Swap mechanics and pricing

1
Swap mechanics and pricing

• National Association of State Treasurers
• State Debt Management Network
• December 3, 2007
• Swap Financial Group
• 76 South Orange Avenue, Suite 6
• South Orange, New Jersey 07079
• 973-378-5500

2
Swap structures

• Many structures, some confusing ones
• 90 of all muni swaps synthetic fixed rate
• Swap is replacement for conventional fixed bonds
• Can start immediately (spot start) or in the
  future (forward start)
  
• Other 10
• Synthetic floating rate
• Basis swaps
• CMS swaps
• Swaptions

3
Swap structure (synthetic fixed)
Swap Dealer
Issuer
Issuer pays Fixed Rate minus the difference
between the two Floating Rates
Bond Holder
4
Tax-exempt bonds vs. swaps
Note Swap rate includes 26 bps cost of annual
floating bond costs. Prices are illustrative.
5
Math Swaps vs. Bonds

• Bonds
• Fixed coupon
• Amortized cost of issuance

• Swap
• Floating bond rate
• Annual costs of floaters (auction fees or
  remarketing and liquidity)
  
• Fixed swap rate
• Floating swap rate

• All-in cost

• All-in cost

6
Plug in some numbers

• Bonds
• 4.55 (fixed coupon)
• 0.05 (amortized cost of issuance)

• Swap
• VR (floating bond rate)
• 0.26 (annual floating bond costs)
• 3.35 (fixed swap rate)
• VR (floating swap rate)

• 4.60 (all-in cost)

• 3.61 (all-in cost)

7
Why does it work?

• Counter-intuitive Why should three steps (issue
  floating, receive floating, pay fixed) be more
  efficient than one (issue fixed)
  
• Swaps allow you to unbundle and take advantage
  of relative efficiencies of different markets,
  and to decide to take certain risks (i.e. greater
  or lesser amount of basis risk)
• Market sensitive It doesnt always work

8
Getting a Fair Price
9
What is mid-market?

• Markets are quoted as bid and offered
• Mid-market (mid) is the hypothetical halfway
  point between the bid side (issuer receives
  fixed) and the offered side (issuer receives
  floating)
• Mid is the starting point for all pricing
  discussions

10
(No Transcript)
11
Dealers spread to mid

• Three components
• Hedge cost
• Credit reserve
• Dealer profit
• Key question Whats fair?

12
What makes hedge cost vary?

• Different products, different hedging costs
• Less liquidity wide bid-offered spread
• Occasional issue Time of day

13
(No Transcript)
14
What makes credit reserve vary?

• Ratings
• Type of credit
• G.O. , water and sewer
• Transit and toll facilities, state HFAs
• Public power generators, solid waste
• Private higher ed, health care
• Nursing homes, convention centers
• Different dealers different standards

15
What makes profit vary?

• Experience (of issuer and issuers advisors) in
  negotiations/competition
  
• Desirability of client
• Deal size
• Deal difficulty and time

16
Spread components
17
Nuances When mid is hard to find

• Mid is often not as certain as is represented
• Reasons
• Product complexity
• Model differences
• Varying skills at hedge execution
• Unhedgeable elements
• Examples to follow

18
Example 1 Simple LIBOR swap

• Tower 111
• Manhattan apartment project
• 100 million, 10-yr amortizing LIBOR swap
• Strong credit guarantee
• Bids (SFG mid was 5.591)
• PNC 5.600
• Bank of New York 5.599
• Bank of America 5.597
• Wachovia 5.595

High-low diff 0.5 bps
19
Example 2 Large BMA swap

• East Bay Municipal Utility District
• Top-tier water and sewer utility
• 392 million, 19-yr amortizing BMA swap
• Bids (SFG mid was 3.393)
• Lehman Brothers 3.454
• Bear Stearns 3.414
• Merrill Lynch 3.412
• Citibank 3.412
• Siebert Brandford Shank 3.407

High-low diff 4.7 bps
20
Example 3 Embedded options

• California Housing Finance Agency
• Nations leading HFA, largest swap user
• 82.5 million taxable, 10-yr LIBOR swap with
  embedded options
  
• Bids (SFG mid was 5.62)
• Merrill Lynch 5.683
• JPMorgan 5.670
• Citibank 5.615
• Goldman Sachs 5.604
• Bank of America 5.594
• UBS 5.590

High-low diff 9.3 bps
21
Example 4 Ultra-long cap

• NewYork-Presbyterian
• NYCs largest non-profit hospital system
• Board wanted to use an interest rate cap for
  protection for its floating rate debt
  
• Bought longest caps in history
• 30 years (record had been 17, most under 10)
• Capped BMA at 6

22
Interest rate cap
Cap Dealer
Issuer
23
How a cap works
6 cap on BMA Index
24
Bid results March 2005
High-low diff 148 bps
25
Bid results November 2006
High-low diff 88 bps
26
Pricing take-aways

• Pricing transparency can be obtained for most
  straightforward swap structures
  
