US TREASURY FLOATING RATE NOTES

February 2012
STRICTLY PRIVATE AND CONFIDENTIAL

The demand backdrop

FRN issuance motivation and estimated benefits

Choice of reference indices and sample structures

19

U S

T R E A S U R Y

F L O A T I N G

R A T E

N O T E S

The demand backdrop for US Treasuries whether FRNs or fixed rate debt should benefit from a structural decline in the stock of high quality assets
Using market-based risk assessment provides

Total debt outstanding for G-10 countries with 5Y CDS spreads less than 100bp, as of end-2007 and end-2011; USD bn
20,000 18,000 16,000 14,000 12,000 10,000 8,000 6,000 4,000 2,000 End-2007 End-2011 United States United Kingdom Netherlands Japan Italy Germany France Canada Belgium

a useful view of the altered investment environment

The stock of bonds issued by sovereigns

with 5Y CDS spreads below 100bp has fallen sharply since 2007, excluding the US (whose CDS spread is below 50bp)
In other words, US sovereign debt is now

a higher fraction of the higher quality sovereign debt universe, likely resulting in a supportive demand backdrop for US Treasuries

THE DEMAND BACKDROP

Note: Sweden and Switzerland were excluded due to lack of data. For end-2011, Canada data is as of March 2011, and Japan data is as of September 2011. For European sovereigns, we assume that the amount of bills and non-domestic bonds outstanding is unchanged between 2007 and 2011.

Within the US fixed income markets, the share of Treasuries is growing while the share of other high-quality assets is falling

Net issuance for various fixed income products by year ; $bn

2300 1800 1300 800 300 -200 -700 -1200

2011E 2012F 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

ABS Net of Fed Purchases BABS Agency MBS Net of Fed & TSY Purchases AGY Net of Fed Purchases Financial Corps (ex ABS) Non-Financial Corps TSY Net of Fed Purchases/Sales Net Fixed Income Supply ex Fed & TSY purchases/sales

THE DEMAND BACKDROP

Will the Basel III LCR requirements trigger increased demand for US Treasuries in general and FRNs in particular?
Market analysts estimated a Liquidity Coverage Ratio of 57% for 30 large bank holding

companies, resulting in a gross shortfall of about $500bn - $1Tn depending on mitigation actions undertaken
While Treasury floaters would be attractive for LCR purposes, the net demand for FRNs

due to LCR provisions is likely to be modest

Treasury floaters are likely to yield less than liquid assets currently held by banks

(fixed rate Treasuries/Agencies/Agency MBS).

While Treasury floaters would be lower duration than current alternatives, many

banks are efficient duration hedgers and may be able to achieve higher returns net of hedging cost using the current mix of assets
With the Fed currently paying banks 25bp on excess reserves, banks are likely

ignore Treasury floaters unless the yield at least matches IOER.

Also, certain regulatory capital implications of Basel III provisions related to AOCI could,
THE DEMAND BACKDROP

on the margin, incentivize banks to buy FRNs over fixed rate debt
However, this is one factor among several driving asset selection, and we expect the

net preference for FRNs over fixed rate debt to be rather modest

Implementation of Dodd Franks central clearing provisions will also create some modest net new demand for high quality collateral such as US Treasuries
Growth of value of total reported and estimated collateral, 2000-2010; $bn

About 82% of gross credit exposure appears

4000 Reported 3000

2126

3957

collateralized alreadyfull implementation of DoddFrank could cause demand for collateral to rise by about $650bn
2934

Estimated
2649

3151

However, this does not directly translate into increased

2150

2000
1017 1209 854 1329 1335 922 924 1470

1984

demand for Treasuriesthe ISDA margin survey also indicates that Treasuries and Agencies make up only about 6% of collateral, which is predominantly composed of cash
Thus, the incremental demand for USTs due to

1000

719 707 437 491 200 250 289 138 145

0 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
Source: ISDA Margin Survey 2011

increased collateral requirements stemming from the implementation of Dodd-Franks central clearing provisions could turn out to be modest.
Caveat it is possible that the move to hold

Value of collateral received and delivered by respondents, $mn

Collateral received 877,552 106,697 38,606 22,943 10,948 21,005 13,196 100,699 1,084,949 Collateral Percent delivered 81% 715,444 10% 154,821 4% 48,409 2% 66,705 1% 13,414 2% 17,438 1% 8,854 9% 29,143 899,408 Percent 80% 17% 5% 7% 1% 2% 1% 3%
5

collateral in custodial accounts rather than on a bilateral basis could lower returns on cash and alter the fraction of USTs in the mix. An increase in this fraction could result in considerable demand for USTs
In addition, the fraction of cash relative to overall

