- Blog Post
- Blog posts represent the views of CFR fellows and staff and not those of CFR, which takes no institutional positions.
Students of bank restructuring quickly learn that a common tactic is to divide a troubled bank into a good bank and a bad bank. The good bank gets the good assets (performing loans, high quality bonds, etc), and the bad bank gets the bad assets. Deposits usually get assigned to the good bank—and thus get backed by the banks’ good assets. Bonds usually get a claim on the bad bank, the bad assets.*
Puerto Rico’s proposed restructuring of the Government Development Bank (GDB), at least to me, looks like good bank/bad bank—but in reverse.
A reminder: the Government Development Bank took in deposits from Puerto Rico’s sprawling public sector and made loans to the public sector. But not everything nets out: some parts of the public sector have funds on deposits at the GDB, while others have large loans outstanding. The GDB also augmented its lending capacity by issuing bonds to the market. The bonds were only tax-free in Puerto Rico, so they tended to attract more on-island investors than a typical Puerto Rican bond. But over time some on-island investors have sold to distressed debt specialists at a deep discount. The Government Development Bank consequently has a very diverse set of current creditors.**
In the restructuring, the Government Development Bank’s creditors will be divided into three groups, and each end up with different claims on the GDB’s underlying assets.
One group of creditors—depositors from the public sector (the retirement fund, the electric company, the tourism development fund, and so on, for a total of $1 billion)—will get what looks like a claim on the bad bank. The bad bank will have as its assets junior loans to Puerto Rico’s government and loans to the medical system. The realistic recovery on these assets is very low. Do not take my word for it: the GDB’s fiscal plan assumes that the Commonwealth’s appropriations-supported debt will not be paid.*** These deposits could, in effect, be wiped out.
The remaining creditors—$3.75 billion in bonds (including about $0.4b billion held by on-island cooperativas) and a bit over $0.5 billion in privileged deposits (from municipalities and “private” actors)—get to choose between two different types of claims on the “good” bank.
One part of this group—my guess is a group that will be composed of mostly on-island investors—will opt for a junior claim on the “good bank” in exchange for a 25 percent haircut.
And a final group of creditors—my guess a group that will include most of the distressed specialists—will opt for senior claim on the “good bank” in exchange for either a 40 or a 45 percent haircut.
And somewhat unusually the senior creditors of the good bank will get a much higher coupon than the junior creditors. The senior bonds with a 45 percent face value haircut get a 7.5 percent coupon, the junior bonds with a 25 percent face value haircut get a 3.5 percent coupon.
If all of the GDB’s creditors able to get a claim on the “good bank” opt for the senior bond, the new “good bank” would have about $2.4 billion in liabilities, and need to pay around $180 million in annual interest. If all these creditors opted for the junior bond, the “good bank” would owe about $3.2 billion, and need to pay about $110 million in annual interest.
This kind of deal might make sense for the junior bonds if the “good” bank actually had enough good assets to cover the bulk of its new bonds. But the “good” bank’s assets aren’t actually all that good—the good bank itself will have a lot of “bad” assets. As a result, the likely value of the junior claim on the good bank is very low.
What assets are going into the "good" bank?
Think of four broad sets of claims: the GDB’s loans to Puerto Rico’s municipalities, a bit of property (unless $100 million), the GDB’s strong (“constitutional”) claims on the Commonwealth, and a set of junior loans to various parts of Puerto Rico’s sprawling government, including a very large and very weak claim on the highway authority. All told these assets have a face value of about $5 billion.
The $1.75b in municipal loans (after netting) have a high anticipated recovery. The GDB’s fiscal plan considers these loans to be the GDB’s best assets, even though some municipalities are likely to themselves need to restructure their debt—as the GDB’s loans to municipalities are backed by some combination of the municipalities’ share of the island’s sales tax and the municipalities property tax collections.
Even if all municipalities are able to repay, the municipal portfolio isn’t big enough to cover all the senior bonds if most eligible creditors swap into the senior bonds. Who knows what the property portfolio will be worth—it doesn’t matter much as it is small. The return for the senior bonds seems likely to depend on the value of $170 million in “constitutional” general obligation bonds and $225 million of “constitutionally guaranteed” bonds issued by the Port of the Americas and $600 million or so in other loans to various bits and pieces of the government (the ports authority, the public building authority, and so on—general obligation bonds currently trade at 60 cents on the dollar, but that price is basically only consistent with the fiscal plan if other bondholders get almost nothing).****
The “good bank” also is getting $1.9 billion in loans to the highway authority. Technically, the unsecured loans are only $1.7 billion—but $200 million in highway bonds held by the GDB are currently in default as well. Highway gasoline tax revenue is subject to the "clawback" and things like license fee revenues are pledged to the bondholders—the unsecured claims on highways are likely to have some of the lowest recoveries in the entire restructuring. Absent these likely worthless loans, the assets in the “good bank” have a face value of just over $3 billion, and an expected value below that.