• For complex structures, swap pricing is more
  translucent than transparent
  
• Knowing if swap pricing is fair often requires
  a judgment call
  

27
Competitive vs. Negotiated
28
Role of swap advisor

• Usually client knows which way he wants to go
• Often, we are brought in after the dealer has
  been chosen sometimes, much after
  
• Client often uses us to confirm his judgment
• Our experience Most governments compete, most
  non-profits negotiate

29
Why negotiate?

• Its my first time, and I want to do it with
  someone I like
  
• A key financial relationship is overwhelmingly
  most common reason clients negotiate swaps
  

30
Why negotiate?

• I want to give the business to someone who
  brought in a great idea
  
• Much less common, as there are few truly original
  ideas
  
• The best ideas are not original product ideas
• Instead, how to apply a product to a clients
  individual circumstances
  

31
Why negotiate?

• My deal could move the market if I bid it
  competitively
  
• Also less common, but can occur if
• Deal is very large
• Deal is in a relatively illiquid part of the
  market
  

32
Case Study University of Texas

• May 2006
• 540,570,000 (amortizing)
• UT pays BMA, receives 67 of 5-Year LIBOR CMS
• BMA market is relatively small deals over 200
  million can move it
  
• Broadly bidding UTs swap moved market 5 to 10
  basis points against it

33
Why negotiate?

• The dealer made me an extremely good price
  proposal
  
• Also less common, but sometimes occurs in the
  context of the negotiated-competitive
  decision-making process
  

34
Case Study Houston CMS

• August 2006
• 249,075,000 (amortizing)
• Houston pays 70 of 1 mo. LIBOR, receives X of
  10-Year LIBOR CMS
  
• No market size issue (LIBOR market very deep)
• Houston wanted to come to market quickly, but
  needed formal A.G. approval
  
• RBC offered to negotiate swap at 2 bps from
  mid-market
  

35
Case Study MBTA

• June 2006
• 280,000,000 (amortizing)
• MBTA unwind of prior fixed-payor swap
  (BMA-based)
  
• Agency wanted to go competitive
• Possibility of adverse market move
• Original dealer (Bear Stearns) offered compelling
  pricing
  

36
Why compete?

• My only goal is best price
• If deal wont move market, competition usually
  provides best pricing
  
• Issue It may take more time to qualify bidders
  (get agreement on key documentation issues)
  

37
Case Study California DWR

• September 2005
• 2,594,000,000 (amortizing)
• DWR pays fixed, receives 66 of 1 mo. LIBOR
• Agency wanted to compete broadly, diversify among
  5 or more dealers
  
• Went out to 17 dealers
• Winners were all non-relationship banks
• Agencys key relationships were ticked off

38
Why compete?

• I need to demonstrate I achieved best price
• Very common reason, esp. in governmental market
• Alternative Fairness Opinion

39
Why compete?

• My product is so unusual that fair price is very
  hard to establish without competition
  
• Real rarity, but sometimes happens
• Best example NY Presbyterian 30-year cap
• New product
• Difficult to hedge
• Priced ranged widely

40
Swap Risks and GASB Reporting
41
Swaps and the bond holder
Swap Dealer
Issuer
The bond holder is not party to the swap, but
must be informed about the swap.
Bond Holder
42
GASB requirements

• Technical Bulletin 2003-1
• General footnote disclosure
• 2007 Exposure Draft (2010 financial statements)
• Measurement of effectiveness
• Reflecting changes in swap value on financial
  statements
  
• If swap is ineffective, then change in value
  must be reflected as change in investment income
  
• If effective, only recognized on balance sheet as
  deferred inflow or outflow

43
Risk 1 Termination Risk

• Termination Risk is the risk that the swap might
  be terminated prior to its scheduled maturity
  
• On day one and at maturity, a swap usually has
  zero value
  
• The story is different in between

44
Measuring swap value

• Swaps can be either assets or liabilities,
  depending on market
  
• Swap value Cost or benefit of terminating under
  specific market conditions
  
• The termination payment is based on
• Interest rates at time of termination
• Remaining years to scheduled maturity
• Notional principal amount

45
How termination works

• Compare original contract swap rate with current
  market rate for a swap ending on the same date
  
• Multiply rate difference times dollar size and
  years remaining, present valued
  
• Example Original rate (5.50) current rate
  (4.50) difference (1.00) times size (10 mm
  100,000) times years remaining (10 years 1
  mm), present valued (at 4.50 770,000)

46
Measuring Termination Exposure

• Assume Issuer has entered into a 100 million
  30-year swap paying 4.50 and receiving the BMA
  Municipal Swap Index. The table shows the
  Replacement Value of the swap at future points in
  time, assuming 200 and 100 basis point increases
  in rates, and no principal amortization.

