Financial Glossary

Current yield is a quick income snapshot: it divides the bond's annual coupon payment by its clean price (the market quote). Think of it as the 'income return' you'd earn if the bond's price stayed flat forever. Unlike YTM, current yield ignores capital gains or losses as the bond moves toward maturity, and it doesn't account for reinvestment of coupons. It's most useful for comparing income generation across bonds quickly, much like dividend yield for stocks.

Running yield measures the annual coupon income as a percentage of the bond's dirty price — that is, the clean price plus any accrued interest. Unlike current yield (which uses the clean price), running yield reflects what you actually pay for the bond at settlement. The difference between the two is typically small, but it widens as accrued interest builds between coupon dates. Running yield is a quick income measure but, like current yield, it ignores capital gains/losses and the pull-to-par effect as bonds approach maturity.

Yield to maturity (YTM) is the bond's internal rate of return — the single discount rate that makes the present value of all future cash flows (coupons and principal) equal the bond's dirty price. Think of it as the bond's 'all-in' yield, but with an important catch: YTM assumes you can reinvest every coupon payment at the same YTM rate, which rarely happens in practice. Because YTM appears inside the discounting formula, there's no closed-form solution — it's solved numerically using methods like Newton-Raphson or bisection, similar to IRR calculations in Excel. Despite its limitations, YTM is the industry standard for comparing bonds because it captures both income and capital gains/losses in a single number.

Dirty price (also called 'full price' or 'invoice price') is what you actually wire to the seller when you buy a bond. It equals the clean price (the quoted market price) plus accrued interest — the interest that has accumulated since the last coupon payment. Bonds are quoted using clean price so comparisons aren't distorted by coupon payment timing, but you always pay dirty price at settlement. For example, if a bond quotes at 98 (clean) with 0.5 in accrued interest, you pay 98.5 (dirty). The name 'dirty' is market slang reflecting that it includes the 'messy' accrued interest component.

Clean price is how bonds are quoted in most markets — it's the price stripped of accrued interest. This convention exists because a bond's full value oscillates between coupon payments (rising as interest accrues, then dropping when a coupon is paid). By quoting clean prices, you can compare bonds fairly without worrying about where each bond is in its coupon cycle. When you actually buy a bond, you pay the dirty price (clean price + accrued interest), but the clean price is what you see on your screen and what traders negotiate. Think of clean price as the bond's 'core value' before adding the interest that's built up since the last payment.

Accrued interest is the daily buildup of interest between coupon payments. When you buy a bond mid-cycle, you compensate the seller for the portion of the next coupon they 'earned' by holding the bond up to the sale date. When the next coupon arrives, you (the new owner) receive the full amount — but you already paid the seller for the days they held it. Think of it like buying a rental property mid-month: you'd reimburse the seller for rent they're entitled to from the 1st through the sale date. Accrued interest ensures fairness for both parties. The exact calculation depends on the bond's day-count convention (30/360, ACT/365, etc.), which determines how to measure the fraction of the coupon period that's passed.

Market value represents what your bond position is worth today. The standard formula is clean price (as a decimal) times notional amount. For example, a $1 million face value bond quoted at 98 has a market value of $980,000. The choice of clean vs. dirty price depends on context: portfolio accounting and risk metrics typically use clean price × notional for consistency, while settlement and cash management use dirty price to capture the full invoice amount. Some systems also add accrued interest separately to the clean MV when calculating total portfolio value. Always confirm which convention your risk system uses to avoid double-counting or missing accrued interest.

Carry is the 'do-nothing' return — the profit you earn from simply holding a bond while nothing else changes. It primarily comes from coupon accrual: as days pass, you accrue interest income. In leveraged contexts, carry is net of funding costs (repo or financing rates). For example, if a bond yields 5% and you finance it at 3%, your carry is ~2%. Traders distinguish carry from price appreciation caused by rate or spread movements. A 'positive carry' position earns money as time passes even if markets are static, while 'negative carry' bleeds value over time. Carry strategies focus on this time-decay component rather than betting on directional price moves.

The coupon rate determines how much interest a bond pays each year, expressed as a percentage of face value. For example, a 5% coupon on a $1,000 bond pays $50 annually, typically split into semiannual payments of $25. The term 'coupon' comes from the old practice of paper bonds with detachable coupons that bondholders would physically clip and redeem for interest payments. Most bonds today pay regular coupons (fixed or floating), but zero-coupon bonds pay no interest — they're issued at a discount and mature at par, with all return coming from price appreciation.

Yield curves are typically built from a limited set of observable bonds (e.g., 2Y, 5Y, 10Y, 30Y Treasuries), but you often need yields at intermediate maturities — say, 7 years — that don't have a liquid benchmark. Curve interpolation fills these gaps. Linear interpolation is the simplest method: draw a straight line between the two nearest points. For example, if 5Y yields 3.0% and 10Y yields 3.5%, linear interpolation estimates 7Y at 3.3%. This keeps the curve smooth without making complex shape assumptions. More sophisticated methods (cubic splines, Nelson-Siegel) produce smoother curves and are used when precision matters for derivatives pricing or portfolio analytics.

A bond benchmark is a standard reference used to compare a bond's yield, price, or performance against a relevant baseline. For individual bonds, the benchmark is typically an on-the-run Treasury bond with similar maturity — for example, pricing a 10-year corporate bond as 'Treasury +150bp' uses the 10Y UST as the benchmark. For portfolios, bond benchmarks are usually total-return indices (Bloomberg Aggregate, corporate indices, etc.) that capture coupon income, price changes, and reinvestment. Some systems use Treasury yield indices (like ^TNX) as simplified rate proxies, but these are yield-only and miss coupons, roll-down, and credit effects — suitable only for directional rate sensitivity, not total return tracking.

In bond markets, 'all-in' refers to the total or comprehensive level. An all-in yield is the bond's complete yield to maturity — the risk-free benchmark rate plus the credit spread. For example, if 10Y Treasuries yield 4.0% and a corporate bond trades at +200bp, the all-in yield is 6.0%. This contrasts with quoting spreads alone (which isolate credit risk) or Treasury yields (which isolate rate risk). The all-in yield captures both components in a single number, making it the actual discount rate used to price the bond's cash flows. In pricing models, the all-in yield can be input directly or constructed by adding a benchmark yield and a spread.

Implied YTM is the yield that, when you discount the bond's future cash flows, produces the bond's current market price. It's solved by working backwards from price to yield — the reverse of typical bond pricing. If you observe a bond trading at 95 (clean price), implied YTM tells you 'what yield the market is demanding given this price.' This is useful for price discovery: if your model assumes 5% YTM but implied YTM is 6%, the market is pricing in more risk than your assumption. Implied YTM is also how traders quickly infer yields when prices update faster than yield quotes.

Implied spread isolates the credit risk premium embedded in a bond's market price. It's calculated as the bond's implied YTM minus the interpolated benchmark Treasury yield at the same maturity. For example, if a corporate bond's implied YTM is 5.5% and the 10Y Treasury yields 3.5%, the implied spread is 200bp. Traders prefer quoting spreads over absolute yields because spreads isolate the bond-specific credit risk, while YTM mixes credit and rate risk. If Treasury yields rise 50bp and a corporate's YTM rises 50bp, the spread is unchanged — meaning credit risk didn't move, just the rate environment. Implied spread is especially useful when market prices move but you want to understand if credit perceptions changed.

Day count conventions are the mechanical rules for measuring time in bond interest calculations. Different markets and instruments use different conventions, which can produce noticeably different accrued interest and yield results for the same bond. Common conventions include: ACT/ACT (ICMA): Actual days / actual days in the period — used for US Treasuries; 30/360: Assumes every month has 30 days and the year has 360 — used for US corporate bonds; ACT/360: Actual days / 360 — common for money markets and floating rate notes; ACT/365: Actual days / 365 — used in UK and some Commonwealth markets. For example, 45 actual days of accrued interest on a 5% annual coupon under 30/360 gives 0.625% (45/360 × 5%), while ACT/365 gives 0.616% (45/365 × 5%). These small differences compound in present value calculations.

Par value (also called face value or principal) is the amount the bond issuer promises to repay the bondholder at maturity. For most corporate and government bonds, par is $1,000 per bond. Bond prices are quoted as a percentage of par: a price of 98 means the bond trades at $980 (98% of $1,000 par). Coupon payments are calculated as a percentage of par value, not market price. For example, a 5% coupon on $1,000 par pays $50 annually, regardless of whether the bond trades at $950 or $1,050. At maturity, bondholders receive par value plus the final coupon, regardless of the purchase price.

