Comprehensive Guide to U.S. Treasury Markets and Securities

U.S. Treasury market: why it matters

Deepest, most liquid government market in the world. Main way the U.S. government finances itself. Used by the Fed for monetary policy. Key investment/hedging vehicle for foreigners. RISK-FREE (from default) benchmark for pricing everything else. Backed by full faith and credit of the U.S.

Primary vs. secondary market

PRIMARY = Treasury auctions NEW securities to the public. SECONDARY = the trading market afterward, through brokers/dealers.

Treasury's debt management goal

Finance the government at the "LOWEST COST OVER TIME," operating in a "REGULAR AND PREDICTABLE" manner. Auctions cut costs DIRECTLY (broad competitive bidding) and INDIRECTLY (liquid secondary market).

Marketable vs. nonmarketable Treasuries

MARKETABLE: tradable, sold at auction, rates set by competitive bidding (bills, notes, bonds, TIPS, FRNs). NONMARKETABLE: can only be sold back to Treasury, sold by subscription, rates set administratively (Savings Bonds, SLGS = State & Local Government Series).

Bills vs. Notes vs. Bonds

BILLS: < 1 yr, NO coupon, sold at a discount, pay par. NOTES: 1-10 yrs (2, 3, 5, 7, 10), semiannual coupons. BONDS: > 10 yrs (20, 30), semiannual coupons. Notes and bonds pay par at maturity.

TIPS vs. FRNs

TIPS: 5, 10, 30-yr. Fixed REAL coupon paid semiannually on inflation-adjusted principal (indexed to NSA CPI-U). At maturity pays the GREATER of adjusted principal or par. FRN: 2-yr, QUARTERLY interest, WEEKLY reset, indexed to the 13-week bill auction rate.

FRN is a floating rate note!

What you bid in each auction

Notes & Bonds → YIELD. TIPS → REAL YIELD. Bills → DISCOUNT RATE. FRNs → DISCOUNT MARGIN.

Auction calendar + reopenings

Bills: WEEKLY (52-wk every 4 weeks). Notes, bonds, FRNs: MONTHLY. REOPENING = selling more of an EXISTING security (same CUSIP, coupon, maturity). 10- and 30-yr new issues in Feb/May/Aug/Nov, reopened in the 2 months after.

Competitive vs. noncompetitive bid

COMPETITIVE: you name the AMOUNT and the YIELD; you may get all, part, or none. Max award = 35% of the offering (less net long position). NONCOMPETITIVE: you accept whatever yield the auction sets; guaranteed the FULL amount. Limit = $10 MILLION par.

Single-price (uniform-price) auction

ALL winners, competitive AND noncompetitive, get the SAME price, set by the highest accepted yield (the STOP-OUT or HIGH yield). Used for every Treasury auction since Nov 1998. A lower bid yield = a more aggressive bid (willing to pay more).

Auction allocation order

(1) ALL noncompetitive bids filled first. (2) Competitive bids filled from the LOWEST yield upward. (3) At the stop-out yield, bids are PRORATED: amount left ÷ total bid at that yield. Any bid above the stop-out gets NOTHING.

AUCTION EXAMPLE: $25B 10-yr, $1B noncomp. A $5B @1.90, B $5B @1.92, C $5B @1.94, D $5B @1.96, E $6B @1.98, F $2B @1.98, G $5B @2.00

Noncomp $1B → $24B left. A, B, C, D all filled ($20B) → $4B left. $8B bid at 1.98% → 4/8 = 50% proration: E gets $3B, F gets $1B. G gets NOTHING. EVERYONE is filled at the 1.980% stop-out.

THE 1/8th RULE: how the coupon is set

Coupon = stop-out yield rounded DOWN to the nearest 1/8 of 1% (0.125%). Eighths: .125 / .250 / .375 / .500 / .625 / .750 / .875. Ex: 1.980% → 1.875% coupon. Because coupon ≤ yield, a new issue prices AT or slightly BELOW par.

