Stay up to Date
Get our weekly insiders guide to political and policy news (Saturdays)
It was 2005, and the recession hadn’t hit yet. Most taxpayers didn’t have opinions on financial derivatives. Few imagined major banks could ever go belly up. Even government agencies began drinking the Kool-Aid.
The San Diego Association of Governments was no different. Looking to maximize the possibilities for TransNet, its newly passed sales tax measure, the public agency did what others were doing: It played around with sophisticated financial arrangements that few understood.
SANDAG bet big that interest rates would go up. Instead, rates went down and stayed down – they’re still down. That unforeseen event – persistent and historically low interest rates – cost the agency millions.
As a result, SANDAG now has a roughly $100 million liability hanging over its head. It’s already spent $3.5 million out of pocket that it didn’t anticipate. And it spent $22 million to get out of a portion of its bad bet using borrowed money that will end up costing $42.5 million to repay.
All of that taxpayer money was supposed to go to regional transportation and infrastructure projects.
But agency leaders say everything is fine. The deal is working as intended.
Financial advisers came to SANDAG in 2005 with an opportunity.
Interest rates were low – very low. There was a consensus among analysts that they’d surely rise soon. After all, they always had. The advisers said SANDAG should bet on this proposition.
The proposal was simple. While issuing $600 million in debt to build new regional projects like highways, trolley lines and habitat preservation, SANDAG would make three 30-year side bets with Wall Street banks Goldman Sachs, Bank of America and Merrill Lynch.
The deals, called interest-rate swaps, were growing in popularity at the time.
In this case, SANDAG would borrow money with a variable interest rate that fluctuates with the market. At the same time, the borrower would agree to separately pay other banks a fixed interest rate on the debt, even though those banks wouldn’t actually lend the money. This fixed rate would be lower than SANDAG would be able to get in a standard bond sale. In return, those same banks receiving fixed-rate payments would pay the borrower a variable interest rate tied to a different market benchmark than the one the borrower agreed to pay the first group it had actually borrowed money from.
SANDAG would pay twice, and get paid once. The idea is that the two variable payments – one coming into the agency and one going out from the agency – would cancel each other out, and it would be left with just an attractive fixed-rate payment to the banks in the swap deals.
That’s the idea.
It didn’t work that way.
Swaps cost public agencies millions of dollars. Now they’ve largely disappeared in the public sector.
“What happened is they got burned by an extreme event no one envisioned: interest rates near zero,” said Steve Malanga, senior fellow at the Manhattan Institute for Policy Research, after reviewing some of SANDAG’s financial disclosures. “That’s why you don’t make speculative bets with taxpayer money.”
When SANDAG made these deals, it was betting rates would go up. It was locking in a low rate for 30 years in case that happened. And, if rates did go up, the deals would turn into an asset and SANDAG might be able to make money selling them.
The flip side, though, is that rates could drop and stay low, and SANDAG would have a liability on its hands. Ultimately, that’s exactly what happened.
On top of that, the variable rates that were supposed to cancel out didn’t, and from 2008 to 2016, SANDAG paid $3.5 million more to its bondholders than it received in the swap.
That’s $3.5 million that could have been spent on the types of capital projects TransNet funds were meant to provide, like dedicated bike lanes connecting Hillcrest to Balboa Park that cost exactly $3.5 million.
“Effectively, they have lost the bet,” Andrew Kalotay, owner of a New York-based debt management advisory firm who testified before the Security and Exchange Commission in 2011 about the dangers of swap deals, said of SANDAG’s swaps. “The timing was terrible … I absolutely guarantee they were hosed.”
SANDAG officials don’t see it that way. The goal was to lock in a low interest rate, they say, and that’s what they did. The problem is that’s not all they did.
Unlike most municipal bond deals that can be refinanced when interest rates fall, the terms of SANDAG’s swap contracts don’t allow renegotiation.