Pull to par is the phenomenon where bond prices gradually move toward face value (100) as maturity nears, assuming no default. A premium bond (price >100) declines toward par over time; a discount bond (price <100) appreciates toward par. This happens because at maturity, all bonds redeem at exactly par — the present value of future cash flows (which includes par repayment) converges to par as time passes. For example, a bond bought at 105 will decline to 100 by maturity, while a bond bought at 95 will rise to 100. Pull to par is a key component of horizon return (roll-down effect). The rate of pull depends on time to maturity and coupon structure — zero-coupon bonds exhibit pure pull to par with no coupon cushion.

A zero-coupon bond (or 'zero') pays no coupons—all return comes from buying below par and receiving par at maturity. For example, a 10-year zero might be issued at $600 and mature at $1,000, implying a yield of ~5.2% annually. Zeros are highly sensitive to interest rates (high duration) because there are no interim coupons to cushion price volatility. Uses: Tax-deferred accounts (to avoid phantom income tax on imputed interest), liability matching (pension funds), and duration management. Examples: US Treasury STRIPS (Separate Trading of Registered Interest and Principal Securities) are zeros created by stripping coupons from regular Treasuries. Corporate zeros are less common due to unfavorable tax treatment.

A discount bond trades below par (price <100), meaning you pay less than face value. This happens when the bond's coupon rate is below current market yields — investors demand a discount to compensate for below-market income. For example, a 3% coupon bond trades at 95 when market yields are 4%. Capital gain at maturity: If held to maturity, discount bonds appreciate to par (100), generating a capital gain in addition to coupon income. Tax treatment: The difference between purchase price and par (called market discount) is taxed as ordinary income at maturity, not capital gain. Duration: Discount bonds have longer duration than premium bonds with the same maturity because more of the return comes from the par repayment (farther in the future).

A premium bond trades above par (price >100), meaning you pay more than face value. This happens when the bond's coupon rate exceeds current market yields — the high coupon makes it attractive, driving the price up. For example, a 6% coupon bond trades at 108 when market yields are 4%. Capital loss at maturity: Premium bonds decline to par by maturity — you lose the premium paid. Your total return comes from high coupons minus the amortization of the premium. Tax treatment: You can amortize the premium over the bond's life, reducing taxable income each year. Callable risk: Issuers often call premium bonds when rates fall, forcing reinvestment at lower yields (call risk).

Settlement date is when the buyer pays cash and receives the bond. In U.S. Treasury markets, settlement is T+1 (one business day after trade). For corporate bonds, it's typically T+2. Settlement date matters for accrued interest: the buyer pays accrued from the last coupon date up to (but not including) settlement. If you buy a bond mid-coupon period, you compensate the seller for interest earned but not yet paid. Think of it as closing day for bonds: the trade date is when you agree to the price, but settlement is when money and securities actually change hands. Delayed settlement allows time for clearance, confirmation, and fund transfers.

The accrual period is the window during which interest 'accrues' on a bond. For a bond paying semi-annual coupons on June 1 and December 1, the accrual period from June 1 to December 1 is 183 days (or 6 months, depending on day count convention). If you buy the bond on September 1 (3 months after the last coupon), you owe the seller accrued interest for those 3 months. Think of it as rent for the bond: every day you hold the bond, you earn a little interest. The accrual period resets on each coupon date. Day count conventions (30/360, Actual/365, Actual/Actual) determine exactly how many days count in each period, which affects accrued calculations.

Flat price (synonymous with clean price) is the bond's quoted market price excluding accrued interest. If a bond is quoted at 98.50, that's the flat price. To settle the trade, you add accrued interest to get the dirty price (invoice price). Flat pricing simplifies comparisons: you can compare two bonds' quoted prices directly without worrying about where they are in their coupon cycles. Think of it as the sticker price on a car: it's the advertised price, but the out-the-door price (dirty price) includes taxes and fees (accrued interest). U.S. Treasuries and most corporate bonds are quoted flat. Exception: bonds in default or distressed situations may trade 'flat' in a different sense — without any expectation of receiving accrued interest.

Invoice price (also called dirty price or full price) is what the buyer actually pays: it equals the clean price plus accrued interest. If a bond is quoted at 98.50 (clean) and has $2.50 of accrued interest, the invoice price is 101.00. The seller receives compensation for the interest earned since the last coupon payment. Think of it as the total checkout price: the clean price is the item price, and accrued interest is like sales tax — not part of the quoted price, but you still have to pay it. Invoice price ensures fairness: without accrued interest, sellers would be penalized for selling between coupon dates, and buyers would get a windfall by collecting a full coupon they didn't fully earn.

Ex-coupon date is when a bond starts trading without the right to the next coupon payment. If you buy a bond on or after the ex-coupon date, the seller keeps the upcoming coupon even though you'll own the bond when it's paid. In the U.S. Treasury market, the ex-coupon date is typically one business day before the coupon payment date. For corporate bonds, conventions vary (often the record date). Think of it as the dividend ex-date for bonds: just like stocks, there's a cutoff for who gets the payment. After the ex-coupon date, the bond's price typically drops by roughly the coupon amount, reflecting the lost payment. Accrued interest calculations adjust accordingly — accrued may go negative between ex-coupon and payment date.

A make-whole call is an issuer-friendly but investor-protective call provision. Unlike a traditional call (which pays par plus a small premium), a make-whole call requires the issuer to pay the net present value of all remaining coupons and principal, discounted at Treasury yield plus a spread (often 10-50 bps). This makes calling the bond expensive unless rates have risen substantially. Example: If a 5% bond has 10 years left and Treasuries are at 3%, the make-whole price might be 115-120 (well above par). Issuers use make-whole calls for M&A flexibility (they can refinance debt if they're acquired) without giving bondholders traditional call risk. Think of it as an expensive escape hatch: the issuer can call anytime, but they have to fully compensate you for the lost income stream. Make-whole bonds trade at tighter spreads than non-callable bonds because call risk is minimal.

OAS is the true credit spread of a bond with embedded options (calls, puts, prepayment). Computed using a Black-Derman-Toy (BDT) binomial interest rate tree: at each node, short rates are lognormal with risk-neutral probability 0.5, calibrated to the zero curve. A constant spread (OAS) is added to every node rate, and backward induction prices the bond with option exercise rules (callable: value capped at call price; putable: floored at put price). Bisection solves for the OAS that matches the market dirty price. For option-free bonds, OAS ≈ Z-Spread (consistency check). For callable bonds, Z-Spread = OAS + option cost — the OAS is lower because the quoted spread includes compensation for call risk that OAS strips out. OAS lets you compare callable and non-callable bonds on equal footing. Lower OAS = lower credit risk.

Option value is the price impact of embedded options on a bond. Computed as the difference between the option-free tree price and the option-adjusted tree price, both at the solved OAS. For callable bonds: Option Value = V(option-free) − V(callable) > 0 (the call limits upside, costing the bondholder). For putable bonds: Option Value = V(putable) − V(option-free) > 0 (the put provides downside protection, benefiting the bondholder). Higher interest rate volatility → higher option value for both calls and puts. A deep out-of-the-money option has near-zero value. Expressed in price points per 100 par — e.g., option value of 1.50 means the embedded option is worth 1.50 points of price.

A bullet bond is the vanilla bond structure: you receive regular coupon payments, and the entire principal is repaid at maturity in one balloon payment (the 'bullet'). Most corporate and government bonds are bullets. Contrast with amortizing bonds (like mortgages), where principal is repaid gradually over time, or sinking fund bonds, where the issuer retires portions early. Think of it as borrowing for a house with interest-only payments and a giant final payment: you pay interest every period, but the loan balance stays constant until the end. Bullet bonds have higher reinvestment risk (you get a big chunk of cash at maturity that must be reinvested) but simpler cash flow modeling and duration calculation.

A sinking fund is a mandatory redemption schedule: the issuer must retire a percentage of the bond issue each year (e.g., 5% annually starting in year 10). Bonds to be redeemed are selected by lottery at par. For bondholders, sinking funds reduce credit risk (less debt outstanding = lower default probability) but create reinvestment risk (your bond might get called early at par, forcing you to reinvest at lower yields). Think of it as forced early retirement for bonds: instead of one big maturity, the issue shrinks over time. Sinking funds were common in the 1970s-80s but are rarer today. If your bond is called via sinking fund and you bought it at 105, you lose 5 points. Conversely, if you bought at 95, you gain 5 points.

