FNM, The derivative spiral,
rate derivatives and counter party risk, an explanation,
capitalstool.com
This is a rather long example which explains how interest rate derivatives work, and how FNM has triggered a derivative spiral. I hope it helps and I am sorry that I couldn't find a way to decrease the length.
Roger
Step back a few months in time to when Bill Gross (www.pimco.com) came forward and identified GE as no more than a highly leveraged hedge fund that had a few sideline industrial businesses. Gross correctly identified that GE had greater than $95B in commercial paper (CP) outstanding and did not have bank credit lines in place to backstop this extremely short term credit risk. A few months later Gross pen another missive where he cleverly deduced that GE and many other US corporation, which at this time had been frozen out of the CP market, were replacing there CP borrowing with 5-10 year debt but were not reflecting any pressure in there earnings. Let me state that I think Gross is a very smart and very good investment manager. I don’t however think that he just recently came to these conclusions. As with all other aspects of the mania, when times were good people looked the other way, now they are concerned but unable to do much about it. Why? Because when you are the major player in the bond market, like Gross, you can’t simply sell as this will set off the stampede. If we use GE as credit derivative example we can explain the current derivative spiral gripping the treasury market and threatening FNM.
So why did Gross serve notice to GE? What does he hope to achieve.
First, lets look at what GE was doing in the CP market. This market serves to fund the day-to-day liabilities of a corporation. It allows corporations to make payrolls every two weeks despite the possibility that incoming cash flows are more lumpy and maybe on a 45day or longer cycle. The CP market is very liquid because the liabilities only exist for days at a time, with 3 months being the longest available security. The flip side of the CP market, the asset side, is your own Money Market (MM) holdings. Since you may demand your money at a any time the manager of your (MM) funds cannot be locked into long-term notes or bonds. This imparts two key characteristics on the CP market. First the rates are extremely low; because the length of the loan is short the riskiness of the loan is reduced. It is highly unlikely that a company will go bankrupt in thirty days or less. With each CP issuance, an insurance policy and due diligence package assures buyers that their money is safe. Secondly, the market is liquid because those who are unable to pay their debts are thrown out of the market within thirty days. (Note: I don’t wish to contradict Mark’s recent comments about his MM account holdings. I am presenting a more idealized version of the way things work. Mark is early in describing how even this market has been corrupted in the quest for yield) Look at your most recent MM statement, it is probably yielding 1.2 – 1.5% annually. Add about 15% for processing and redemption fees and you get an effective financing rate of 1.4 – 1.8% per year. At that rate GE can finance turbines or whatever else at 5% per year and make a killing. And in fact that is how this “industrial” company is heretofore recession proof.
So what is the catch? As we have been hearing from FNM recently, you may end up with a duration mismatch. That is, in the case of GE, a company places an order for a $50m jet engine. They of course do not have the cash to pay for it so they agree to enter into a structured finance arrangement whereby GE lends the company $50m and demands two payments a year consisting of principle and interest. At the same time GE must pay for the materials and labor required to build the engine. Therefore GE goes out to the CP market and borrows an amount equal to the cost of the engine + the amount needed to finance that cost until the engine is delivered and the company begins making payments.
This highly simplified transaction demonstrates the duration mismatch. In this case GE borrows in thirty-day increments to fund the construction and financing of a jet engine that has a service life of twenty years and a loan to value of ratio of 1.0. Why does GE do it? To earn the spread between CP 1.5% and 10year bonds 5% (at the time) which is 3.5%. On a $50 million dollar jet engine financed for ten years that amounts to roughly $22m. I am willing to bet that the margins on a jet engine are far less than the profits from financing it. GE still must make payments every thirty days even though the company only has to pay every six months.