“Once you enter into a swap, if you want to unwind it, it’s going to be very expensive,” Kalotay said. “These deals were never advisable.”
There is one way out of the deals before the term is up, in 2038, when the bond debt is fully repaid.
To terminate the 30-year swaps early, one side must pay the other side the current fair market value of the swap contract. That could mean millions or even hundreds of millions of dollars.
Holding onto a losing swap deal means continuing to overpay the banks, and continuing to face the risk of a huge termination payout. That risk can tie up funds that would otherwise be available for projects.
At the time SANDAG did the deal, its swaps had no value at all. If interest rates went down, the swaps would have a negative value and become a liability. If rates went up, so would the value of the swaps and they’d become an asset.
When PFM Financial Advisors presented the deals to the SANDAG board, they emphasized the latter.
“Swap agreements become assets that gain in value as interest rates rise,” said a presentation made to the SANDAG board of directors in November 2005. Had interest rates risen just 1 percent from November 2005 to March 2008, SANDAG’s swap deals would have been worth more than $33 million, the board was told in 2005, meaning banks would have to pay SANDAG that much to end the swaps.
The fair market value of SANDAG’s swaps – which were agreed to in 2005 and took effect in 2008 when the first round of new TransNet bonds was sold – has fluctuated over time, but public records show the agency has been on the losing end from the start.
The value of the 2008 swaps dropped to negative $126 million last fiscal year, an amount large enough to replace bridges at San Onofre, the Batiquitos Lagoon, the Los Peñasquitos Lagoon and Rose Canyon. Recent rate increases have helped recover some value.
By January of this year, the swaps were worth negative $90.5 million, according to recent SANDAG board materials, still more than the cost of replacing the San Diego River Bridge.
This March, SANDAG’s financial adviser assured the board – which oversees a $1.4 billion budget – it shouldn’t be worried about that liability.
“I wouldn’t expect us to act on that in the near future,” Peter Shellenberger, managing director of PFM, told the board.
SANDAG staff echoed that sentiment last month, without ruling out the possibility of an early termination at some point.
The negative value “does not have any impact on the agency’s budget unless SANDAG decides to terminate the swaps, and its fluctuations do not affect debt payments over time,” Helen Gao, then a SANDAG spokeswoman who has since left the agency, wrote in an email. “Currently, SANDAG has no plans to terminate the swaps, as they continue to serve their purpose of providing budget certainty over the long term at a reasonable rate. However, SANDAG always monitors the market for advantageous opportunities and will continue to do so moving forward.”
Efforts to downplay the liability, though, ignore the ongoing risk the swaps pose if SANDAG is forced to terminate. That could happen if certain credit ratings aren’t maintained by all parties, if SANDAG or the banks in the swap file for bankruptcy or if other requirements aren’t met.
SANDAG’s continued certainty that it isn’t a concern is also, in essence, the same bet the agency made at the beginning: that rates will surely increase soon, at which point the swap value will rise.
“We are making a bet with Goldman Sachs that we know better than them, that we know which way the market is going to go,” said Malanga, with the Manhattan Institute. “Should we be making these bets with sophisticated banks like Goldman Sachs?”
But where Goldman Sachs is a private entity, which answers to its shareholders if a bet turns bad, SANDAG is a public agency.
“You are betting with public money,” said Lisa Washburn, managing director for the Massachusetts firm Municipal Market Analytics, who formerly worked at Moody’s Investors Service. “As long as the swap liability persists, it’s continued evidence of the bad bet they made before.”
In recent years, when SANDAG has gone back to the bond market to borrow money to build more projects, it has opted for old-fashioned fixed-rate debt. Altogether, swaps and related variable-rate bond debt now account for 23 percent of all the agency’s debt, Gao said.
SANDAG’s swap liability could have been worse today, if not for a big move in 2012. But reducing its swap headache came at a price.
In 2012, SANDAG terminated a portion of its swap contracts while refunding some of its old bonds.