G-Spread measures the yield premium a bond offers above the risk-free government curve. Unlike a quoted spread (which uses a single benchmark tenor), G-Spread interpolates the zero curve at the bond's exact maturity. Formula: G-Spread = Bond YTM − Interpolated Govt Yield(maturity). Example: A 7-year corporate bond yields 5.50%. The interpolated UST zero rate at 7 years is 4.20%. G-Spread = 130 bp. G-Spread is simpler than Z-Spread (which discounts each cash flow separately) but gives a quick read on credit compensation. CFA L1 tests this as the most basic spread measure.

Simple yield (also called Japanese yield) is a quick approximation: (Annual Coupon + Annual Amortization) ÷ Clean Price. It ignores compounding and reinvestment, so it's less accurate than YTM but useful for quick mental math. For a par bond, simple yield equals the coupon rate. For discount bonds, simple yield > YTM (amortization is spread too evenly). For premium bonds, simple yield < YTM. Think of it as back-of-the-envelope yield — good for sanity checks but not pricing.

YTC is the yield an investor would earn if the issuer exercises its call option — redeeming the bond before maturity at a predetermined price (usually par or a small premium). Calculation: Solve for the discount rate that equates the present value of cash flows up to the call date (plus the call price) to the market price. YTC is lower than YTM when the bond trades at a premium (issuer is incentivized to call). Yield to Worst (YTW) = min(YTM, all YTCs) — the most conservative yield scenario. Bloomberg YAS prominently displays YTW for callable bonds.

YTP is the yield an investor would earn if they exercise the put option — selling the bond back to the issuer at a predetermined price (usually par) on a scheduled put date. Calculation is identical to YTC: Solve for the discount rate that equates the present value of cash flows up to the put date (plus the put price) to the market price. Puts protect the investor — if rates rise and bond prices fall, the holder can put the bond back at par. YTP is typically higher than YTM when the bond trades below par (the put is 'in the money'). Putable bonds trade at a premium to non-putable bonds because the embedded put option has positive value to the investor.

YTW is the minimum of YTM and all YTCs — it answers: 'What's the worst yield I could get?' For callable bonds, if rates drop, the issuer calls the bond and you get the lower YTC. If rates rise, the bond runs to maturity and you get YTM. YTW is the most conservative estimate of return. Why it matters: Premium callable bonds often show attractive YTM but may be called early, giving you much less. Bloomberg shows YTW as the primary yield for callable bonds. CFA L1–L2 emphasizes YTW as the appropriate yield for callable bond analysis.

Forward rates are implied by the relationship between zero rates at different maturities. The 1Y rate, 2 years from now (the '2y1y forward') is derived from today's 2Y and 3Y zero rates. Formula (continuous): f(t1,t2) = (z₂×t₂ − z₁×t₁) / (t₂ − t₁). Interpretation: If you can lock in 4% for 2 years or 4.5% for 3 years, the implied 1-year rate starting in year 2 must be 5.5% (to make both strategies equivalent). Forward rates aren't forecasts — they're arbitrage-implied rates. But they're critical for: pricing FRAs and swaps, building term structure models, and the CFA L2 'three curves' framework (zero, forward, par).

The par yield at tenor T is the coupon rate that makes a bond price exactly at 100 (par), given the current zero curve. Formula: c = freq × (1 − DF(T)) / Σ DF(tᵢ) where DF is the discount factor from the zero curve. Par yields complete the three curves framework (CFA L2): Zero curve (spot rates), Forward curve (implied future rates), and Par curve. The par curve is what you'd see quoted for new bond issues — it's the fair coupon rate at each maturity. If the zero curve is upward sloping, par yields are slightly below zero rates (because interim coupons are discounted at lower short rates).

I-Spread measures the yield premium a bond offers above the interest rate swap curve at matching maturity. Unlike G-Spread (which uses the government curve), I-Spread uses the swap curve as the benchmark, making it more relevant for corporate bond relative value since swaps reflect interbank credit risk. Formula: I-Spread = Bond YTM − Interpolated Swap Rate(maturity). Example: A 5-year corporate bond yields 5.80%. The interpolated swap rate at 5 years is 4.60%. I-Spread = 120 bp. I-Spread is widely used by credit traders because the swap curve is smoother and more liquid than the government curve. CFA L2 covers I-Spread alongside G-Spread and Z-Spread as the three key spread measures.

The Asset Swap Spread represents the credit component of a bond's yield in swap terms. In an asset swap, an investor buys a fixed-rate bond and enters a swap to receive floating (e.g., SOFR) plus a spread. The ASW captures the difference between the bond's fixed cash flows and the par swap rate, adjusted for any premium or discount to par. Formula (simplified): ASW ≈ (Coupon − Swap Rate) + (100 − Clean Price) × freq / (TTM × freq) in percentage, converted to bp. For par bonds, ASW ≈ I-Spread. For discount bonds, ASW > I-Spread (the price discount adds extra spread). For premium bonds, ASW < I-Spread. Asset swaps are fundamental to credit relative value trading — they strip out interest rate risk, leaving pure credit exposure priced in floating-rate terms.

DV01 (Dollar Value of 01) measures interest rate risk in dollar terms. It tells you how much money you make or lose if yields move 1 basis point. For example, a $1 million bond position with 7.5 duration has DV01 ≈ $750 (7.5 × $1M × 0.0001). If yields rise 10bp, you lose ~$7,500. DV01 is more intuitive than duration for portfolio managers because it directly shows P&L impact. It scales linearly with position size: double your notional, double your DV01. Traders use DV01 to aggregate rate risk across different bonds and construct hedges. Convention: usually reported as a positive number for long positions, understanding that rising yields cause losses.

CS01 (Credit Spread 01) measures credit risk sensitivity in dollar terms: it tells you how much money you make or lose if the bond's credit spread widens or tightens by 1bp, holding the risk-free curve constant. For example, a corporate bond with CS01 = $500 loses $500 if its spread widens 1bp (from say +150bp to +151bp over Treasuries). CS01 differs from DV01: DV01 measures total yield sensitivity (rates + spread), while CS01 isolates spread risk only. It's most useful when bonds are priced as 'Treasury benchmark + credit spread,' allowing you to separate rate risk from credit risk. Traders use CS01 to hedge credit exposure independently of rate exposure.

Macaulay duration is the present-value-weighted average maturity of a bond's cash flows. It measures how long it takes, on average, for the bond's cash flows to 'pay back' the bond's price. Macaulay duration is always less than or equal to the bond's time to maturity (equal only for zero-coupon bonds). Modified duration is derived from Macaulay duration.

Modified duration measures interest rate sensitivity: it estimates how much a bond's price changes when yields move. A duration of 5.0 means a 1% yield increase causes roughly a 5% price drop. This is a first-order (linear) approximation — it works well for small yield changes but becomes less accurate for large moves (where convexity matters). For example, if a bond has modified duration of 7.2 and yields rise 0.5%, expect price to fall about 3.6% (7.2 × 0.5%). Textbooks often derive modified duration from Macaulay duration as MacDur / (1 + y/m). In practice, Strata computes it numerically using symmetric finite difference: the bond is repriced at yield ± 1bp and the slope of the price-yield curve is estimated from the two prices. This numerical approach handles irregular cash flows, odd first coupons, and day-count conventions that the closed-form formula cannot.

Convexity captures the smile curve in bond pricing: the price-yield relationship isn't a straight line (duration's assumption), it's curved. When yields rise, prices fall less than duration predicts; when yields fall, prices rise more than duration predicts. This asymmetry is valuable — convexity is a 'good' thing. Including convexity improves price estimates for larger yield moves: ΔP/P ≈ −Dur×Δy + 0.5×Conv×(Δy²). For example, with duration 6 and convexity 50, a 1% yield rise causes ~5.5% price drop (not 6%), while a 1% yield fall causes ~6.5% price gain (not 6%). Higher convexity means the bond exhibits less downside and more upside, all else equal.