So, where is the downside? First lets consider your MM holdings, these aren’t invested assets they are the monies that you rely on for your daily living expenses: rent, mortgage, food & tuition, etc. Remember the old adage: if you need the money within five years invested in bonds not stocks. Where do highly specific assets such as made to order jet engines fit in the equity bond risk spectrum? Now lets consider GE. Once the engine is delivered it loses a substantial portion of its value to depreciation, it also is not entirely fungible due to its custom nature. Should the airline go out of business GE’s risk goes beyond the jet engine loan, because it will also lose any expected income from spare parts or other future orders. Finally the CP market, which was providing short-term low cost funds against liquid receivables, is now left holding a risky unsecured loan on an illiquid asset. So it is clear to see why Gross, a major player in the CP market, was outraged and was determined to stop this practice.
All things being equal, once GE was cut off from the CP market we should have expected a massive earnings miss and an implosion in the stock price. Of course this never happened. So the question is how can GE possibly continue to make profits from interest rate spreads when it must borrow at a rate greater than 10yr treasury rates. For that matter we can now extend this example to most large non-financial corporations that were pushed out of the CP market at about the same time. Enter JPM, BAC, GS, C, BSC, AIG, LEH, etc and the derivative and corporate bond bankers.
To make a rather complicated story short any of these derivative players would be happy to sell an interest rate derivative to the CFO of GE or any other major corporation. These derivatives fall under three major categories:
1)Rate swaps – Here GE issues ten year bonds and immediately enters into a swap. GE agrees to pay the floating rate, while JPM agrees to pay the fixed rate. Think of it like a fancy adjustable rate mortgage. We will go through this example in more detail below.
2)Rate caps – Think of this like an option on interest rates. GE could buy a put on interest rates; if rates decline from the GE bond issue price they would capture the benefit without calling the bonds and reissuing. Or they could buy a call option, which would allow them to rollover bonds at the same rate even if interest rates were to rise.
3)Rate spreads – Very fancy rate caps that are dependent not only on the rate of interest but also on the differential between various treasury bonds and notes.
Keep in mind that derivatives can be very complicated and impossible for those not involved with details of any particular deal to understand. We are going to concentrate on rate swaps, as they are the most relevant to our discussion of the derivative spiral in treasuries.
Going back to our GE example we are at the point in time last spring when GE had to replace some amount of CP with corporate bonds. Without all of the details we will assume the CP rate to be 1.75%, the GE bond rate to be 5%, and the cost of the swap to be 120 basis points. First, why does GE need to replace CP borrowing with a bond and a rate swap? The above example suggests that GE was borrowing in the CP market to fund current projects. Those projects and the expected financing profits were incorporated in earnings projections that were presented to Wall Street. Any changes to the underlying financing would require a substantial change in earnings forecasts. Second and more important, if interest expense were to rise and rise quickly this would further depress earnings and change GE’s cash flow substantially. So, GE is willing to pay maintain low floating rate payments.
Every derivative contract has a counter party; in this case we will choose GS. Why would GS enter into this derivative contract, which would force them to receive floating rate payments, which at the time were 1.75%, every six months and make fixed rate payments, which are 5%. This is a very complicated question which I can only summarize possible answers:
1)GS may have several short positions in similar derivatives and looks to hedge away exposure by entering long on this contract
2)GS feels they can effectively hedge using the cost of the swap (120 basis points) and produce a profit
3)GS internally has made a macro call on increasing interest rates and is speculating that short rates will rise faster than long rates
We cannot know for certain what GS’ reasoning is. Let’s look at the mechanics. GE issues $10B in 10yr notes yielding 5%. GE simultaneously purchases a $10B notional rate swap from GS paying 120 basis points. The swap is good for ten years. On the day the transaction is executed GE receives $10B (excluding costs and fees) from bondholders and pays GS $12m for the swap. The counter parties do not have any further obligations until six months later when the first interest payments are due. GS is obligated to pay GE $25m which is 5% (annualized) of the $10B loan. GE is obligated to pay GS the prevailing short-term interest rate that we cannot know at this point. If rates remain the same then it is 1.75%. So then GS is entitled to receive $17.5m or in actuality GS must pay to GE $7.5m. If short-term rates decrease then GS must pay GE more, if short-term rates increase GS pays GE less. If short-term rates skyrocket above 5% then GE must pay GS.