SANDAG sold a new round of TransNet sales tax bonds with fixed interest rates totaling $420 million, and used a portion of the new money to refund some of its old 2008 debt. Bond documents also show nearly $22.6 million was used to terminate a portion of the initial swap contracts.
That means nearly $23 million that could have gone toward transit projects went to Bank of America and Goldman Sachs instead, simply to get out of some of the swaps. That’s roughly equivalent to the cost of dedicated bikeway projects to connect Old Town and Hillcrest, or in Pacific Beach along Rose Creek, or building a trail system along the San Diego River near Santee.
But it gets worse. Because the money came from new debt that will be paid back 1.88 times when interest is factored in, the actual cost to taxpayers for the swap termination is nearly $42.5 million.
SANDAG officials said the payout was cost-neutral, because the bond debt refunding resulted in savings. Gao even argued “It would be misleading to add $3.5 million to $22.6 million and assert that the swaps have cost SANDAG $26.1 million.”
But multiple experts familiar with swaps said the termination fee is absolutely part of the cost of the swap, and since SANDAG used debt, adding the repayment multiplier is an accurate representation of the true cost to taxpayers.
“SANDAG achieved savings through a fixed-rate refunding of a portion of the Series 2008 variable rate bonds, and the savings from the refunding were then used to fund the swap termination costs on a portion of the swaps,” Gao said. “As a result, the termination of the swaps was cost-neutral for SANDAG, meaning the partial swap termination did not increase the agency’s annual debt service.”
No matter how you slice it, the termination payment was SANDAG paying up its end of a lost bet.
Swaps were sold to government agencies as a way avoid risk and save money. Many didn’t work out that way.
When the market soured, some government agencies locally and beyond found themselves on the losing side of swap deals (like SANDAG) and terminated all of their swaps (unlike SANDAG).
“Since the credit crisis began in 2008, failed swaps deals have cost more than $4 billion as hundreds of borrowers, from the Bay Area Toll Authority in California to Harvard University, quietly pay Wall Street firms to end interest-rate swap agreements,” according to a 2010 Bloomberg report called “A Wall Street Gimmick That Soaks Taxpayers.”
Close to home, San Diego County paid Morgan Stanley and Citibank $22 million to terminate its last interest rate swap contracts in 2008, which were tied to 2002 pension obligation bonds. The termination fee came from the county general fund.
Going forward, “There are no immediate plans to issue any variable rate debt and no consideration of entering into an interest rate swap,” county spokesman Michael Workman wrote in an email.
The San Diego Metropolitan Transit System paid UBS $3.24 million to get out of an interest rate swap tied to 2004 pension obligation bonds. The termination fee came from the operating budget.
“As the bond market began to deteriorate in 2008, MTS leadership recognized both the variable bonds and the interest rate swapped tied to them were no longer good deals,” MTS spokesman Rob Schupp said in a statement. “And there are NO plans to have variable debt in the future.”
When PFM was marketing swaps to SANDAG and other governments before the recession, it presented charts showing the growing popularity of swap deals by businesses and public agencies. In March of this year, Shellenberger again emphasized to the board SANDAG was not the only government in California to make swap deals.
That’s true. But since the financial crisis, the number of municipal swap deals has plummeted.
SANDAG, however, has kept the bulk of its swaps – with their outsized monthly payments to banks on top of debt payments to actual bondholders – even as money runs dry for new projects.
SANDAG leaders did take another step to try to mitigate the damage years ago, though, and turned to yet more swaps for help.
After the financial crisis, SANDAG gave swaps another try in March 2009 and entered into two more deals that may turn out to be winners.
Those deals, called swap overlays, don’t take effect until 2018 but are currently assets.
Thanks to the favorable terms of the swap overlay deals, “the only way SANDAG would be worse off is if tax-exempt rates are greater than taxable rates,” an extremely rare occurrence in the marketplace, then-finance director Renee Wasmund explained to the SANDAG board in 2009, according to meeting minutes.