Stress testing estimates how a bond or portfolio would perform under extreme but plausible scenarios — typically large parallel yield shifts (+/-100bp, +/-200bp). A simple stress uses duration: a 100bp yield rise with duration 6 implies ~6% price loss. But for large moves, you should include convexity for accuracy: ΔP/P ≈ −Dur×Δy + 0.5×Conv×(Δy²). Limitations: assumes parallel shifts (all maturities move equally), ignores spread changes, and can't capture options or embedded features. More sophisticated stress tests model non-parallel shifts (curve steepening/flattening), credit spread shocks, and historical stress scenarios (2008 crisis, 1994 bond selloff).

A parallel shift moves the entire yield curve up or down by the same number of basis points at every maturity. It's the simplest and most common stress scenario for fixed-income portfolios. While real yield curve moves are rarely perfectly parallel, this assumption provides a useful first-order risk estimate using duration.

A steepening yield curve occurs when the spread between long-term and short-term rates increases. This can happen through a 'bear steepener' (long rates rise faster than short rates) or a 'bull steepener' (short rates fall faster than long rates). Steepening typically signals expectations of future economic growth or inflation.

A flattening yield curve occurs when the spread between long-term and short-term rates decreases. This can happen through a 'bear flattener' (short rates rise faster) or a 'bull flattener' (long rates fall faster). Persistent flattening may signal economic slowdown expectations or tightening monetary policy.

Key rate duration measures how a bond's price changes when a single point on the yield curve shifts by 1bp while all other points remain fixed. Unlike modified duration (which assumes parallel shifts), key rate durations reveal where on the curve a bond is most sensitive. The sum of all key rate durations approximately equals the bond's effective duration.

Effective duration measures interest rate sensitivity for bonds with embedded options (callable, putable, mortgage-backed securities). Modified duration assumes cash flows are fixed, but callable bonds have uncertain cash flows — if rates fall, the issuer may call the bond. Effective duration captures this by calculating duration using option-adjusted cash flows. Calculation: Shock yields up/down by small amount (e.g., ±25bp), reprice bond with option model, measure price sensitivity. Example: A 10-year callable bond might have modified duration of 7, but effective duration of 4 if it's likely to be called in 5 years. When to use: Always use effective duration for MBS, callables, putables. Use modified duration only for option-free bonds.

Spread duration (credit duration) measures how much a bond's price changes when its credit spread widens or tightens, holding the risk-free curve constant. This isolates credit risk from rate risk. For example, a corporate bond with spread duration of 6 loses ~6% if its spread widens 100bp over Treasuries. Difference from modified duration: Modified duration measures total yield sensitivity (rates + spread). Spread duration isolates spread. Calculation: Similar to modified duration, but shock spread not yield. Use: Credit portfolio management, measuring pure credit exposure, hedging credit risk with CDS. For high-yield bonds, spread risk often dominates rate risk.

OAD measures a bond's true interest rate sensitivity after accounting for embedded options. Computed by bumping the entire zero curve ±25bp, rebuilding the BDT tree at each bump, and repricing with the same OAS. The finite difference gives the duration that includes option exercise behavior. For callable bonds near the call price: OAD < modified duration (the call caps upside, reducing sensitivity). For putable bonds: OAD < effective duration (the put floors downside). For option-free bonds: OAD ≈ effective duration (no option exercise in tree). OAD is the correct duration measure for callable/putable bonds — using modified duration overstates interest rate risk.

OAC measures the curvature of the price-yield relationship after option effects. Like OAD, it bumps the curve ±25bp, rebuilds the tree, and computes the second-order finite difference. For option-free bonds: OAC is always positive (prices are convex in yields). For callable bonds near the call price: OAC can be negative — as yields drop, the bond approaches the call price ceiling and price appreciation slows, creating negative convexity. This is a key CFA Level II concept: callable bonds exhibit negative convexity in the region where the call is in-the-money. For putable bonds: OAC is typically positive and higher than for option-free bonds.

The market-implied default probability is derived from credit spreads: PD ≈ Spread / (1 − Recovery Rate). This is the simplified CFA approach (risk-neutral PD). Example: 150 bp spread, 40% recovery → PD = 0.0150 / 0.60 = 2.5% annually. Key assumptions: constant hazard rate, spread = pure credit compensation (no liquidity premium). In practice, CDS-implied PDs are more accurate. Recovery rate varies by seniority: senior secured ~65%, senior unsecured ~40%, subordinated ~25%. Think of it as the break-even default rate that justifies the spread over Treasuries.

Recovery rate is what you get back after a default, expressed as a percentage of face value. Industry standard assumptions (Moody's long-run averages): Senior secured: ~65%, Senior unsecured: ~40%, Subordinated: ~25%. The complement (1 − Recovery Rate) is Loss Given Default (LGD). Recovery rates drive credit spread decomposition: Spread ≈ PD × LGD. In practice, recovery varies enormously: Lehman Brothers recovered ~28%, while most corporate defaults recover 30–50%. The 40% standard assumption is used for CDS pricing and CFA exam problems.

Horizon analysis estimates your total return from holding a bond for a specific period (say, 6 months or 1 year) and then selling it — rather than holding to maturity. Why use it? Most bonds are traded, not held to maturity, so horizon returns matter more than YTM. The total return decomposes into four components: carry (coupon income), roll-down (price change from aging along the curve), price effect (price change from yield/spread moves — your speculation/bet), and reinvestment (income from reinvested coupons). This breakdown shows what you're earning from 'automatic' sources (carry + roll-down) versus what depends on market moves (price effect). Traders use horizon analysis to compare bonds on a forward-return basis rather than static YTM.

Carry return is the portion of total return attributable to coupon payments received between settlement and the horizon date. It represents the income earned simply by holding the bond, before any price changes.

Roll-down return is the price change from the bond getting closer to maturity and 'rolling down' the yield curve, assuming the curve shape stays unchanged. If the curve is upward-sloping (normal), a 10Y bond yielding 4% becomes a 9Y bond yielding 3.5% after a year — price rises even with no yield change. This is 'positive roll-down.' If the curve is flat, roll-down is zero. If inverted, roll-down is negative. Curve dependency is key: steep curves offer more roll-down; flat curves offer none. For premium bonds (price >100), roll-down competes with pull-to-par (negative); for discount bonds, both effects are positive. Roll-down is a 'free lunch' in upward-sloping curves — you earn it just by holding the bond.

Price return (or 'price effect') is the uncertain, speculative component of horizon return: it's the gain or loss from yield/spread changes you don't control. If yields fall 50bp, price return is positive (you win); if yields rise 50bp, price return is negative (you lose). This is separate from roll-down (which assumes no yield change). Critically: carry and roll-down are somewhat predictable (you earn them if markets don't move much), but price return is pure speculation — you're betting on rate direction. Many traders focus on maximizing carry + roll-down while minimizing exposure to adverse price moves (negative convexity, duration mismatches). Horizon analysis helps you see which part of your expected return is 'locked in' versus dependent on your yield view being correct.

Reinvestment return is the additional income earned by reinvesting coupon payments at a given rate until the horizon date. It equals the future value of reinvested coupons minus their face value.

Total horizon return combines carry, roll-down, price effect, and reinvestment returns to give the all-in return from holding and selling a bond at the horizon. It's expressed as a percentage of the initial investment (dirty price × notional).

The efficient frontier is a curve in risk-return space representing portfolios that are 'mean-variance optimal'. Each point on the frontier offers the maximum expected return for its level of volatility (risk). Portfolios below the frontier are suboptimal because you could achieve higher return for the same risk or lower risk for the same return.

The Sharpe ratio measures how much excess return (above the risk-free rate) an investment provides per unit of risk (standard deviation). Higher Sharpe ratios indicate better risk-adjusted performance. It's commonly used to compare portfolios or strategies.

The covariance matrix is the core input for portfolio optimization: it shows how asset returns co-move. Diagonal elements are variances (each asset's standalone volatility), while off-diagonal elements are covariances (how pairs move together). For example, stocks and bonds often have low or negative covariance — when stocks fall, bonds may rise — enabling diversification. Diversification benefit: A portfolio of two 20%-volatility assets with 0.5 correlation has ~17.3% volatility, less than either asset alone. The matrix is symmetric (Cov(A,B) = Cov(B,A)) and typically estimated from historical returns, but estimation error is huge—small sample changes can radically alter results. Advanced methods (shrinkage, factor models) improve stability.