1/8th practice: stop-out 4.412%? 4.249%? 4.834%?

4.412 → 4.375% (4 3/8). 4.249 → 4.125% (4 1/8), since it hasn't reached 4.250. 4.834 → 4.750% (4 3/4). (In a REOPENING the coupon is already fixed, so only the price changes.)

AUCTION PRACTICE #2: $20B 10-yr, $2B noncomp. A $6B @4.40, B $6B @4.42, C $4B @4.45, D $4B @4.45, E $5B @4.47. Who gets what? Stop-out? Coupon?

Noncomp $2B → $18B left. A and B filled ($12B) → $6B left. $8B bid at 4.45% → 6/8 = 75%: C gets $3B, D gets $3B. E gets NOTHING. STOP-OUT = 4.45%, everyone filled there. COUPON = 4.375% (4 3/8, rounded DOWN to the nearest 1/8).

EXAM CHECKLIST for any auction question

(1) Fill noncomp first. (2) Fill competitive from the LOWEST yield up. (3) STOP-OUT = the highest yield you need to accept. (4) Prorate at the stop-out: amount left ÷ amount bid there. (5) Bids above the stop-out get zero. (6) Everyone gets the stop-out yield. (7) COUPON = stop-out rounded DOWN to the nearest 1/8.

TAAPS vs. TreasuryDirect

TAAPS (Treasury Automated Auction Processing System): INSTITUTIONS, accepts competitive AND noncompetitive bids. TreasuryDirect: RETAIL/individuals holding directly with Treasury, accepts NONCOMPETITIVE bids only.

Book-entry + Fed's role in auctions

All Treasuries are electronic ledger entries (BOOK-ENTRY), held via the Commercial Book-Entry System (indirect, through your bank/broker) or TreasuryDirect (direct). Federal Reserve Banks act as Treasury's FISCAL AGENTS to run the auctions.

Primary dealers

Banks and broker-dealers designated by the NEW YORK FED that trade directly with it. They must: be counterparties in open market operations, give market commentary, bid in ALL Treasury auctions, and make markets for the Fed's foreign official accounts.

Bidder categories: Primary Dealer / Direct / Indirect

PRIMARY DEALER: dealers bidding for their OWN house account. DIRECT: non-primary-dealers bidding for their own house account. INDIRECT: customers bidding THROUGH a direct submitter (e.g., calling JPMorgan), including foreign central banks bidding through the Fed.

Who buys at auctions? Does the Fed bid?

Biggest buyer class = INVESTMENT FUNDS (mutual, money market, hedge funds, asset managers). The Fed does NOT bid competitively. It "ROLLS" maturing holdings at auction as an "ADD-ON," and can buy in the secondary market.

SOMA + QE

SOMA = SYSTEM OPEN MARKET ACCOUNT, the Fed's securities portfolio. QE = the Fed buys Treasuries and agency MBS with newly created money → prices ↑, yields ↓, borrowing costs ↓, pushes investors into riskier assets. Only the FED creates money; Treasury pays with REAL money.

Foreign holders of Treasuries

Largest: JAPAN (#1), UK, CHINA. WHY? TRADE: they export to the U.S., earn dollars, and park them in safe, liquid Treasuries. Tracked by the TIC (Treasury International Capital) System.

Average maturity of U.S. marketable debt

About 70.67 MONTHS (~5.9 years). The unit is MONTHS!

Coupon, YTM, current yield

COUPON: annual % of face, set at auction, paid SEMIANNUALLY ($100 at 4% = two $2 payments). YTM: the IRR that makes the PV of cash flows = price (the focus for this exam). CURRENT YIELD: coupon ÷ price.

Bond pricing + price/yield

Price = PRESENT VALUE of the cash flows. INVERSE relationship: yields ↑ → prices ↓. Premium: price > par (coupon > yield). Discount: price < par (coupon < yield).