THE DEMAND BACKDROP

Cash Government securities US EU UK Japan Other Others Total collateral

collateral could decline as short-term rates rise, leading to higher fractions of other assets such as USTs
Here again, any such incremental demand will

not discriminate between FRNs and fixed-rate securities

Money market investors have been increasing Treasury holdings recently thanks to a plunge in supply ex-Treasuries
Estimate of MMMF Treasury and agency securities holdings, $bn
Money fund AUM is modestly lower over the past four years, but money

market supply has shrunk by 18%

MMMFs are significant holders of Treasury and Agency securities MMMFs hold about $400bn in Treasury and Agency securities each

for combined $800bn, representing about 1/3 of AUM In addition, repo holdings backed by Treasury securities and Agency securities (including Agency MBS) are approximately $150bn and $250bn, respectively Preference for T-bills and Agency discount notes, but funds do own coupon securities including about $125bn in Agency FRNs with maturities 2 years or less There could be demand for Treasury floaters yielding more than Tbills, but MMMF preference is likely under 2 year final maturity
Other significant money market investors not bound by rule 2a-7 may have

Source: iMoneyNet, Crane Data, J.P. Morgan Note: Agency securities holdings includes discount notes and floating rate notes.

2a-7 Taxable MMF AUM vs. Money Market Supply, $bn, cumulative since 2007 year-end

THE DEMAND BACKDROP

Source: iMoneyNet, J.P. Morgan, Bloomberg

interest in Treasury floaters as a substitute for lower yielding bank deposits, money market fund shares, fixed-rate Treasuries and agencies, or other money market instruments. Securities Lending operations of custodial banks. Historically buyers of floating rate ABS, corporates and agencies with maturities out to 3 years. Market analysts have estimated total investments by these securities lenders exceed $1tn, of which floaters currently account for about $200bn State and local governments. Commonly invest operating and other funds in Treasuries, Agencies and MMMF shares. The mortgage GSEs. Actively invest excess cash by selling Fed funds, investing in Treasuries and repo and time deposits with a limited number of financial institutions. Treasury floaters with yields in excess of the Fed funds rate could be attractive. Corporate cash. The recent growth of US corporate cash balances has led to growth in bank deposits, MMMF and other liquid investment strategies. For many firms, Treasury floaters could be an easy to use alternative to these investments.
6

... but while this could be an initial tailwind, it is unlikely to be a long-term positive for demand
Over the past 2 years, declining supply of T-Bills and Agency obligations have forced

short-term liquidity investors into credit products such as CP/CDs, and any new supply would likely be well received
However, in the longer term, regulation of money-market funds could alter their appeal to

investors, subsequently altering their demand profile for any floating rate note issuance.
One offset to the extent that money market regulation could cause assets to flow

into short-term bond funds, those funds could in turn emerge as a demand source for FRNs

THE DEMAND BACKDROP

The demand backdrop

FRN issuance motivation and estimated benefits

Choice of reference indices and sample structures

19

U S

T R E A S U R Y

F L O A T I N G

R A T E

N O T E S

Revisiting the case to term out debt maturities

Treasury terms out debt in order to reduce debt rollovers as well as the uncertainty regarding future debt

service costs. Thus, extending the average maturity of outstanding Treasury debt is most beneficial when current term rates are low and/or the risk of large future changes in Treasury rates are at risk to the upside.
Term Treasury yields are a composite of three things Expectations of the future path of the Fed Funds rate Sovereign credit spreads, and Term premium
FRN ISSUANCE MOTIVATION AND ESTIMATED BENEFITS

By replacing T-Bill issuance with nominal fixed rate Treasury issuance, the Treasury locks in: The current path of the expected increase in Fed Funds (currently benign) Current credit costs as reflected in the current sovereign CDS credit spread (currently low) The current yield curve term premium (currently low)

FRNs can help reduce the debt roll-over burden, without paying the yield curve term premium, but at the expense of retaining exposure to rising interest rates and credit costs
By issuing FRNs with a term of (say) 2 years, Treasury can capture the low funding costs of T-bills,

while effectively terming out issuance and reducing roll-over requirements

As an example, monthly issuance of (say) $10bn of 2-year FRNs raises $240bn over two years,

and increases the rollover burden by $10bn/month once the auction cycle is fully phased in (i.e., after two years)
Issuing an equivalent $240bn of securities by increasing 3 and 6 month bill offerings could boost

the monthly roll-over requirement by $40-80bn.

