rating.
Use of direct bank placements has increased in the past two years, primarily because they offer hospitals access to a variable rate without introducing put risk, and also because they are private transactions that price off of an index, in contrast to VRDBs that trade in the public markets at rates that reflect the credit quality of the bank.
As with any product, the devil is in the details. There are certain business and credit terms that can have a major impact on the net risk position attached to a direct-lending structure. Nevertheless, the product has gained a lot of traction across the healthcare landscape and is likely to remain popular as long as banks continue to make it available.
The web extra to this article provides an additional variable-rate option-windows VRDBs-that has been used selectively, with very good pricing results, by certain double A-rated healthcare credits. It also provides a chart comparing the risks and benefits of direct bank placements versus VRDBs.
Positioning Your Organization to Participate in the Variable-Rate Market
The current state of the tax-exempt variable-rate market is mixed. It remains a market that is highly dependent on bank financing and one where all types of alternatives are limited, especially for weaker credits. It is therefore important for healthcare organizations to track whether these and other more established structures emerge as robust alternatives to VRDBs. The emergence of VRDB alternatives will depend on how VRDB pricing and risk metrics shift over time under the influence of factors such as bank credit quality and regulatory developments.
In all scenarios, it will remain important for hospital finance leaders to actively position their organizations to participate in the variable-rate market.
To this end, the finance leader should ensure that the right baseline variable-rate exposure target is established for the organization based on its credit and risk profile. This step requires an understanding of the costs and benefits of risks related to variable-rate exposure versus other risks (operating or asset-based) that the organization might be forced or choose to assume.
The finance leader also should be thoroughly familiar with the available products and know the circumstances (pricing and terms) under which the organization would be willing to deploy them within the overall capital structure. This knowledge involves understanding what risks are embedded in various products, how they relate to other risks across the organization, whether these risks can be managed or mitigated, and if so, how.
Getting to the right answer will require focus and effort-but the result will be a debt portfolio that makes intelligent use of variable-rate products to secure a lower cost of debt financing.
Eric A. Jordahl is a managing director and co-leader of the financial advisory practice, Kaufman, Hall & Associates, Inc., Skokie, Ill. (firstname.lastname@example.org).
Risks Incurred with Bank Facility-Supported VRDBs
Hospital credit risk. The risk that a decline in the hospital or health system credit rating might result in an increase in the bank facility fee, and a corresponding increase in the net variable-rate funding cost.
Bank credit risk. The risk that a decline in the bank's credit rating might lead to an increase in the variable interest rate on the bonds, or worse, an increase in tender activity on the bonds.
Regulatory risk. The risk that a shift in regulations, such as the pending Basel III requirements, might increase the bank fees for offering a particular product.
Acceleration risk. The risk that the bonds might be purchased by the bank and begin to bear interest at the "bank bond rate," which is typically significantly higher than what is available in the public market.
Bank renewal risk. The risk that the bank facility might not be renewed at its scheduled expiration date, in which case, if no replacement could be found, the bonds would be subject to mandatory tender and conversion into a term loan with a much shorter amortization than would likely be available under the variable-rate bonds.
Acceleration risk. The risk that the hospital or health system might breach a covenant within the bank agreement, allowing the bank to terminate its obligation to provide liquidity, in which case, if nothing were done in response, the inevitable result would be a mandatory tender and conversion to a term loan.
Publication Date: Tuesday, May 01, 2012