Portfolio volatility measures how much your portfolio's returns fluctuate — it's the standard deviation of returns. Unlike simply averaging individual asset volatilities, portfolio vol accounts for correlations: assets that don't move in lockstep reduce total risk. For example, a 50/50 portfolio of two 20%-vol assets with 0.3 correlation has ~16% vol, not 20%. This is the diversification benefit—the whole is less risky than the sum of parts. The formula σₚ = √(wᵀΣw) shows portfolio vol depends on weights (w), individual vols (Σ diagonal), and correlations (Σ off-diagonal). Lower correlation = better diversification. This is why global portfolios (stocks + bonds + alternatives) can achieve lower vol than stock-only portfolios.

The minimum variance portfolio sits at the leftmost point of the efficient frontier. It has the lowest volatility among all possible portfolios given the available assets. This portfolio prioritizes risk reduction over return maximization.

The maximum Sharpe ratio portfolio (also called the tangency portfolio) offers the best trade-off between excess return and risk. It lies where the Capital Market Line is tangent to the efficient frontier. Investors seeking optimal risk-adjusted returns often target this portfolio.

The Capital Market Line shows the risk-return trade-off for efficient portfolios that combine the risk-free asset with the tangency portfolio. Points on the CML represent the best possible risk-return combinations achievable by mixing cash and the optimal risky portfolio.

Maximum drawdown (MDD) measures pain: the biggest percentage drop from a portfolio's peak to its lowest point before recovering to a new high. A 30% MDD means at some point, your portfolio fell 30% from its peak. Unlike volatility (which treats upside and downside symmetrically), MDD captures the actual loss experience investors endure. For example, the S&P 500's MDD in 2008 was ~57% (peak Oct 2007 to trough Mar 2009). Recovery time matters: that drawdown took 4+ years to recover. Behavioral research shows MDD drives redemptions more than volatility — a 20% MDD feels worse than 15% annualized vol. Hedge funds and alternatives often target low MDD (10-15%) to keep clients invested through downturns.

The Sortino ratio is similar to the Sharpe ratio but uses downside deviation instead of total volatility. This focuses on 'bad' volatility (returns below the minimum acceptable return) rather than penalizing upside volatility. A higher Sortino indicates better risk-adjusted returns considering only downside risk.

The Calmar ratio compares a portfolio's annualized return to its maximum drawdown, measuring return per unit of drawdown risk. It's popular among hedge funds and CTAs. A Calmar ratio above 1.0 means the annualized return exceeds the worst historical drawdown.

Downside deviation only considers returns below a threshold (usually 0% or the risk-free rate), ignoring positive deviations. This provides a more intuitive measure of risk for investors who are primarily concerned about losses. It's used in the Sortino ratio as a replacement for standard deviation.

Rebalancing is the process of realigning portfolio weights to desired targets after market movements cause drift. For example, if equities outperform bonds, the equity allocation may exceed the target, requiring selling equities and buying bonds to restore balance. Rebalancing enforces discipline and can capture a 'rebalancing premium' from systematically buying low and selling high.

Asset allocation is the most important decision in portfolio management: how to divide capital across major asset classes (stocks, bonds, cash, real estate, alternatives). The famous Brinson study (1986) found that ~90% of portfolio return variability comes from asset allocation, not security selection or market timing. Strategic allocation sets long-term targets based on goals and risk tolerance (e.g., 60% stocks, 40% bonds). Tactical allocation makes short-term tilts based on market views (overweight stocks when optimistic). Classic frameworks: 60/40 stocks/bonds for balanced investors, 80/20 for growth, 40/60 for conservative. Modern approaches add alternatives (private equity, hedge funds) and inflation hedges (commodities, TIPS) for further diversification.

Risk parity allocates capital so each asset class contributes equally to portfolio risk, not equally by dollar amount. Traditional 60/40 (stocks/bonds) is dominated by equity risk (stocks are 3x more volatile). Risk parity might be 30% stocks / 70% bonds, using leverage on bonds to equalize risk contributions. Calculation: Allocate weights such that w_i × σ_i × ρ_i,p = constant for all assets. Example: If stocks have 15% vol and bonds have 5% vol, you'd hold 3x more bonds by weight to equalize risk. Benefit: Better diversification — no single asset dominates risk. Popularized by: Bridgewater's All Weather Fund. Criticism: Requires leverage, performs poorly when all assets decline together (2022: stocks and bonds both down).

The Black-Litterman model solves mean-variance optimization's biggest problem: extreme sensitivity to expected return inputs. Small changes in return assumptions cause wild weight swings. Black-Litterman starts with market-implied returns (reverse-engineering the CAPM equilibrium) as a neutral baseline, then tilts toward your views (e.g., 'I think EM will outperform by 2%'). Key innovation: Uses Bayesian updating to blend market equilibrium with your views, weighted by confidence. This produces stable, diversified portfolios instead of concentrated bets. Output: Expected returns that balance market consensus and your insights. Adoption: Widely used by institutional investors. Limitation: Still requires subjective view inputs, but handles them more gracefully than raw mean-variance.

The Kelly Criterion determines the optimal fraction of capital to risk on a bet (or investment) to maximize long-term wealth growth. Formula: f* = (p×b − q) / b, where p = win probability, q = loss probability, b = win/loss ratio. Example: A bet with 60% win chance and 1:1 payoff → Kelly = (0.6×1 − 0.4)/1 = 20% of bankroll. Insight: Never go all-in (Kelly is always <100% for realistic probabilities). Overbetting reduces growth; underbetting is suboptimal. In finance: Rarely used at full Kelly (too volatile). Practitioners use fractional Kelly (e.g., half-Kelly) for smoother equity curves. Criticism: Assumes you know p and b precisely (impossible in markets), ignores leverage constraints and path dependency.

VaR estimates the worst-case loss over a time horizon at a specified confidence level. For example, a 95% 1-day VaR of $100,000 means there's a 5% chance of losing more than $100,000 tomorrow. Critical limitation: VaR tells you the threshold, but not how severe losses beyond that threshold could be. A portfolio with $100k VaR could lose $101k or $1 million in the worst 5% of outcomes — VaR doesn't distinguish. This is why CVaR (Expected Shortfall) is often used alongside VaR — it measures the average loss in the tail. VaR is widely used for regulatory capital (Basel III), risk limits, and reporting, but should never be the only risk metric. Use it with stress tests, max drawdown, and scenario analysis.

Conditional Value at Risk (CVaR), also called Expected Shortfall, measures the average loss when losses exceed the VaR threshold. It provides information about the tail risk that VaR misses. For risk management, CVaR is considered more informative than VaR alone.

Monte Carlo simulation is a brute-force forecasting method: instead of using formulas, you run thousands of random scenarios and observe the distribution of outcomes. The process: (1) Estimate return distributions for each asset (mean, volatility, correlations); (2) Generate random draws from these distributions; (3) Simulate portfolio evolution over time (daily/monthly steps); (4) Repeat 10,000+ times; (5) Analyze the distribution of final wealth. For example, running 10,000 simulations of a 60/40 portfolio over 30 years shows the range of outcomes — median $2.5M, 10th percentile $1.2M, 90th percentile $4.8M. Monte Carlo captures non-linearity, path dependency, and rebalancing effects that formulas miss. Used for retirement planning, VaR estimation, and stress testing.

Skewness measures asymmetry in return distributions. Positive skew (right tail longer) means occasional large gains, many small losses — desirable. Negative skew (left tail longer) means occasional large losses, many small gains — undesirable crash risk. Zero skew is symmetric (normal distribution). Typical values: Equity indices ~0 to −0.5 (negative skew from crash risk). Individual stocks −0.5 to 0. Long-only portfolios typically negative. Options: Selling puts creates extreme negative skew (small premiums, rare blow-ups). Buying calls creates positive skew (small losses, rare home runs). Skewness matters because investors are asymmetrically sensitive to losses — a −30% crash hurts more than a +30% gain helps.

Excess kurtosis measures tail risk: how much more likely extreme events are compared to a normal distribution. Positive excess kurtosis (leptokurtic) means fatter tails — more frequent outliers. Most financial returns have excess kurtosis of 3-10, meaning 3-sigma+ events happen far more than the normal distribution predicts. Example: A normal distribution says a 4-sigma event occurs once in 15,000 days (~60 years). With excess kurtosis of 5, it's once every 1,000 days (~4 years). This is why VaR underestimates risk — it assumes normality. The 2008 crisis was a supposed '25-sigma event' under normality, but given realistic kurtosis, it was more like 4-sigma. Key: Higher kurtosis = don't trust VaR, use stress tests.