However, by choosing to substitute T-Bill issuance with FRNs rather than with fixed rate debt of
FRN ISSUANCE MOTIVATION AND ESTIMATED BENEFITS

similar maturity, the Treasury:

Does not lock in a future path of short rates, and instead takes on that risk in exchange for

not paying the (usually positive) interest rate risk premium priced into the curve
It does lock in term funding, and thus takes advantage of its currently low sovereign CDS

spread.
However, it may retain exposure to a widening in its sovereign CDS spread if its floating debt

costs are indexed to some benchmark that is affected by it (e.g., T-bill yields)

10

The benefits of issuing Floating Rate Notes versus fixed-rate debt

Trailing hypothetical savings (relative to a 2Y fixed rate note) from issuing a 2Y FRN indexed to 6-month T-bill yields at a zero spread, versus the trailing change in the fed funds rate over the 2-year period
3.0 2.5 2.0
% Savings of Issuing FRN

Trailing hypothetical savings (relative to a 5Y fixed rate note) from issuing a 5Y FRN indexed to 6-month T-bill yields at a zero spread, versus the trailing change in the fed funds rate over the 5-year period
5.0 4.0
Change in Fed Funds Rate (2yr)

2yr - 6mo

2yr Change in Fed Funds

-6 -4
% Savings of Issuing FRN

5yr - 6mo 5yr Change in Fed Funds

-6 -5 -4 -2 -1 0 1 2 3 4
Change in Fed Funds Rate (5yr)

1.5 1.0 0.5 0.0 -0.5 -1.0 -1.5 90 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09

-2 0 2 4 6

3.0 2.0 1.0 0.0 -1.0 -2.0 90 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06

-3

FRN ISSUANCE MOTIVATION AND ESTIMATED BENEFITS

FRNs allow Treasury to term out its funding while lessening average funding costs in the long run Given typically positive term premium in the yield curve, the realization of short rates over a fixed term

(say, 2- or 5-years) will on average be lower than the ex-ante term rate
Thus, issuing FRNsassuming the issuance spread is not too high, and assuming that FRNs substitute

for fixed rate notescan produce cost savings on average. This has historically been the case, as shown in the charts above
Last, to the extent that UST FRNs draw in new incremental demand from investors who cannot hedge the

interest rate risk in fixed rate Treasuries, this should result in an aggregate benefit to Treasury
11

Savings depend less on the choice of the reference rate index, and more on the tenor of issuance that FRNs will replace - issuing FRNs instead of Bills will yield maturity extension benefits but not cost reduction
3-month T-bill yield and ex-post 3-month average of GC index (GCFRTSY Index) versus cost/savings of floater
0.25 0.20 0.15 0.10 0.05
FRN ISSUANCE MOTIVATION AND ESTIMATED BENEFITS

FRNs whether indexed to bill yields or an overnight

3M T-bill yield (%; left axis) 3M avg GC (%; left axis) Differential (bp, right axis)

20 15 10 5 0 -5

index such as fed funds or GCF would have historically produced largely similar cost savings
Charts alongside show that the rolling differential

between GC/fed funds with respect to bills is stable and relatively small. Thus, savings from issuing FRNs might be expected to be less dependent on the choice of index
Savings with respect to rolling bill issuance would

0.00 Oct 09

Feb 10

May 10 Aug 10

Dec 10

Mar 11

Jun 11

Sep 11

3-month T-bill yield and ex-post 3-month average of effective Fed funds (FEDL01 Index) versus cost/savings of floater
0.20 15

be expected to be negative; the price of terming out debt while still paying short term rates will be reflected in the spread over bill yields that Treasury will need to pay on an FRN, which is discussed later.