From this example we can see that if rates remain the same in the first six months GS will have a profit of $4.5M. However they are committed for ten years and cannot withstand the potential losses if short-term rates remain the same or go lower. For this reason they must hedge there exposure buy selling the swap. Likely buyers would include FNM or hedge funds who tend to have cash flows that are opposite GE. Or they could hedge using rate caps and rate swaps. Alternatively they could hedge by netting against other derivative contracts in there portfolios. Finally GS could attempt to buy yielding instruments to make the 5% interest payments although it is easy to see that with the $12m payment received it would be impossible to generate $25m in interest payments that are due every six months.
How in the world does this relate to FNM??!!
FNM has the exact opposite problem (we will exempt credit quality and concern ourselves with prepayment risk). FNM has a portfolio of mortgages which she holds directly or as pass throughs. Unlike GE however her counter parties the mortgage payers of the United States have an unlimited option embedded in their loans. At anytime you can prepay your mortgage or sell your house thereby dissolving your mortgage. FNM bondholders however only allow her to call a certain percentage of her bonds and only allow those bonds to be called infrequently. Hence fannies liabilities are relatively fixed while her assets may fluctuate substantially.
Fannie routinely issues 10,15&20yr bonds that are priced about 20-50 basis points above comparable treasuries. So lets look at her bonds issued in the last three years, before Sir Printsalot decided to slash rates. In the interest of time lets agree that they yield more than 5.5% Now we focus on the current melt-up in 10yr treasuries as current mortgages are indexed to this rate. At a level of 4% on the 10yr we can expect mortgages to price below 7% on thirty years. 7% is quite a bit lower than mortgages were pricing three years ago. It make sense the mortgage payers would execute their option and refinance their mortgage at a level below 7%. Given the recent melt-up in treasuries, which I describe as a derivative spiral, the new mortgages will be at a level substantially below 7% with many below even 6%. Fannie’s liabilities remain the same and hence we get a duration gap which is a mathematical way of analyzing the mismatch just described. Fannie has three options:
1)Stop buying mortgages, fold the tent and go away. Many a stool may dream but unlikely.
2)Buy as much agency debt as possible to reduce the liability portion of the duration gap. This does not solve the problem it just limits it in the short term.
3)Attempt to hedge using treasuries. As shown in the GS example it is impossible for FNM to recreate 7% interest payments by purchasing government securities yielding 3.75% or below. Further the agency market now exceeds the treasury debt therefore hedging is not possible. Thus the derivative spiral begins.
The derivative spiral:
1)For whatever reason, in this case economic weakness, treasury yields decrease.
2)Treasury yields enter a zone that is likely to trigger mortgage refinancing.
3)Mortgage bankers fearing prepayments start buying treasuries hoping to hedge the losses in their portfolios from prepayments with capital gains from treasuries as rates decrease.
4)GS and other derivative counter parties that have entered into rate swaps purchase treasuries in an attempt to offset short-term interest rate payment liabilities.
5)FNM and other mortgage intermediaries having relatively fixed liabilities temporarily deploy prepayments in the treasury market in attempt to replicate eliminated interest payments.
6)Speculators seeking capital gains enter the treasury markets and interest rate futures market
7)The combination of 3-6 depress interest rates to the point that second order derivatives (options) cause rate caps to become in the money requiring counter parties to hedge using treasuries securities.
8)Steps 3-7 become self-reinforcing regardless of demand for mortgage finance.
Keep in mind that derivative trading operations settle P/Ls at the end of each trading day. Therefore losses beget losses and gains beget gains in the spiral until an outside force exerts itself and either stops all trading or relieves the distressed counter parties. |