Price quotes as % of par

Quoted per $100 face. 93 = $93 per $100 = $930 per $1,000 = $93,000 per $100,000.

32nds: "98-20" (or "98.20")

DOLLARS before the dash, 32nds AFTER: 98 + 20/32 = 98.625. Even with a decimal point, the digits after are STILL 32nds, not cents! 1/32 = $0.03125.

"+" / "2" or "¼" / "6" or "¾" in quotes

The third digit is EIGHTHS of a 32nd. "+" = HALF a 32nd: 98-20+ = 98.640625. "2"/¼: 98-202 = 98 + 20.25/32 = 98.6328125. "6"/¾: 98-206 = 98 + 20.75/32 = 98.6484375.

Quote practice: 98-19¼? 101-08+? 99-16?

98 + 19.25/32 = 98.6015625. 101 + 8.5/32 = 101.265625. 99 + 16/32 = 99.50.

Bid / Ask / Spread / "98-20 / 21"

BID = highest price a buyer pays (you SELL here). ASK/OFFER = lowest price a seller accepts (you BUY here); offer > bid. SPREAD = offer − bid, a LIQUIDITY measure. "98-20 / 21" = bid 98-20, offer 98-21.

Clean vs. dirty price + accrued interest

CLEAN/FLAT = quoted price, no accrued interest. DIRTY/FULL/INVOICE = clean + accrued (what you actually pay). ACCRUED = semiannual coupon × (days since last coupon ÷ days in period). Day count = ACTUAL/ACTUAL.

Why accrued interest exists (the "11-month" idea)

If you've held a bond 11 months toward a coupon, selling at the clean price gives away interest you earned. So the buyer pays the clean price PLUS accrued interest. Treasuries use an ACTUAL/ACTUAL day count.

Settlement + nomenclature "T 4 ⅜ 05/15/36" + GT10

SETTLEMENT = when cash and securities swap. Treasuries settle T+1 (trade date + 1 day). "T 4 ⅜ 05/15/36" = Treasury, 4.375% coupon, matures May 15, 2036 (ticker, coupon, maturity). GT10 = Bloomberg's GENERIC 10-yr.

On-the-run vs. off-the-run

ON-THE-RUN ("CURRENT") = most recently issued at a maturity; most liquid, traded electronically. OFF-THE-RUN = older issues: "OLD" (the one just before), "OLD-OLD," "OLDS." Off-the-runs trade by voice and RFQ.

Secondary market: then vs. now

THEN: principal-based, OTC, bilateral, voice. NOW: over $30T outstanding, ~$1.2T daily volume, hybrid PRINCIPAL/AGENCY (principal = trade from own book, agency = match buyers and sellers), increasingly ELECTRONIC.

D2C vs. Dealer-to-Dealer

DEALER-TO-CUSTOMER: slightly over HALF of volume, voice or electronic RFQ (request-for-quote), mostly bilateral clearing. DEALER-TO-DEALER: through INTERDEALER BROKERS (IDBs), heavily automated.

Electronic / Automated / HFT + Internalization

ELECTRONIC: screen-based, orders sent by computer. AUTOMATED: subset using algorithms. HFT: automated with low latency and high message rates. INTERNALIZATION: a broker fills an order from its own inventory or other customer flow instead of sending it to the interdealer market.

IDBs vs. PTFs

IDBs: platforms that aggregate bids/offers, stand as BLIND principal, provide ANONYMITY and price discovery. PTFs (Principal Trading Firms): trade their OWN money with automated/HFT strategies, huge volume, tiny end-of-day positions, now the MAJORITY of IDB volume.

When-issued (WI) trading

Trading a new security after the auction is ANNOUNCED but BEFORE it's issued (settles on the issue date). Serves as PRICE DISCOVERY and reduces auction uncertainty. The WI yield at 1 p.m. is the benchmark used to judge the auction.