My rough read on this is that there just might be enough assets in the “good bank” to cover most of the claims on the senior bond, but anyone taking the junior bond is likely to get very little.
And that’s the rub. If the GDB’s assets are worth say between 35 and 45 cents on the dollar, it isn’t clear why it is in the interest of Puerto Rico to provide some GDB creditors a return of over 55 cents (taking into account the high coupon) and others close to zero.
Especially as the creditors at risk of getting close to zero are likely to be disproportionately Puerto Rican financial institutions and municipalities.
The public sector depositors are clearly making a sacrifice—they are giving up their claim on the municipal loan portfolio, the property, the constitutionally backed bonds, and some other potentially performing loans—in order to smooth the overall restructuring. The logic of that sacrifice can be questioned. The fiscal plan doesn’t assume any recovery on the public sector’s deposits at the GDB, but to the extent the GDB still has good assets, some parts of the public sector look to be leaving a bit of cash on the table.
And I worry that a lot of unsophisticated “on-island” investors will take the junior bond issued by the “not-so-good bank” and ultimately get close to nothing, while sophisticated investors walk away with the good collateral.
Only an investor that values face value rather than cash flow would take the junior bond. One such group of investors may be Puerto Rico’s cooperativas—small local credit unions (they are locally regulated, they are NOT backed by the national credit union association). They can value their exposure to Puerto Rico’s government at face value for regulatory purposes.*****
But swapping into the 75 cent bond is a false economy if there are no good assets backing the junior bond. I think Puerto Rico would be better off if the cooperativas went into the deal at 55 cents on the dollar (plus a high coupon) even if the associated upfront losses exceeded the cooperativas’ current capital. Puerto Rico should add a plan to fill the cooperativas’ capital holes into its fiscal plan (their regulator/insurer would need to be recapitalized through the budget).
And, well, the same point applies to any municipal depositor that chooses the high face value junior bond rather than the good senior bond—picking the junior bond potentially is a ruinous financial choice.
The alternative to this deal is simply giving all bonds and deposits an equal claim on the eventual recovery of the Government Development Bank’s assets.
I can see why those who get the first claim on the “good” bank's good assets are better off as a result of the deal, but those who are left with the second claim on the good bank, or with a claim on the bad bank seem likely to be left worse off.
Do not get me wrong: there is a need to simplify the GDB’s mess of interconnecting claims. A deal through Title VI (the mechanism for collective voting outside of bankruptcy in PROMESA) that essentially gets the bondholders off the balance sheet could help make the negotiation between the municipalities and the other public depositors and the rump GDB easier. But any settlement with the bondholders should be at a valuation that is close to the likely resolution value of the GDB’s full portfolio—and it seems at least to me that the proposed deal may give the creditors with the good sense to swap into the senior claim on the "good bank" more than that.
Technical note: all the numbers here come from the Restructuring Support Agreement, which is available on EMMA’s website (EMMA aggregates information on the municipal bond market). Schedules 7 and 8 set out the liabilities “public sector deposits” and the assets “public entity loans” of what I called the bad bank. Schedule A lists the GDB’s bonds and schedule 1 lists the deposits (after netting) able to get claims on the good bank. The assets of the good bank for in schedules 4 and 5. Please email me if any of my numbers seem off.
*There are some legal wrinkles—depending in large part on whether the deposits formally have priority over the bonds—but that is the idea. Even if deposits and bonds have equal rank bonds can be assigned to the bad bank so long as they ultimately recover a sum equal to the value of the banks’ underlying assets in liquidation.
** It should be noted that the public sector in Puerto Rico is both a creditor to the GDB—it has lent the GDB money, by putting the public sector’s deposits at the GDB—and a borrower from the GDB. In the analogue to bankruptcy for banks (“resolution”), the government would both have a claim on the GDB’s estate, and the GDB’s estate would have a set of legal claims on different parts of the commonwealth. On net though the GDB is a creditor to the Puerto Rico’s public sector. Ballpark, the GDB now has $3.5 billion in deposits and $3.75 billion in bonds, according to its fiscal plan. Some of the deposits are escrow accounts against certain loans, and other deposits and loans have been netted out. As a result the restructuring seems to cover about $5.5 billion in bonds and deposits.
*** There are different categories of claims on Puerto Rico. The commonwealth itself has constitutional “general obligation” bonds, and appropriation supported debt (the appropriation supported debt had no effective remedy in the event of default even in the absence of PROMESA’s title III, it clearly is a junior claim). Puerto Rico also has pledged various revenue streams to back particular bonds, though most these pledges can be clawed back to support the central government and thus these bonds are effectively junior as well. However, the sales tax backed bonds are exempt from clawback—though that is currently the subject of litigation. Puerto Rico’s debt structure is unquestionably complex.
**** I have left out $300 million in loans with proceeds assigned to the issuer (schedule 6 of the restructuring support agreement).
***** In October of last year the previous Puerto Rican government estimated that one third of the GDB’s bonds are held on island, with about $400 million of the GDB’s bonds held by the cooperativas (see p. 71).