0.15

10

0.10

0.05

0.00 Oct 09

3M T-bill yield (%; left axis) 3M avg FF (%; left axis) Differential (bp, right axis) Feb 10 May 10 Aug 10 Dec 10 Mar 11 Jun 11 Sep 11

-5
12

Estimating the forward-looking benefits to Treasury from FRN issuance

Three factors determine the costs/savings of FRNs versus fixed-rate debt The pricing spread e.g., what fixed spread over (say) floating 3-month bill yields Treasury

pays to issue a par priced FRN

The level of interest rate risk premium at time of issuance The Feds monetary policy stancesavings are likely to be greater when the change in the

funds rate is negative, and especially when such change is more negative than the expectations priced into forwards
FRN ISSUANCE MOTIVATION AND ESTIMATED BENEFITS

The first of these pricing spreads will likely not be onerous enough to materially reduce the

savings from issuing FRNs

Even if FRNS are initially less liquid, market participants will likely arbitrage away any

significant differences from the spread implied by fixed rate note asset swap levels, thanks to a highly developed interest rate derivatives market
Experience in other fixed income product sectors that have fixed rate notes as well as

FRNs suggests that this is historically true

13

A closer look at yield curve term premium

Term premium itself may be thought of as being composed of two parts the cost of maturity extension, as well as the premium for the privilege of fixing funding costs, which we may think of as just the

interest rate risk premium

We may estimate the former by looking at the asset swap spread of term Treasury debt over

Bills e.g., if 2Y notes swap to 3M bill yields + 8bp to term, then 8bp represents the cost of maturity extension
FRN ISSUANCE MOTIVATION AND ESTIMATED BENEFITS

In near-zero rate regimes, we may estimate the interest rate risk premium via cap costs It is reasonable to assume that the risk to short-rates is one-sided, there is similarity

between the cost of an at-the-money cap on short rates (expressed in yield terms) and the portion of term premium attributable to the uncertainty in short rates
Estimation is subject to basis risks, since the cap market is based on Libor and not OIS

forwards; Libor cap costs likely overestimate interest rate risk premium currently, because of higher Libor rate volatility
FRNs will incur the costs associated with maturity extension, while saving on interest rate risk

premium (relative to issuing term debt)

14

Is this the best time in the cycle for FRN issuance from Treasurys perspective?
Estimated interest rate risk premium* by maturity; bp of yield Rolling 2-year savings from issuing a 2-year FRN linked to 6-month T-bills relative to issuing fixed-rate notes versus 2Y interest rate risk premium; bp of yield

110 100 90 80 70 60 50 40 30
FRN ISSUANCE MOTIVATION AND ESTIMATED BENEFITS

2Y 3Y 5Y

20 10 Jan 10 Jul 10 Jan 11 Jul 11 Jan 12

240 220 200 180 160 140 120 100 80 60 40 20 Jul 10

Trailing cost savings Savings attributable to term premium

Jan 11

Jul 11

Jan 12

* Estimated as the cost of an at-the-money cap on short rates, in yield terms. This is premised on the notion that in near-zero policy rate regimes, the one-sided nature of policy rate risk makes interest rate risk premium comparable to cap costs.

Current levels of interest rate risk premium are low, and the risk to the expected path of policy rates is likely asymmetrically biased higher

The Feds commitment to low rates until late-2014, as well as its new communications policy of projecting a path for the funds rate, have already lowered interest rate risk premium, and this is unlikely to rise for several years. Given de minimis monetary policy rates currently, the next move by the Fed is only likely to take the funds rate higher With interest rate risk premium currently near all time lows, savings are likely to be marginal

15

What about the cost where might Treasury FRNs price?

Estimated pricing spread on a hypothetical 2Y FRN linked to 3M Treasury bills*; bp

25 20 15 10 Estimated pricing spread; bp Average = 11.7

Regardless of the choice of floating rate index used to specify the coupons in any potential FRN issued by Treasury, it is useful to consider the par priced FRN spread-over-bills for purposes of analysis

I.e., if the basis swap market is used to transform the FRN into a floater linked to bill yields, what would the pricing spread be for a par priced FRN at time of issuance This represents the direct cost incurred by Treasury, for the sole purpose of terming out its debt It is reasonable to assume that this will not fall below zero; should it do so, Treasury has a strong incentive to issue FRNs in place of T-bills

5
FRN ISSUANCE MOTIVATION AND ESTIMATED BENEFITS

-5 Jul 10 Jan 11 Jul 11 Jan 12

* Assumes that FRNs will price at the same asset swap level as a maturity matched bullet Treasury.

