Past article: Derivatives in court - Collared IRS

This article appeared in Corporate page of The Edge Malaysia, Issue 804, May 3 - 9, 2010

Derivatives in court: “Collared” Interest Rate Swaps

By Jasvin Josen

On March 17th, four big banks, 11 bankers and two former city officials were charged with fraud in connection with the sale of interest-rate derivatives to the city of Milan. The trial is due to start on the 6th of May.

It all started with the Italian municipality hiring four banks to find investors for its €1.69 billion in bonds. Public records show that all four agreed to charge the lowest fee that had been proposed, 0.01 percent. Similar deals in Italy in the same period have carried fees of 0.3 and 0.45 percent of the face value of the bonds. “An underwriting fee of 0.01 percent looks unreasonably low indeed, particularly so in relation to the underwriting risk of a fixed-rate, 30-year bond,” said Milan-based Piero Burragato, a council member.

The prosecutor will contend on 6 May that there was in fact, a significant fee built into an interest rate derivative packaged with the bond. This article will reveal how it was done.

The deal

As part of the deal, the same four banks were hired by the city to also advise it on how to use the new bonds to restructure its existing debt to cut costs.

The banks proposed two phased approach: Step 1, the city could save money by buying interest-rate swaps. Step 2, the institutions best prepared to sell them those swaps were none other than the banks themselves. Without seeking competitive bids, the city agreed in June 2005, to let the four banks sell them swap contracts.

The Interest rate Swap...and the collar

The 30-year bond carried an annual interest of 4.019 percent. With the interest rate swap (IRS), the city swapped the fixed interest rate for a floating rate set at 12-month EURIBOR. With the low sentiment in interest rates then, it made sense to swap fixed interest rates with variable rates. Furthermore the city would minimise interest rate risk with the swap. (IRS and interest rate risk is explained in my previous article on “Financial Wizardry in swaps – the Greek case” March 8, 2010)

However there was slightly more to this particular IRS. . Part of the contract involved a “collar”. This meant that the variable rates that Milan was supposed to pay had an upper and lower limit, referred to as a cap and floor. If the EURIBOR rate rose above a certain level, the banks would pay Milan the difference between the actual rate and the capped rate. As for the floor, if the EURIBOR went below a certain level, Milan would have to compensate the banks instead. The combination of the cap and the floor protected Milan from rising rates but this also meant it would have to pay out if the EURIBOR fell. Chart 1 illustrates the whole deal. Do note that the cap and the floor rates in the chart are only hypothetical examples.

Chart 1

Caps and Floors are really Interest Rate Options

A cap is actually a portfolio of interest rate options, to be specific, call options. Milan actually bought a portfolio of 12-month EURIBOR call options, expiring every year, for the thirty year period. The strike rate of the call was the cap rate. If the 12-month EURIBOR went above the strike rate, Milan would benefit by getting a payoff from the banks. For example, say the strike rate was 4.2% and the 12-month EURIBOR at that time was 2.5%. Now assume that after four years, the EURIBOR rose to 5.0%. The banks would then compensate Milan for 0.8% (5% - 4.2%) of the face value of the bond.

Whereas a floor is a portfolio of interest rate put options. Milan sold a portfolio of 12-month EURIBOR put options to the bankers, expiring every 12 months. Now say the strike rate of the puts was set at 2%. If the EURIBOR for any of the periods fell below 2%, Milan would have to make a payoff to the banks.

Valuation of caps and floors

Caps and floors being interest rate options are most often valued using the Black Scholes model. Without going into the workings of the model, we review below, the important inputs of the formulae:

 The underlying rate (the 12-month EURIBOR rate) and the strike rate. This will ascertain if the option is in-the-money, at-the-money, or out-of-money.

 Maturity of the option (12 months)

 Volatility rate (this is usually implied volatilities from similar market caps and floors)

The caps and floors price will consist of thirty caplets and floorlets respectively, expiring year after year. The underlying 12-month EURIBOR rate for the caplets that start in the future (for example starting Year 2 expiring Year 3) are obtained from the forward EURIBOR curve.

To buy a cap and sell a floor is a common strategy. In the usual case, the strategy involves a zero cost collars. This product cost nothing to the client as the price to buy the call (or premium paid) would completely be offset by the price to sell a floor (premium received). The level of the cap and floor would be chosen such that the premiums set off each other.

What went wrong?

  • EURIBOR hit the floor

Apparently in the beginning, Milan did generate some savings as the EURIBOR went above the cap rate. However in 2009 the EURIBOR rates fell, triggering payments to the banks.

  • Huge upfront payment for the collared IRS being questioned

What the city of Milan is really unhappy about is that it paid a huge price as upfront payment for the collared IRS. The prosecutor, Alfredo Robledo, claims that Milan had to pay an upfront payment of more than €100m. In investment banking terms, this was day-one profit to the banks.

If I were to perform an analysis in retrospect, the upfront payment could have come from two sources, the IRS or the collar. I do not expect the IRS to be very off-market, as the fixed rate of the bond and the variable EURIBOR rates are derived from the market itself.

It is more likely that the bulk of the price Milan paid came from the collar. It is unlikely that the collar was a zero cost collar, meaning that the upfront premium receivable for the cap purchased did not offset the premium payable for the floor sold. The cap level may be set at a high EURIBOR rate, making the call options more out-of-money and Milan receiving a small premium for the cap. Alternatively, the floor may be placed at a high EURIBOR rate, resulting in the put options to be more in-the-money and thus commanding a high premium, which Milan would have to pay upfront to the banks. We can see now that the cap and floor levels could be easily tailored to generate any upfront payment required.

  • Mark to market losses

Today, Milan faces mark-to-market losses of €231 million on its swaps, according to council member Davide Corritore [source: Bloomberg]. This value is potentially made up of the IRS and the collar. The IRS is likely to have a positive value resulting from the decreasing EURIBOR rates payable to the banks. However it is very possible that this positive value is being wiped out by the much higher negative market value from the collar. The market value of the collar could be interpreted as the cost to exit the deal. If Milan would want to exit the collar deal now, it would have to sell the cap and buy the floor as a contra transaction. With the extremely low EURIBOR rates these days, the cap or call options would be worthless. But I am guessing that the floor or put options is deeply in-the-money. The exit cost to Milan hence would be very high, translating into mark-to-market losses in its books.


We like to believe that financial innovation producing complex derivatives were primarily intended to channel risk to parties best suited to hold it. It is turning out that the risk seems to be ending up with those least familiar with it.

Does that mean that Milan has a right to sue? Probably not. Despite the complicacy of the derivative, the responsibility still lies with the buyer to be wary of the seller. The caveat emptor principle is valid everywhere and anywhere.

1 comment:

  1. U.S. drillers added oil rigs for a sixth consecutive week, extending a nine-month recovery, energy services firm Baker Hughes Inc said. capitalstars