STRIPS

Separate Trading of Registered Interest and Principal of Securities. The coupons and principal of a note/bond trade separately as ZERO-COUPON securities (deep discount, pay face at maturity). NOT issued directly to investors; available only through dealers/institutions.

C-STRIP / P-STRIP / fungible

C-STRIP = one coupon payment. P-STRIP = the principal payment at maturity. FUNGIBLE = interchangeable: coupon strips paying on the same date are identical no matter which bond they came from.

TIPS basics

Protect investors from INFLATION. Can be CHEAPER for Treasury because they remove the inflation risk premium. First 10-yr TIPS: JAN 1997. Index: NSA CPI-U (CPURNSA) with a 3-MONTH LAG. BOTH coupon and principal are adjusted. TIPS coupons are LOWER than nominal coupons (they're real rates).

Nominal Treasuries vs. TIPS: what moves prices

NOMINAL: Nominal yield = Real yield + Inflation. Real yields ↑ → price ↓. Inflation ↑ → price ↓. Inflation HURTS nominals. TIPS: Real yields ↑ → FLAT price ↓. Inflation ↑ → GROSS price ↑ (inflation is a FLOATING component).

TIPS: deflation + taxes

DEFLATION: the index ratio can drop below 1, cutting coupon payments, but at maturity you get at least ORIGINAL PAR. TAXES: inflation adjustments to principal are taxed YEARLY even though you only receive them at maturity, so hold TIPS in TAX-ADVANTAGED accounts.

Breakeven inflation + Fisher equation

BREAKEVEN ≈ NOMINAL yield − TIPS REAL yield = expected inflation + an inflation risk premium. FISHER (exact): (1+y/2) = (1+r/2)(1+i/2). Just know Fisher EXISTS; use the APPROXIMATION on the exam.

Fisher: Real Yield = nominal - inflation

EXAM Q: nominal or TIPS — which do you buy?

Expected inflation > breakeven → buy TIPS. Expected < breakeven → buy NOMINAL. Ex: 10-yr nominal 4.40%, TIPS 2.00% → B/E 2.40%. You expect 3%? Buy TIPS. You expect 2%? Buy NOMINAL.

Basis points + yield curve + 2s10s

1 bp = 0.01%. 1% = 100 bp. YIELD CURVE = snapshot of yields across maturities right now. 2s10s = 10-yr yield − 2-yr yield. 5s30s = 30-yr − 5-yr. STEEPENING = the spread widens; FLATTENING = it narrows.

Inverted yield curve

Short yields > long yields (negative spread). The market expects LOWER rates ahead, often recession. If the Fed hiked to 10%, the 10-yr would NOT go to 10%: rates that high can't last. Decent but imperfect predictor ("predicted 20 of the last 10 recessions").

10-yr yield = ? + term premium

10-yr ≈ AVERAGE EXPECTED SHORT RATES + TERM PREMIUM. Ex: 3% now, 5% expected next year, 7% the year after → 3-yr ≈ 5%. TERM PREMIUM = extra yield for holding long bonds beyond the expected rate path (issuance/deficit risk). Longer maturity = bigger price swings = more risk.

Treasury buybacks

PURPOSES: (1) LIQUIDITY SUPPORT, buying off-the-runs (long run). (2) CASH MANAGEMENT, adjusting the balance sheet (short run). (3) Lately, to push DOWN long-term yields (demand ↑ → price ↑ → yield ↓). Treasury pays with REAL money, so it has far less firepower than the Fed.

Why stocks fall when yields rise

(1) Higher discount rate → lower PV of future earnings (growth/tech hit hardest). (2) ~5% Treasuries compete with stocks. (3) Higher borrowing costs hurt profits and demand. FLIP SIDE: a stock crash → FLIGHT TO SAFETY → Treasury prices ↑, yields ↓.

How Fed HIKES can LOWER the 10-yr

Short-term hikes → economy cools, inflation expected to fall → lower expected FUTURE rates → long yields fall. Short rates ↑ make short-term borrowing more expensive, which slows demand and inflation.