The Student's t distribution has heavier tails than the normal distribution, making extreme events more likely. It's characterized by degrees of freedom (df): lower df means fatter tails. At df=3-5, tail events are 2-3x more likely than normal. As df increases, the t-distribution approaches normal. It's widely used in finance to model realistic return distributions and stress scenarios.

Beta measures systematic risk - how sensitive the portfolio is to market movements. Beta of 1 means the portfolio moves with the market. Beta > 1 indicates higher volatility than the market (aggressive); Beta < 1 indicates lower volatility (defensive). Beta is calculated as the covariance of portfolio returns with market returns, divided by market variance.

Alpha measures skill-based outperformance: the return earned above what CAPM predicts given your beta. Positive alpha means you beat the risk-adjusted benchmark; negative alpha means you underperformed. Calculation example: Your portfolio returns 12%, risk-free rate is 2%, market returns 10%, your beta is 1.2. Expected return = 2% + 1.2×(10%−2%) = 11.6%. Your alpha = 12% − 11.6% = 0.4%—you outperformed by 40bp after adjusting for risk. Reality check: Most active managers have negative alpha after fees. Alpha is zero-sum (before costs)—every dollar of outperformance comes from someone's underperformance. Generating consistent alpha is extremely difficult; most 'alpha' is actually beta in disguise or luck.

Tracking error is the standard deviation of the difference between portfolio and benchmark returns (active returns). It measures the consistency of active management. Low tracking error means the portfolio closely follows the benchmark; high tracking error indicates significant deviation. Often expressed as an annualized percentage.

The Information Ratio (IR) measures consistency of outperformance: it's alpha divided by tracking error. High IR means you beat the benchmark consistently without wild swings; low IR means erratic results. Benchmarks: IR > 0.5 is good (difficult to achieve), IR > 0.75 is excellent, IR > 1.0 is exceptional. Why it's hard: Achieving IR > 0.5 means generating 5% alpha with 10% tracking error, or 2% alpha with 4% tracking error — both are tough. Compare to Sharpe ratio (total return / total risk): IR measures active return per active risk, while Sharpe measures total return per total risk. For index-hugging managers, IR is more relevant than Sharpe — it isolates the value-add from deviating from the benchmark.

R-squared (coefficient of determination) measures the percentage of portfolio return variance explained by the benchmark. R² of 1.0 means the portfolio moves perfectly with the benchmark; lower values indicate the portfolio has significant idiosyncratic risk. High R² makes beta and alpha estimates more reliable.

Historical stress testing applies actual market conditions from past crises (e.g., 2008 Financial Crisis, COVID-19 Crash) to estimate how your portfolio would have performed. It uses portfolio beta to scale market losses and provides a reality check for risk models based on actual extreme events rather than statistical assumptions.

VaR backtesting compares actual losses against the VaR threshold to validate model accuracy. A 95% VaR should be breached about 5% of the time. If breaches occur significantly more often, the VaR model is underestimating risk. The Kupiec test provides a statistical measure of whether the breach rate is acceptable.

Scenario analysis tests portfolio performance against hypothetical market conditions - either custom shocks (e.g., rates +200bp, equities -20%) or predefined scenarios (e.g., stagflation, risk-off). Unlike historical stress tests, scenarios can combine multiple simultaneous shocks to test specific concerns.

Expected Tail Loss (ETL), also called Conditional VaR (CVaR) or Expected Shortfall (ES), measures the average loss when losses exceed VaR. For example, if 95% VaR is $100k, ETL might be $150k — meaning when you're in the worst 5% of outcomes, you lose $150k on average. ETL is superior to VaR because it quantifies tail severity, not just the threshold. VaR says 'you'll lose more than $100k 5% of the time' but doesn't say how much more. ETL answers that: $150k on average. Basel III uses ETL (called ES) for trading book capital requirements because it's more sensitive to tail risk. Coherent risk measure: Unlike VaR, ETL satisfies all axioms of coherent risk measures (subadditivity, monotonicity, etc.).

Stress VaR estimates potential losses using historical stress scenarios (2008 financial crisis, COVID-19 crash, etc.) rather than recent market conditions. While standard VaR uses the past 1-2 years of data, Stress VaR uses extreme historical periods to ensure capital adequacy during crises. Basel III requirement: Banks must calculate both standard VaR and Stress VaR; capital requirements use the higher of the two. For example, standard 99% VaR might be $50M using calm 2019 data, but Stress VaR using 2008 conditions could be $200M — capital is based on $200M. Benefit: Prevents complacency during calm periods. Limitation: Assumes future crises resemble past ones (the next crisis is always different).

Marginal VaR measures how much portfolio VaR changes when you add $1 (or 1 unit) of a specific position. It answers: 'If I increase my Apple position by $1M, how much does my portfolio VaR increase?' Use: Risk budgeting — allocate risk to positions with the best return-per-unit-of-marginal-risk. Diversification insight: Marginal VaR can be negative—adding a position that's negatively correlated with the portfolio reduces total VaR. For example, adding bonds to a stock portfolio might have negative marginal VaR due to diversification. Component VaR: Marginal VaR × Position Size = Component VaR (each position's contribution to total VaR). Sum of all component VaRs = Total Portfolio VaR.

FX (foreign exchange) risk is the uncertainty in returns caused by currency movements when you hold assets denominated in foreign currencies. Example: You buy a EUR bond yielding 3%. The bond performs as expected, but EUR/USD falls 5%—your USD return is roughly −2% (3% yield − 5% FX loss). FX risk can amplify or dampen asset returns unpredictably. For US investors holding European stocks, FX risk can contribute 30-50% of total volatility. Key: FX risk is bidirectional — EUR appreciation boosts USD returns, depreciation hurts. Proper portfolio risk models must include FX volatility and FX-asset correlations (safe-haven currencies like CHF/JPY often appreciate during equity sell-offs, providing natural hedging).

Currency exposure measures how much of your portfolio's value is denominated in each foreign currency. A 30% EUR exposure means 30% of portfolio market value is in EUR-denominated assets. This exposure determines how sensitive your portfolio is to EUR/USD exchange rate movements. Unhedged currency exposure adds both risk and potential diversification benefits.

FX volatility measures how much an exchange rate bounces around, expressed as annualized standard deviation of log returns. Developed currencies (EUR/USD, GBP/USD, USD/JPY) typically have 6-12% annual vol — lower than equities (15-25%). Emerging market currencies (TRY, BRL, ZAR) often have 15-25% vol — comparable to equities. Why it matters: A 30% EUR position with 10% FX vol contributes ~3% to portfolio volatility (30% × 10%). This is often underestimated — FX can be a major risk source. Comparison: FX vol is roughly 1/2 to 2/3 of equity vol for developed markets, but similar for emerging markets. FX vol spikes during crises (EUR/USD hit 20%+ vol in 2008).

FX correlation determines whether currency exposure adds or reduces portfolio risk. Risk-on currencies (AUD, NZD, EM currencies) strengthen when equities rise (+0.3 to +0.6 correlation with stocks)—they amplify equity risk. Safe-haven currencies (JPY, CHF) strengthen when equities fall (−0.3 to −0.5 correlation)—they provide natural hedging. Example: Holding Japanese equities with JPY exposure creates a hedge — when Japanese stocks fall, JPY often appreciates, cushioning losses. Currency pairs: EUR/USD and GBP/USD are highly correlated (+0.7 to +0.9), while USD/JPY and AUD/USD are weakly correlated. Understanding FX correlations is critical for global portfolio construction — unhedged FX can either enhance or destroy diversification depending on correlation regime.

A cross rate is calculated when there's no direct quote between two currencies. Instead, both currencies are quoted against a common base (usually USD), and the cross rate is derived by dividing one rate by the other. For example, EUR/GBP can be derived from EUR/USD and GBP/USD quotes.

The spot rate is the prevailing exchange rate for a currency pair for settlement typically within two business days (T+2). It reflects current supply and demand conditions in the foreign exchange market. Spot rates are the basis for pricing all FX derivatives and forward contracts.

The base currency (or reporting currency) is the currency used to express total portfolio value, income, and risk metrics. When a portfolio holds assets in multiple currencies, all values are converted to the base currency using prevailing exchange rates for aggregation and reporting purposes.