The chart alongside shows the hypothetical pricing spread, if FRNs were to price at the same asset swap spread as a maturity matched fixed rate Treasury note. This is a reasonable estimate of where FRNs might price

16

A stylized illustration of the relationship between pricing spreads and the attractiveness to Treasury
A schematic illustration of the attractiveness of issuing FRNs from Treasurys perspective, for various pricing spreads (versus a T-bill floating index)

Bills + maturity extension premium + interest rate risk premium

Not attractive FRNs deliver maturity extension at a higher cost than term fixed-rate debt

FRN ISSUANCE MOTIVATION AND ESTIMATED BENEFITS

Likely pricing spread = bills + maturity extension premium

Attractive FRNs still deliver savings from term premium but give some of it back for the privilege of extension

Bills + 0

Very attractive FRNs deliver maturity extension as well as cost savings relative to term debt or rolling bills

17

The impact of FRN issuance on the weighted average maturity of Treasury debt will likely be relatively modest in the initial years

We consider 3 stylized issuance policies: Treasury issues FRNs at the expense of reduced issuance in (i) T-bills, (ii) matched maturity fixed rate issuance, or (iii) a reduction in fixed rate coupon Treasury issuance in proportion to current gross issuance Replacing matched maturity fixed rate debt does not alter WAM, but the other two policies will alter WAM

Projected WAM in different scenarios; months Date Base case: matchedmaturity nominal coupon sizes are reduced to issue floaters 63.4 65.4 68.3 70.0 Bill issuance is reduced to issue floaters 63.5 65.5 68.6 70.3 Nominal coupon sizes are reduced proportionally across the maturity stack 63.4 64.8 67.1 68.4

Issuing FRNs wholly at the expense of bills will have the greatest impact on WAM

Sep-12 Sep-13 Sep-14 May-15

FRN ISSUANCE MOTIVATION AND ESTIMATED BENEFITS

Assumes $50bn in annual FRN issuance beginning in Note: The base case assumes that coupon sizes are unchanged while net bill issuance adjusts May 2012 ($25bn in 2s, $15bn in 3s and $10bn in 5s), as the budget deficit increases/decreases. Also, FRN issuance is assumed to occur at the expense of maturity matched fixed rate coupon issuance . which is increased to $100bn annually by 2014 The increase in WAM is likely to be modest even if done wholly at the expense of T-bills

Historical and projected WAM of marketable Treasury debt based on scenarios #1 and #2 in table above
75 70 65 60 55 50 45 Dec 80 Base case/ matchedmaturity nominals
18

Historical and projected share of T-bills as %ge of marketable Treasury debt based on scenarios #1 and #2 in table above
35%

All from bills

30% 25% 20% 15% 10% 5% Dec 94 All from bills Jan 99 Feb 03 Mar 07

Base case/ matched maturity nominals

Nov 87

Sep 94

Aug 01

Jun 08

May 15

Apr 11

May 15

The demand backdrop

FRN issuance motivation and estimated benefits

Choice of reference indices and sample structures

19

U S

T R E A S U R Y

F L O A T I N G

R A T E

N O T E S

19

The majority of the Agency floater market is linked to Libor or Fed funds
Distribution by original maturity (years) Distribution by underlying benchmark type

CHOICE OF REFERENCE INDICES AND SAMPLE STRUCTURES

0-1 1-2 2-3 3-4 4-5 5-6 6-7 7-8 8-9 10-11 12-13 15-16 20-21 >30

0.2% 16.7% 75.5% 4.9% 0.3% 1.5% 0.1% 0.0% 0.0% 0.1% 0.1% 0.4% 0.1% 0.2%

1M Libor FF Effective 3M Libor Prime rate 3M T-bill CPI

64.8% 25.9% 5.4% 2.8% 0.7% 0.1%

About $153bn of Agency floaters are outstanding

currently, which is about 7.2% of the total Agency debt market

Most of these structures reference Libor or Fed

funds as an index
Demand for floaters linked to Libor and FF

Distribution by reset frequency

Monthly Daily Quaterly Weekly Semi-annually

64.9% 28.6% 5.6% 0.7% 0.2%

may be due in part to the deep and liquid derivatives markets based on these indices, allowing for efficient hedging of risks
These indices have disadvantages too

FNMA, FHLMC and FHLB floaters outstanding; $bn

350 300 250 200 150 100 50 0 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 125 98 123 70 45 152 318 267 250 214 137

exposure to banking system credit risk (in the case of Libor), and the Fed funds effective rate is distorted by IOER and related inefficiencies