SANDAG spokesman David Hicks said in an email last month SANDAG expects the new swaps to bring in more than $1 million annually from 2018 to 2038 “assuming current market conditions.”
The overlays had a positive fair market value of $10.8 million in fiscal year 2016, meaning SANDAG would get paid that much if those swap contracts were terminated.
With the new swap contracts factored in, the fair market value of SANDAG’s entire swap portfolio totaled negative $115.5 million at the end of last fiscal year. It was better in January, at negative $73.5 million.
Stewart Halpern chairs the volunteer taxpayer oversight committee that monitors TransNet, and has a background in public finance. He declined an interview, but wrote in an email the swap liability does not worry him.
“Since the intent is to maintain the position until maturity, I am not concerned about the current valuation as long as the (swap) counterparty remains creditworthy,” Halpern wrote.
But Halpern also said public agencies need to exercise extreme caution when investing in complex financial instruments like interest-rate swaps.
He pointed to the collapse of Long Term Capital Management, a hedge fund run by Nobel Prize winners that, as chronicled in the book “When Genius Failed,” lost over $4 billion of its capital within one year, triggering a Federal Reserve bailout.
“Even Nobel Prize-winning experts in derivatives’ valuation have difficulty pricing the risk of highly unlikely events, so I believe that, if instruments such as swaps are to be used at all, a very conservative approach should be taken,” Halpern wrote.
When SANDAG approved the deals, there was a recurring sentiment among the elected officials on the agency’s board: This is nearly a risk-free proposition.
The idea was that SANDAG in some ways couldn’t lose.
Current board chairman Ron Roberts declined to discuss the swaps and did not answer several specific questions sent by email.
Roberts issued a written statement praising SANDAG for its AAA credit rating and for using swaps “to lock in a favorable interest rate of about 4.2 percent including fees.”
Roberts did not acknowledge the looming liability the swaps represent, or that SANDAG was overpaying on its swap deal at 4.2 percent. When SANDAG made its bet, financial adviser PFM said the agency would achieve a 4 percent interest rate. With debt as high as $600 million, that seemingly small difference means millions more is spent on interest than planned.
PFM did contemplate things not going exactly according to plan.
In a 2005 memo, PFM gave an “extreme example” and explained SANDAG could have to pay $55 million to the banks to get out of the swaps if interest rates decreased 1 percent.
“Given historical performance, we would not expect rates to reach this low level,” the memo said.
As we now know, PFM’s “extreme” was not the worst it could and would get.
At a November 2005 board meeting approving the swaps, then-SANDAG board Chairman Mickey Cafagna likened them to an insurance policy. If rates go up, the banks pay SANDAG. If rates go down, SANDAG has to pay to terminate the contracts, but can then renegotiate its debt at historically low rates.
“If interest rates drop, and we have to buy out of this for $55 million, it’ll probably be the best thing that has happened to us, because we’ll have tremendous savings,” he said at the board meeting. “In my opinion, it’s fairly risk-free, if you can have anything that’s risk-free – there certainly is risk here. I think it’s a tremendous opportunity.”
Rep. Scott Peters, then a San Diego city councilman on the board, felt similarly.
“The implication of that $55 million termination is that it wouldn’t matter in a low interest-rate environment anyway, based on the amount of money we’d be able to throw around at those low rates,” Peters said at the same board meeting.
Even in a low interest-rate environment, the scenario described by Peters and Cafagna is far from guaranteed.
Market conditions at the time of the refunding could make it impossible, and SANDAG would have to structure the new debt in a way that saves enough money to cover the termination, and doesn’t just push the debt down the road. Even then, you’d have to be OK with sending millions of dollars to Wall Street for swap terminations instead of paying for transportation projects, making payouts voters didn’t approve.
“It’s one thing when you’re betting with your own money,” said Malanga, with the Manhattan Institute for Policy Research. “It’s another when betting with taxpayer or ratepayer money.”