FX hedging uses derivatives (usually forward contracts) to lock in exchange rates and eliminate FX risk. A fully hedged position removes all FX exposure — you lock in the asset return in local currency terms. Partial hedging (say, 50%) reduces but doesn't eliminate FX risk. Cost/benefit: Hedging EUR exposure costs ~interest rate differential (if EUR rates are lower than USD, hedging costs you the difference — currently ~0.5-2% annually). Key tradeoff: Hedging eliminates downside FX risk but also upside. Common practice: Many institutional investors hedge 50-100% of developed-market FX and 0-30% of emerging-market FX (EM hedging is expensive and illiquid). The formula F = S × (1 + r_d)/(1 + r_f) shows forward rates embed interest differentials.

Forward points are the premium or discount added to the spot rate to determine the forward rate, reflecting interest rate differentials between two currencies (covered interest parity). If EUR rates are lower than USD rates, EUR trades at a forward premium (positive points). Calculation: Forward = Spot + Points. For example, EUR/USD spot = 1.1000, 1-year forward points = +0.0050, so 1-year forward = 1.1050. Units: Often quoted in basis points or pips. Why they exist: Arbitrage ensures forward rates embed the interest differential — otherwise you could borrow in low-rate currency, convert to high-rate currency, and lock in risk-free profit. Use: FX hedging, calculating hedge costs.

Covered Interest Parity (CIP) is an arbitrage relationship: the forward exchange rate must equal the spot rate adjusted for interest rate differentials, otherwise arbitrage opportunities exist. Formula: F/S = (1 + r_domestic)/(1 + r_foreign). Example: If USD rates are 5% and EUR rates are 2%, EUR must trade at a 3% forward premium — otherwise you could borrow EUR at 2%, convert to USD, earn 5%, and lock in the future exchange rate via a forward, capturing risk-free profit. Covered means the FX risk is hedged with a forward. Empirical reality: CIP held tightly pre-2008 but broke down post-crisis due to bank balance sheet constraints, regulation, and cross-currency basis (the deviation from CIP).

Market capitalization (market cap) is the total dollar value of a company's outstanding shares. It equals share price × diluted shares outstanding. For example, a company with 100 million shares trading at $50 has a $5 billion market cap. This is the market's current valuation of the equity, not the whole company (which includes debt). Market cap determines size classifications: mega-cap (>$200B), large-cap ($10B-$200B), mid-cap ($2B-$10B), small-cap ($300M-$2B), micro-cap (<$300M). Larger caps tend to be more liquid and less volatile.

Enterprise value (EV) measures the total cost to acquire the entire company, not just the equity. It equals market cap (equity value) + debt − cash (+ minority interest and preferred stock, if material). Think of it as the theoretical takeover price: you buy all the equity (market cap), assume the debt, but get to keep the cash. For example, a company with $10B market cap, $2B debt, and $1B cash has EV = $11B. EV is used in valuation multiples (EV/EBITDA, EV/Sales) because it's capital-structure neutral — it values the operating business regardless of how it's financed.

The P/E ratio is the most widely quoted valuation metric: it tells you how many years of earnings you're paying for. A P/E of 20 means the stock costs 20x annual EPS. Trailing P/E uses TTM earnings (actual). Forward P/E uses next year's estimated earnings (consensus). Interpretation: Low P/E (<15) may signal value or pessimism; high P/E (>25) suggests growth expectations or optimism. Sector variation: Tech/growth stocks often trade at 25-50x, while banks/utilities trade at 8-15x. Limitation: P/E breaks down when earnings are negative, volatile, or manipulated by accounting. Use alongside EV/EBITDA and P/S for completeness.

EV/EBITDA is a levered-agnostic valuation multiple: it values the entire business (EV) relative to operating cash profit (EBITDA), ignoring capital structure. This allows apples-to-apples comparisons across companies with different debt levels. Typical ranges: Software 15-25x, Consumer goods 10-15x, Industrials 8-12x, Mature/cyclical 5-8x. Lower EV/EBITDA may signal undervaluation or low growth; higher suggests growth or quality. Why use it: EBITDA approximates cash flow before financing decisions, making it cleaner than net income (which includes interest, taxes, D&A). Limitation: Ignores capex intensity — high-capex businesses look cheap on EV/EBITDA but expensive on EV/FCF.

EV/FCF compares the total enterprise value to free cash flow, indicating how many years of current FCF the market is paying for.

FCF yield expresses free cash flow as a percentage of enterprise value. Higher yield implies more cash generation relative to valuation.

The PEG ratio adjusts the P/E multiple for expected earnings growth. A PEG of 1.0 suggests the stock is fairly valued relative to its growth; below 1.0 may indicate undervaluation. It is most useful when comparing companies with different growth profiles within the same sector.

EV/Revenue compares total enterprise value to top-line sales. It is commonly used for high-growth or unprofitable companies where earnings-based multiples are not meaningful. Lower EV/Revenue may indicate a cheaper valuation relative to sales.

The price-to-sales ratio values a company's equity relative to its revenue. Unlike P/E, it works for companies with negative earnings. It is capital-structure-dependent (unlike EV/Revenue), so it should be used with caution when comparing companies with different leverage levels.

The price-to-book ratio compares a company's market value to its book value (shareholders' equity). A P/B below 1.0 suggests the stock trades below net asset value — common for banks and capital-intensive industries. Growth companies often have P/B well above 1.0 because much of their value is intangible (brand, IP, moat).

DCF is the fundamental valuation method: it estimates a company's intrinsic value by projecting future free cash flows and discounting them to present value using WACC (the cost of capital). The process: (1) project FCF for 5-10 years explicitly, (2) estimate terminal value for all cash flows beyond the forecast (typically 60-80% of total value), (3) discount both back to today. For example, if you project $100M FCF growing to $150M over 5 years, then $3B terminal value, discounted at 10% WACC, you get enterprise value. Sensitivity is extreme: a 1% change in WACC or terminal growth can swing value 20-30%. Always run sensitivity tables. DCF reflects your view of the business, not the market's current mood.

The DDM values a stock by discounting expected future dividends at the required rate of return. The Gordon Growth Model is the simplest form, assuming constant dividend growth in perpetuity. It works best for stable, mature dividend-paying companies and breaks down when growth approaches or exceeds the discount rate.

SOTP valuation breaks a diversified company into its individual business segments, values each using the most appropriate method (comparable multiples, DCF, etc.), then sums the segment values. It is especially useful for conglomerates where applying a single multiple to the whole business would be misleading.

WACC is the required return a company must earn to satisfy all capital providers (equity and debt holders). It's the weighted average of cost of equity (via CAPM: Rf + β×MRP) and after-tax cost of debt, weighted by market values. Example: If cost of equity is 12%, cost of debt is 5%, tax rate is 25%, and the company is 60% equity / 40% debt, WACC = 0.6×12% + 0.4×5%×(1−0.25) = 8.7%. Use in DCF: WACC is the discount rate for free cash flows — it represents the opportunity cost of capital. Typical values: Stable companies 6-10%, Growth/tech 10-14%, High-risk/EM 14-20%. Sensitivity warning: A 1% change in WACC can swing DCF value by 20-30%—run sensitivity tables!

Terminal value (TV) captures the company's value beyond your explicit forecast (usually years 6-10 onward into perpetuity). Two methods: (1) Gordon Growth (perpetuity): TV = FCF_final × (1+g) / (WACC−g), assuming constant growth forever. Use g = GDP growth or inflation (~2-3%). (2) Exit multiple: TV = EBITDA_final × exit multiple (e.g., 10x), implying you sell the company at year 10. Critical: TV is typically 60-80% of total enterprise value, so terminal growth assumptions dominate the valuation. A 1% change in terminal g can swing value 20-30%. Always sensitivity-test terminal assumptions. Never use terminal g > long-term GDP growth (~2-3%) unless you believe the company will eventually become larger than the economy.

The Gordon Growth Model (GGM) is the simplest form of the dividend discount model. It values a stock as next year's dividend divided by the difference between the required return and the constant growth rate. The model is most appropriate for mature, stable companies with predictable dividend policies. It produces NaN when growth equals or exceeds the required return.

The H-Model is a two-stage dividend discount model where the growth rate declines linearly from an initial high rate to a long-term sustainable rate. Unlike abrupt two-stage models, the H-Model produces smoother transitions. The parameter H equals half the high-growth period, representing the midpoint of growth decay.

Cost of equity represents the rate of return investors expect for holding a company's equity, compensating for both the time value of money and the risk premium. It is commonly estimated using the Capital Asset Pricing Model (CAPM), which adds a risk premium (beta times the equity risk premium) to the risk-free rate.