20

The corporate floater market is also predominantly linked to Libor as an index rate
Distribution by original maturity (years) Distribution by underlying benchmark type High grade corporate floaters outstanding*; $bn

0-1 1-2 2-3 3-4 4-5 5-6 6-7

CHOICE OF REFERENCE INDICES AND SAMPLE STRUCTURES

3.42% 23.46% 34.89% 1.86% 13.76% 2.03% 5.24% 1.50% 0.28% 12.02% 0.22% 0.24% 0.63% 0.03% 0.00% 0.32% 0.09%

3M Libor 1M Libor FF Effective 6M Libor Prime Rate 3M T-bill

Distribution by reset frequency

96.10% 3.38% 0.38% 0.09% 0.04% 0.01%

550 500 450 400 350 300 250 256 444

509

392

7-8 8-9 9-10 10-11 11-12 19-20 12-13 28-29 29-30 30-31

267 212 216 2011

Quarterly Monthly Daily Semi-annually Weekly

96.13% 3.39% 0.42% 0.05% 0.01%

200 2005 2006 2007 2008 2009 2010

* Includes only index-eligible floaters. (Floaters with <$300mn outstanding and less than one year to maturity are excluded.)

21

Possible Options for a Reference rate index

Index Decrease Treasury's rollover risk Diversify Treasury's funding costs Reduce basis risk in the system Already Will likely Will likely used in appeal to appeal to retail existing money market investors markets investors Additional Notes Provides for a variety of reset frequencies from overnight to 12-month. More attractive to some investors as more closely linked to their liabilities. Subject to banking system funding pressures. Diversification benefits will be somewhat limited, since issues around sovereign credit concerns would likely also result in higher LIBOR. Nonetheless, with only 3 US banks in the USD Libor panel, some degree of diversification is likely. Typically indexed to weekly auction clearing rates. Treasury could also explore daily resetting to secondary/ constant maturity T-bills data released by the Fed, but this could reduce transparency. Would enable frequent resets keeping price of floater close to par and thus making it more attractive to investor types that value price stability such as money market funds. Will be of interest to investors who typically roll T-bills. While this market exists, very small percentage of Agency and Corporate FRNs are linked to T-bills (0.7% and 0.1% , respectively). Changes in Bill auction schedule would result in changes in floaters in which resets are linked to bill auctions. Not significantly different than TBills, risking cannibalization of T-Bill demand. Daily resets. Would enable daily resets keeping price of floater close to par and thus making it more attractive to investor types that value price stability such as money market funds. Predictably low in current rate environment. Subject to changes in the Fed's monetary policy. Future of Fed Funds market is uncertain, as future of GSEs and Fed chosen policy tool is uncertain. A highly visible rate, but would will be hard to hedge given basis risk with tradable markets. Could enhance Repo market itself. Could decrease demand for repo product and indirectly for nominal Treasuries.

LIBOR

yes

partially

yes

yes

yes

yes

CHOICE OF REFERENCE INDICES AND SAMPLE STRUCTURES

T-bills

yes

no

yes

yes

yes

yes

Fed funds eff. rate

yes

yes

yes

yes

partially

yes

Fed funds target GCF repo rate

yes

yes

no

no

yes

yes

yes

no

yes

no

partially

yes
22

Comments on structural characteristics of FRNs

Issuance in the existing floater market has been concentrated in maturities 5-years and in. More than 92%

of the Agency market and 61% of the corporate floater market were issued with an original maturity of less than 3-years.
Having a more frequent reset frequency will result in lower interest rate duration and thus lower price vol,

and could be more desirable to investors seeking stable value assets. Daily and Monthly resets are more typical in Agency FRNs, while quarterly resets are more typical in Corporate FRNs
Treasury should floor coupons payments at zero
CHOICE OF REFERENCE INDICES AND SAMPLE STRUCTURES

This does note necessarily mean a zero floor on observations of the floating index rate. For instance,

a note paying semiannual coupons, with daily accruals could result in negative observations on one or more days between coupon payment dates. Only the ultimate coupon payment needs to be floored at zero

23

Choice of index rate and final maturity could also be a determinant of incremental demand for the product
1-week average of GCF Treasury index versus 1-week average of par amount traded; %
0.30 0.25 0.20 0.15 0.10
CHOICE OF REFERENCE INDICES AND SAMPLE STRUCTURES

$bn
Index level Par amount traded

Depending on final maturity there could be significant demand for a Treasury Floater Indexed to either overnight fed funds or GC Repo, primarily from Money-market funds and liquidity portfolios. Demand for an FRN linked to either of these indices would likely be driven by:

