Would your organization issue bonds due in a hundred years? That's what New York and Presbyterian Hospital and Cleveland Clinic did in the last 24 months. Locking in interest rates for the next century may be tempting to some, but ultra-long bonds could prove costly if called.
Last June, New York and Presbyterian Hospital sold $850 million in taxable bonds. $250 million was issued as interest-only with principal due in 2116, a hundred years from now.
NYP was the second healthcare provider to issue century bonds. In 2014, Cleveland Clinic sold $400 million maturing in 2114 on similar terms.
Century bonds and other ultra-long bonds can be attractive to investors such as pension funds and life insurance companies looking to match liabilities with long-duration assets.
Ultra-long bonds also earn higher yields. NYP paid a 70 basis points (0.70%) premium over the 40-year yield sold on the same date.
In the US, close to 70 century bonds have been issued since the structure was first used by Walt Disney and Coca-Cola in 1993.
In the municipal sector, activity has been mostly in higher education.
For NYP and Cleveland Clinic, century bonds provided an opportunity to lock in interest rates for a very long time, and at a relatively low premium.
In doing so, these borrowers looked past one of the basic tenets of finance: don't leverage an asset past its useful life. Ultra-long debt sets the stage for debt to be outstanding long after any assets have exceeded their useful lives.
But century bonds present a much greater downside: they can be very expensive to call.
It's safe to say issuers of century bonds are not planning to call them anytime soon, and some may not care what legacy they leave to future management.
But a hundred years is a long time, particularly in healthcare where consolidation among providers is accelerating, and when organizations merge or consolidate balance sheets, most bond documents require outstanding debt to be paid off.
When called, century bonds are subject to a make-whole payment equal to the greater of par and the present value of all remaining debt service payments.
The present value is based on a discount rate equal to the Treasury yield for a maturity similar to the remaining maturity plus a small premium (for NYP, the premium was 35 basis points).
The complication arises when the bonds are called more than 30 years from final maturity (within the first 70 years from the sale date) and there is no actual yield available because Treasuries only go out to 30 years.
When a yield is not available, the call provision language says to use an assumed yield quoted by a primary dealer based on how corporate ultra-long bonds price for a similar maturity.
This provision leaves the dealer –who has zero fiduciary duty to the borrower– with complete discretion on what discount rate to use, not a good situation when the make-whole amount is highly sensitive to it.
As our chart shows, if interest rates rise 100 basis points (+1.00% parallel shift across the Treasury yield curve), the call premium tops at 20% or $50 million on $250 million par. This is not cheap compared to traditional call premiums which seldom exceed 102%, but it may still be acceptable to strong, well-rated borrowers.
But if rates stay flat, the premium goes up to 40% of par or $100 million.
This is unlikely in today's low rate environment, but if rates were to decline by 100 basis points, the premium shoots up to 80% or $200 million.
We assumed a normal, upward-sloping yield curve. It gets even more costly if the dealer user a flatter yield curve instead.
The analysis reveals another oddity: in a normal, upward-sloping yield curve environment, the NYP call premium initially rises over time, which is the inverse of traditional call provisions.
So when looking into ultra-long bonds, healthcare providers should ask themselves how likely their organization is to be around in its current form in a hundred years, and if interest rates will rise enough to make a call scenario affordable.
Skipping the analysis could spell trouble for future management.