FOMC basics

FEDERAL OPEN MARKET COMMITTEE: the Fed's rate-setting body. Sets the target range for the FED FUNDS RATE (the overnight bank lending rate) at 8 scheduled meetings a year. 12 voting members. Its decisions most directly move SHORT-term rates.

FED HIKES → 2-yr vs. 10-yr (KEY EXAM CONCEPT)

The 2-YR is the most sensitive to the Fed: it tracks where markets expect policy over the next 2 years, so it RISES A LOT. The 10-YR rises LESS, or can even FALL if hikes look credible (lower inflation and growth ahead). Result: the curve FLATTENS (2s10s narrows) and can INVERT.

FED CUTS → 2-yr vs. 10-yr

The 2-YR FALLS the most (tracks the Fed). The 10-YR falls less, since it still carries the term premium and longer-run growth/inflation expectations. Result: the curve STEEPENS (2s10s widens).

Bear/bull + flattener/steepener

BEAR = yields RISING (prices falling). BULL = yields FALLING. FLATTENER = spread narrows. STEEPENER = spread widens. Fed hiking → typically a BEAR FLATTENER (2-yr up more than 10-yr). Fed cutting → typically a BULL STEEPENER (2-yr down more than 10-yr).

SEPT 16, 2026 FOMC: what happened

The Fed HIKED 25 bp to 3.75%-4.00%, a unanimous 12-0 vote and the first hike since 2023. Chair KEVIN WARSH: inflation "too high for too long." The 10-yr briefly dipped below 4.95% and then went back to ~5%; the S&P 500 fell ~1%. Markets price more hikes ahead, mainly because of OIL.

Bond deals: why heavy CORPORATE issuance matters for Treasuries

More bond SUPPLY competes for the same investor dollars, and underwriters/investors often SELL or short Treasuries to hedge rate risk on new corporate deals. Both push Treasury yields UP and can weaken Treasury auction demand (cited as a headwind for the 10-yr auction).

AUCTION RESULTS: tail vs. stop-through

Compare the auction's high yield to the WHEN-ISSUED yield at the 1 p.m. deadline. TAIL = high yield ABOVE WI = WEAK demand (Treasury had to pay more). STOP-THROUGH = high yield BELOW WI = STRONG demand.

AUCTION RESULTS: bid-to-cover

Total bids ÷ amount sold. HIGHER = STRONGER demand. Compare to the average of recent auctions of the same tenor.

AUCTION RESULTS: reading the bidder shares

HIGH PRIMARY DEALER award = WEAK (dealers must absorb what others didn't want). HIGH INDIRECT share = STRONG (foreign/end-investor demand). Direct share = domestic end-investors.

AUCTION RESULTS: "concession"

Yields rising (prices cheapening) in the days/hours BEFORE an auction to make the new supply attractive. A concession HELPS demand; "no concession" hurts it.

READING: $13B 20-yr reopening tailed (Bloomberg)

1. The auction was weak. It tailed: the high yield of 5.420% came in above the 5.400% when-issued yield, so Treasury had to pay more than the market expected.

2. The other signals confirm it:

• Bid-to-cover was 2.57, below the 2.73 average, so there was less demand relative to supply.

• Primary dealers took 16.9%, the most since February. Dealers are stuck with whatever real investors don't want, so a high dealer share is a bad sign.

• Indirect bidders fell to 52.5%, meaning weaker demand from foreign and end investors.

3. It was a record-high yield for the 20-year, the highest since the tenor was reintroduced in 2020.

4. Why it was weak:

• There was no concession: yields hadn't cheapened ahead of the auction to attract buyers.

• The long end had recently outperformed, so it already looked expensive.

5. The aftermath: Yields rose further and the 5s30s curve steepened. Long bonds got cheape

READING: why the 20-yr tailed

Helped by the high outright yield and strong 10-yr/30-yr demand the week before, but HURT by NO CONCESSION on the day and the long end's recent OUTPERFORMANCE (it was already rich).

