For as long as most people can remember, UK municipal finance has been safe and boring. In the wild days of the 1980s, Hammersmith & Fulham council almost went bust speculating in derivatives, and was saved by a landmark House of Lords ruling. Since then, UK council borrowing has been tightly constrained by central government while derivative use has been banned.
Then again, the finance industry loves to innovate around regulation. It turns out that UK local authorities are still involved in derivatives after all. Many years ago, bankers found a way of embedding the pesky contracts inside perfectly legal loans, using so-called lender’s option borrower’s option (LOBO) agreements. These bank loans attracted councils because they offered a cheaper rate than offered by central government.
LOBO agreements are typically only a couple of pages long. A local council – Newham or Camden for example – agrees to borrow £10 million from a bank, such as Barclays. The loan is very long term, repayable in 40, 50 or even 60 year’s time. However, starting a few years after the loan is taken out, the bank has the option to annually increase the interest rate. In response, the council can repay the loan immediately without any extra cost.
In practice, a bank like Barclays doesn’t agonise over whether to increase the rate of a £10 million loan to Camden council. It lets the market determine its economic interest, and when the loan is made, the bank replicates the option internally and sells in the market, locking in a margin. Some of that margin helps reduce the headline rate payable on the loan, but most of it is booked by the bank as upfront profit. If the option subsequently goes in-the-money it gets exercised against the bank, which in turn exercises its LOBO option on the council. Whether it realises it or not, the council is on the other side of a derivative called a swaption (see Note 1).
The best way to look at LOBOs is not as a long-term loan that can get cancelled if rates rise but rather as a short term loan that suddenly becomes extremely long-term when rates fall (see Note 2). When LOBOs first took off in the late 1990s and 2000s, sterling interest rates were already falling, and indeed there was a lot of demand for swaptions at the time. The buyers were UK life assurers who had sold guaranteed minimum-rate annuity policies to savers, and back then I wrote about some of their big hedging transactions with City investment banks.
This was also a time when the UK government considered selling sterling options itself, as part of a 2004 report by economist David Miles looking at the long-dated mortgage market. Nothing came of that idea, but by 2005 the Bank of England noted that swaption prices were falling because of new supply from local authority LOBO contracts.
What probably no-one expected was that by the end of the decade rates could fall much further, with gilt yields breaching 3 per cent levels from 2012, and 30-year swap rates falling to 2.5 per cent last week (see chart). For the life assurers, buying the hedges turned out to be a smart move as the swaptions kicked in. But what about the councils? What might have been 5-10 year loans if rates had ‘mean-reverted’ back to the levels of the 1990s now stretch out to 50 years, far longer than the cycle of capital spending at most local authorities. Through a daisy chain of counterparties, council taxpayers are paying out the annuities sold decades ago by life assurers.
Yet unlike the interest rate ‘hedges’ sold by banks to UK small and medium-sized enterprises, LOBOs haven’t emerged as a mis-selling scandal with borrowers tipped into bankruptcy. The derivative is hidden from view, and only if the council decides to repay a loan early, outside the narrow window of the LOBO agreement, does the mark-to-market impact make itself felt. And so, unsurprisingly, LOBOs sit there like Snow White after eating the QE-poisoned apple, slumbering on municipal balance sheets until our grandchildren repay them.
However, prompted by scandals in the financial sector, a few princes are trying to wake Snow White up. Joel Benjamin and Ranjan Kumaran, campaigners at finance reform group Move Your Money, have unleashed a flood of freedom of information requests to English councils. With a few exceptions, most have been forthcoming about releasing LOBO documents. For the first time (outside of an investment bank that is) it has been possible to value the contracts on a mark-to-market basis, to discover the kind of profits banks made on them and what their repayment cost might be today.
With the help of derivatives consultant Gary Kendall, I recently valued a portfolio of 25 LOBO contracts on the balance sheet of Kent county council, the third-biggest LOBO borrower in the UK. Allowing for a credit spread of 25 basis points, we estimate that the Barclays and the other lenders booked a profit of £23 million on the embedded derivatives. (see Note 3)
How does Kent value the loans? In its latest annual report, Kent says its LOBOs are worth £200.7 million. When pressed, Kent adds that this figure represents just part of the loans with ‘uncertain duration’. That’s conveniently smaller than the £441.8 million face value of the loans. But compare that again to the mark-to-market repayment cost of the loans, which on 15 October went above £716 million, according to our estimates (see chart). Putting it another way, the difference in value between what Kent says the LOBOs are worth and what the banks would legally require today is more than the council’s annual families and social care budget.
In a response to my questions, Kent insists “We do not believe that this has any bearing on the financial position of the council”. Indeed. Since no-one is actually forcing Kent to take that mark-to-market hit, why should we care? One might say that having to pay £1 billion in interest on Â£441 million of borrowing isn’t a particularly good deal for the council, and the mark-to-market penalties remove any flexibility to seek better financing terms elsewhere. In its response, Kent says it was advised to enter the loans by its former treasury adviser Butlers, and that LOBOs “represented the most cost-effective way of funding large amounts of capital expenditure at that time”. Beyond saying that the LOBOS are being “closely monitored”, Kent is staying tight-lipped.
But the real pressure to get out of LOBOs may come from the banks that originally issued them. Remember how these banks sold on the embedded derivative at the time of issue? Back in 2005 or so, that might have seemed clever but today it puts the banks in a worse position than the councils. Consider how the sharp falls in government bond yields and swap rates since 2010 revealed risk management deficiencies in some banks. Belgium’s Dexia, one of the biggest LOBO lenders in the UK, was forced to seek a government bailout after being hit by â‚¬15 billion of derivative margin calls in August 2011. The other big LOBO lender Depfa is another basket case, bailed out by the German government via its owner Hypo Real Estate and now being wound up.
The risk and accounting of these failed banks remains murky, but both will be included in the ECB’s forthcoming stress tests and asset quality review. That adds to pressure to sell volatile assets such as LOBOs to new investors, along with UK lenders such as Barclays that already mark their LOBOs to market. But who would buy such assets unless they were immunised against volatility? Doing so would require the bank to buy the swaption hedge back at its current market prices, locking in a massive loss.
The upshot, say industry practitioners, is that bankers are quietly sounding the councils out about commuting or restructuring LOBOs, negotiating a derivative unwind cost in an attempt to clean up their balance sheets. Kent won’t say whether any bankers have traipsed down the M20 to Maidstone yet. But perhaps a few years from now, UK municipal finance will be genuinely boring again.
Note 1: Strictly speaking, the options embedded in LOBO contracts are Bermudan swaptions, so-called because they can be exercised at annual or semi-annual intervals until the loan matures.
Note 2: Put-call parity says that a swap cancellable when rates rise is equivalent to a swap extendible when rates fall.
Note 3: What we did is work out the market value of each Kent LOBO loan assuming that its coupon payments were swapped into a floating rate by a bank trading desk. The key assumption here is whether the floating rate is Libor – the rate banks lend to each other – or whether it includes an additional credit spread, which reduces the potential upfront profit on the trade.
The Kent FOI documents indicate that the council was able to borrow floating-rate money at 25 basis points above Libor during the period the loans were taken out. At the recommendation of a former banker familiar with the deals, we added a 25bps spread when calculating the upfront profit. However, for the break costs, Barclays – Kent’s biggest LOBO lender – is quite explicit where it states in its loan documents that break costs are to be computed at Libor flat, so for that part of our analysis we omitted the spread.