47
Risk 2 Basis Risk

• Basis Risk is the risk that the floating rate the
  issuer receives on its swaps doesnt offset the
  floating rate the issuer pays on its bonds
  
• Basis Risk is the key issue in effectiveness
  testing

48
Review of swap structure
Swap Dealer
Issuer
Fixed Rate
Basis Risk comes from the difference between the
two Floating Rates
Bond Holder
49
SIFMA (BMA) basis risk

• SIFMA, the Securities Industry and Financial
  Markets Association Municipal Swap Index, is an
  arithmetic average of a wide range of tax-exempt
  floating rate bonds (non-AMT)
• Most floating muni bonds normally trade very
  close to SIFMA
  
• If the trading relationship changed, the Issuer
  would be exposed to risk
  
• Potential sources of change Issuer credit
  problem, scandal involving other same-state
  issuer
  
• GASB will allow automatic effectiveness for SIFMA

50
LIBOR basis risk

• Tax-exempt floaters normally trade at a
  percentage of the taxable LIBOR index (i.e. 67)
  
• Using LIBOR generates big savings
• What would happen if munis will lost preferential
  tax treatment?
  
• Bondholder bears risk with fixed-rate bonds
• Issuer bears risk with unhedged floating rate
  bonds and of LIBOR swaps

51
Tax risk scenario
Current Tax Structure
Radical Tax Change
Swap
Issuer
Net funding cost 5 minus 4 of LIBOR LIBOR t
oday is 4.75, so 4 of LIBOR 19 bps Bottom l
ine cost 4.81
Net funding cost 5 plus 17.5 of LIBOR LIBOR
goes up to 7.00, so 17.5 of LIBOR 123 bps
Bottom line cost 6.23
52
Tax risk events

• Small effect Reduction in federal rates
• Larger effect Exemption of all investment income
  corporate bond interest, dividends, capital
  gains from income tax
  
• Largest effect Taxation of munis under a Flat
  Tax, with no grandfather clause
  
• Key Questions How real is the risk? Does the
  benefit more than compensate for the risk?
  
• GASB will require quantitative demonstration for
  effectiveness

53
Risk 3 Liquidity rollover risk

• Definition Risk of inability to renew liquidity
  facility on floating rate bonds
  
• Most floaters depend on third-party liquidity
  facilities to backstop the put
  
• Liquidity could become expensive or difficult to
  obtain if
  
• General credit crunch
• Systemic banking problem
• District credit problem
• Disclosure is bond-related, not swap-related

54
Liquidity risk mitigation

• Various alternative approaches exist
• Convert bonds to index floaters
• Convert bonds to auction rate
• Convert bonds to fixed rate
• Call bonds

55
Risk 4 Counterparty Risk

• Definition Swap provider fails to perform or
  defaults
  
• The 1 risk -- These are long contracts
• Risk Measurement Replacement cost, not notional
  principal amount

56
Swap dealer universe

• Goldman Sachs
• GS Capital Markets (Aa3/AA-)
• GS Mitsui Marine Derivative Products (Aaa/AAA)
• Morgan Stanley
• MS Capital Services (Aa3/AA-)
• MS Derivative Products (Aaa/AAAt)
• Bear Stearns
• BS Capital Markets (A1/A)
• BS Financial Products (Aaa/AAA)
• BS Trading Risk Management (Aaa/AAAt)
• Lehman Brothers
• LB Special Financing (A1/A)
• LB Financial Products (Aaa/AAA)
• LB Derivative Products (Aaa/AAAt)
• Merrill Lynch
• ML Capital Services (A1/A)
• ML Derivative Products (Aaa/AAA)

• Citigroup
• Citibank N.A. (Aaa/AA)
• Citigroup Financial Products (Aa2/AA)
• Salomon Swapco (Aaa/AAAt)
• JPMorgan
• JPMorgan Chase Bank (Aaa/AA)
• UBS
• UBS AG (Aaa/AA)
• Others
• Bank of America (Aaa/AA)
• Royal Bank of Canada (Aaa/AA)
• Wachovia Bank (Aa1/AA)
• Deutsche Bank (Aa3/AA-)
• Dexia (Aa1/AA)
• Depfa (Aa3/A)

57
Counterparty Risk Mitigation

• Do business only with strong counterparties
• Natural double-As
• Synthetic triple-As
• Require downgrade collateralization provisions
• amount equal to the Replacement Value
• frequent mark-to-market of both collateral value
  and swap replacement value
  
• Get right to terminate early at Districts side
  of bid-ask spread at second downgrade threshold
  
• Diversify counterparty exposure
• GASB requires disclosure only of counterparty
  ratings

58
Swap Scandals
59
Dept of Justice Investigation

• Bid rigging and price fixing
• Swaps and GICs
• Three firms raided
• CDR Financial (Chambers Dunhill Rubin), Beverly
  Hills, CA
  
• IMAGE (Investment Management Advisory Group),
  Pottstown, PA
  
• Sound Capital, Eden Prairie, MN
• Almost every major dealer subpoenaed
• No charges yet, but its a criminal investigation

60
Biola University

• Medium-sized sectarian university in California
• IRS investigation found hidden mirror swap
• Disclosed swap was between Biola and BNP Paribas
• Hidden identical swap was between BNP and Bank of
  America
  
• Bank of America was Biolas banker and advisor
• IRS remedy Declared bonds taxable
• Biola is suing B of A and BNP

61
Narvik

• Eight tiny towns in Norway bought structured
  bonds, with embedded highly leveraged muni swap
  
• When market turned upside down, towns suffered
  50 loss
  
• Citigroup, Terra Securities are likely targets