READING: $39B 10-yr auction stopped through (Bloomberg)

Awarded at 4.834% (highest since 2007) vs. 4.849% WI → 1.5 bp STOP-THROUGH = STRONG. Dealers only 4.3% (among lowest ever), indirects 79.25%, directs 16.5%. Bid-to-cover 2.71 (highest since 2016) vs. 2.52 avg. Yields fell after the auction.

READING: why the 10-yr was strong

Helped by the high OUTRIGHT YIELD, a CONCESSION on the day (yields up ~5 bp), and speculative SHORT positions in futures (shorts buy to cover). Offset by a heavy CORPORATE new-issue calendar (competing supply) and recent 2s10s flattening.

READING: Quarterly Refunding (what it is)

Every quarter (Feb/May/Aug/Nov) Treasury announces the refunding auctions: 3-yr, 10-yr, 30-yr. Both 2026 statements: $125B total = $58B 3-yr + $42B 10-yr + $25B 30-yr, auctioned on a YIELD basis at 1 p.m. on consecutive days.

READING: refunding math — new cash raised

New cash = amount offered − maturing securities held by the public. MAY: $125B − $83.3B = ~$41.7B new cash. AUG: $125B − $96.3B = ~$28.7B new cash.

READING: refunding — auction sizes + bills/CMBs

Treasury expects to KEEP coupon and FRN auction sizes the same "for at least the next several quarters" (regular and predictable). Bigger in new-issue months (10-yr $42B vs. $39B reopenings). Surprise borrowing needs are handled with BILL sizes and CMBs (cash management bills).

READING: refunding — TGA, buybacks, 20-yr change

TGA (Treasury General Account) = Treasury's cash balance at the Fed, ~$900B-$1.05T. Buybacks each quarter: up to $38B OFF-THE-RUN for LIQUIDITY support + up to $25B (1-month to 2-yr) for CASH MANAGEMENT. May: 20-yr reopenings now settle sooner, SHORTENING the WI period to reduce repo specialness.

READING: WSJ "Bond Yields Could Come Down as Fast as They've Climbed"

The recent rise in the 10-yr toward 5% is mainly markets pricing MORE FED HIKES (odds of 3+ hikes went from 16% to ~85%), NOT inflation fears (breakeven ~2.4%, 5y5y ~2.3%) or term premium (slightly DOWN since July). So it could REVERSE quickly (oil falls, the Fed looks credible, recession).

READING: WSJ "What Comes Next, Now the 10-Year Has Crossed 5%?"

5% is a CROSSROADS: a ceiling or a new era? Four forces: (1) FED policy (2) GROWTH and INFLATION/OIL (Brent ~$109) (3) STOCKS: a crash → flight to safety → yields ↓ (4) Treasury BUYBACKS. The 10-yr touched 5%, a 19-yr high; before this it had hit 5% only once since 2008 (2023).

READING: WSJ buyback backfire

Treasury said it would buy up to $6B of 10-20 yr bonds, but yields ROSE (markets expected more). It then bought only $5.2B (not enough offers at market prices), and yields rose AGAIN. Expectations it couldn't meet made it backfire.

READING: WSJ "Hedge Funds Are the Wild Card"

PENSIONS are pulling back from bonds (U.S. fixed income ~40% → 10-15% of assets) and HEDGE FUNDS are filling the gap: ~$2T, a record 7% of the market. The main strategy is the BASIS TRADE: cash Treasuries vs. futures, amplified with LEVERAGE. The NY Fed is asking about the risks.

READING: hedge funds — pros vs. cons

PROS: add LIQUIDITY, act as counterparties in swaps, help auctions go smoothly. CONS: SHORT holding periods, PRICE-SENSITIVE (demand higher yields), and LEVERAGED, so they can dump bonds fast and AMPLIFY stress in a crisis. Backdrop: deficit ~6% of GDP; China and Japan own fewer Treasuries.