250 200 150 100 50

2a-7 Money Market Funds (if the contingent final maturity is less than 397 days) Corporate Treasury accounts not set-up to trade repo Investment funds / Foreign accounts looking for a high quality floating rate asset

0.05 0.00
May 10 May 11 Feb 10 Nov 09 Aug 10 Nov 10 Feb 11 Aug 11 Nov 11

GC Index (GCFRTSY Index) versus effective Fed funds (FEDL01 Index); 1-week moving average; %
0.30 0.25 0.20 0.15 0.10 0.05 0.00 Jul 10 Jan 11 Jul 11 Jan 12 GC index Fed funds

The basis between GCF and fed funds is small on a smoothed basis, so returns on an FRN linked to either would likely be similar Both indices are amenable to daily resets, which would produce very low interest rate duration risk (but not spread duration), and thus lower price volatility. However, ratings agency guidelines favor indices that are more than 95% correlated to either fed funds or Libor, possibly making fed funds a better choice

24

Description of a sample 2Y FRN linked to 6M T-bill yields

Characteristics
Maturity: 2-years Coupon: Floating Payment Frequency: Semi-

Hypothetical annualized funding cost for a 2-year Treasury FRN linked to 6-month T-bill yields versus actual 2-year Treasury yield; %

6 5 4 3 2 1 0

Hypothetical annualized funding cost for 2Y FRN linked to 6M Tbills 2Y Treasury yield

Annual
Reference Index: The average
CHOICE OF REFERENCE INDICES AND SAMPLE STRUCTURES

auction yield of 6-mo T-bill Auctions during reference period

Day Count: Act / Act

Aug 01 Dec 01 Apr 02 Aug 02 Dec 02 Apr 03 Aug 03 Dec 03 Apr 04 Aug 04 Dec 04 Apr 05 Aug 05 Dec 05 Apr 06 Aug 06 Dec 06 Apr 07 Aug 07 Dec 07 Apr 08 Aug 08 Dec 08 Apr 09 Aug 09 Dec 09
25

Description of a sample 2Y FRN linked to the overnight fed funds and GCF rate indices
GC Index Floater Characteristics
Maturity: 2-years Coupon: Floating Payment Frequency: Semi-Annual Reference Index: GCFRTSY <Index> The Index is the weighted average interest paid

Average Fed funds Floater Characteristics

Maturity: 2-years Coupon: Floating Payment Frequency: Semi-Annual Reference Index: FEDL01 <Index> The index represents the volume-weighted

each day on General US Treasury Collateral in the dealer to dealer repo market.
CHOICE OF REFERENCE INDICES AND SAMPLE STRUCTURES

Average current daily volume is approximately

average of interest rates at which depository institutions lend balances at the Federal Reserve to other depository institutions overnight.
Day Count: Act / Act

$150bn.
Day Count: Act / Act

26

Conclusions

The demand backdrop is currently favorable for US Treasuries, but it is prudent for Treasury to consider broadening its issuance strategy to draw in more incremental demand Floating Rate Notes issued by Treasury are one such avenue, and could be attractive to money funds, investors seeking bonds with low duration risk, and possibly banks seeking to mitigate the accounting effects of some of the Basel III provisions

That said, any such incremental demand is likely to be modest in the near term

The current timing does not appear ideal, although initiating an issuance program now could allow Treasury to position itself to capitalize on a more favorable market environment

Term premium in the yield curve is currently at all time lows, and the risk to the path of the funds rate is biased asymmetrically towards higher rates since further Fed easing is not possible However, initiating a program now could help position Treasury for a future environment marked by higher term premia

CHOICE OF REFERENCE INDICES AND SAMPLE STRUCTURES

The choice of a floating rate index must balance the need for simplicity and transparency with the need to diversify Treasurys funding risk

Indexing to T-bills offers simplicity and transparency, but does not fully diversify funding cost risk GCF offers the prospect of daily resets and very low duration risk, but is a more complex choice that is mostly unknown to retail investors Libor offers simplicity & transparency, but this index creates exposure to banking sector credit for Treasury Indexing to average Fed funds rate offers simplicity and transparency, overnight reset frequency, and a viable derivatives market for risk management. In addition, a reasonably well developed FRN market exists in other sectors. Its appeal to retail investors needs to be further studied

27