Implied share price is calculated by dividing the model's equity value by shares outstanding. In a DCF, equity value is enterprise value minus net debt. In a comps model, equity value comes from applying peer multiples to the subject's fundamentals. Comparing implied price to market price reveals potential upside or downside.

A football field chart is the standard investment banking summary slide: it shows implied valuations from different methods (DCF, DDM, comps, precedent transactions) as horizontal bars, with the current market price as a vertical reference line. Interpretation: If most methods cluster at $50-60 and the stock trades at $40, it may be undervalued. If valuations are widely dispersed ($30-$80), there's high uncertainty. Narrow clustering suggests consensus; wide ranges suggest model sensitivity or disagreement. Typical layout: Y-axis lists methods (DCF bull/base/bear, EV/EBITDA comps, P/E comps, etc.); X-axis is price per share; bars show min-max range; current price line shows where market is relative to your analysis. Used in fairness opinions, pitch books, and equity research.

Precedent transactions (precedent M&A, transaction comps) values a company based on prices paid in actual M&A deals for similar companies. Unlike trading comps (which reflect minority stake, public market valuations), precedent transactions include a control premium—the extra amount acquirers pay for 100% ownership and strategic value. Methodology: Find recent deals (last 2-3 years) in the same sector. Calculate EV/EBITDA, EV/Revenue paid. Apply those multiples to your company. For example, if recent deals traded at 12x EBITDA and your target has $100M EBITDA, implied EV ≈ $1.2B. Control premium: Typically 20-40% above trading price. Use in M&A: Establishes floor valuation for acquisition bids. Limitations: Every deal is unique (synergies, timing, strategic fit), market conditions change.

Diluted shares outstanding represents the total number of shares that would be outstanding if all dilutive securities (stock options, restricted stock units, convertible bonds, warrants) were exercised or converted. This gives a more conservative estimate of ownership and per-share metrics. Treasury stock method for options: Assume options are exercised, company uses proceeds to buy back shares at market price — net new shares are added. Example: 100M basic shares, 5M options at $20 strike, stock at $40. Proceeds from exercise = $100M, buys back 2.5M shares at $40, so diluted shares = 100M + 5M − 2.5M = 102.5M. When to use: EPS calculations (always use diluted), DCF equity value (divide by diluted shares), valuation multiples (P/E uses diluted). Anti-dilution: Out-of-the-money options aren't included (not dilutive).

Net margin shows how much of each revenue dollar is retained as net income after all expenses, taxes, and interest.

EBITDA margin measures operating profitability before depreciation and amortization. It is often used to compare operating performance across firms.

Dividend yield measures the cash income you earn annually from holding the stock, expressed as a percentage of the current price. A 4% yield means $4 per year for every $100 invested. Typical ranges: Growth stocks 0-2% (they reinvest earnings), Dividend aristocrats 2-4%, High yield / REITs 4-8%. Watch out for value traps: An abnormally high yield (>8%) often signals distress — either the dividend is about to be cut or the stock has crashed. Payout ratio check: Dividend / Earnings should be <70% for sustainability. A 90% payout ratio with high yield is a red flag — limited margin for safety.

Beta quantifies systematic risk: how sensitive a stock is to broad market movements. Beta = 1 means the stock moves in line with the market (S&P 500). Beta > 1 means amplified moves (high-beta/aggressive). Beta < 1 means dampened moves (low-beta/defensive). Calculation: Regress stock excess returns on market excess returns; beta is the slope. Sector examples: Tech stocks 1.2-1.5 (cyclical, high growth), Utilities 0.5-0.8 (defensive, stable), Consumer staples 0.6-0.9, Financials 1.0-1.3. Use: CAPM cost of equity = Rf + β×(Rm − Rf). Higher beta requires higher expected return to compensate for risk. Limitation: Assumes linear relationship and stable correlations — both can break down in crises.

ROE measures how much profit a company generates per dollar of shareholders' equity. An ROE of 15% means the company earns $15 for every $100 of equity capital. Higher ROE signals efficient capital use, but beware: high leverage inflates ROE mechanically. The DuPont decomposition breaks ROE into three components: ROE = Net Margin × Asset Turnover × Equity Multiplier (or ROE = Profitability × Efficiency × Leverage). This reveals how a company achieves its ROE: through margins (pricing power), asset efficiency (capital-light model), or leverage (financial engineering). Typical values: 10-15% is average, 15-25% is strong, >25% is exceptional (high-moat businesses). Banks and leveraged firms often show >15% ROE structurally.

ROA measures how efficiently a company uses its total assets to generate profit. Unlike ROE, it is not affected by leverage, making it useful for comparing capital-intensive businesses. Higher ROA indicates more productive use of assets.

ROIC measures how effectively a company deploys its total invested capital (equity plus debt minus cash) to generate operating returns. It strips out capital structure effects by using NOPAT (net operating profit after tax). ROIC above the cost of capital indicates value creation.

FCF margin is the purest profitability measure: what percentage of revenue becomes free cash flow (operating cash flow − capex) available for dividends, buybacks, debt paydown, or M&A? A 15% FCF margin means $15 of every $100 in sales is distributable cash. FCF vs. EBITDA margin: FCF margin is always lower because it subtracts capex. A company with 25% EBITDA margin and 10% capex intensity has 15% FCF margin. Asset-light businesses (software, services) have FCF margins close to EBITDA margins. Capital-intensive businesses (telecom, utilities) have much lower FCF margins. Sector benchmarks: Software 20-35%, consumer goods 10-15%, industrials 5-10%.

Gross margin measures the percentage of revenue retained after subtracting the direct cost of goods sold. It reflects pricing power and production efficiency. Gross margins vary widely by industry — software companies often exceed 70%, while retailers may be below 30%.

Operating margin shows how much profit a company earns from its core operations before interest and taxes. It captures both gross profitability and operating expense control. Expanding operating margins often signal improving unit economics or operating leverage.

The current ratio is a liquidity metric that measures a company's ability to cover short-term obligations with short-term assets. A ratio above 1.0 indicates more current assets than liabilities. Extremely high ratios may suggest inefficient use of working capital.

The debt-to-equity ratio measures financial leverage by comparing total borrowings to shareholders' equity. Higher ratios indicate greater reliance on debt financing, which amplifies both returns and risk. Acceptable levels vary by industry — capital-intensive sectors typically carry higher D/E.

Interest coverage measures a company's ability to service its debt from operating earnings. A ratio below 1.5 is a warning sign; below 1.0 means the company cannot cover interest payments from operations. It is a key metric in credit analysis and bond covenants.

Asset turnover measures how efficiently a company generates revenue from its asset base. Higher turnover indicates more productive use of assets. It is one of the components of the DuPont decomposition of ROE (margin x turnover x leverage).

The quick ratio (acid-test ratio) is a conservative liquidity metric that excludes inventory from current assets, answering: 'Can the company pay bills if inventory can't be quickly sold?' Quick Ratio = (Cash + Marketables + Receivables) / Current Liabilities. A ratio >1.0 means the company can cover short-term obligations without selling inventory. Why exclude inventory? Inventory can be slow to liquidate or may sell at a discount in distress. Benchmarks: >1.5 is strong, 1.0-1.5 is adequate, <1.0 signals potential liquidity stress. Sector variation: Asset-light businesses (tech, services) often have >2.0; manufacturers may operate safely at 0.8-1.2 if inventory turns quickly. Compare to current ratio — if current is 2.5 but quick is 0.9, the company is heavily reliant on inventory (potential red flag).

Payout ratio measures what percentage of earnings is paid out as dividends: Dividends / Net Income. A 60% payout means $0.60 of every $1 earned goes to shareholders; $0.40 is retained for reinvestment. Sustainability check: <70% is generally sustainable, 70-90% is aggressive, >90% is risky (limited cushion for earnings volatility). Growth stage matters: Mature companies (utilities, consumer staples) often have 60-80% payouts — stable earnings, limited growth opportunities. Growth companies have 0-30% payouts — they reinvest in growth. Danger signs: Payout >100% means the company pays more than it earns — unsustainable, likely dividend cut ahead. Alternative for REITs: Use Dividends / FFO (funds from operations) since REITs have large non-cash D&A.

DSCR tells you if a company (or project) generates enough cash to cover its debt obligations. Formula: EBITDA ÷ (Interest + Principal Repayments). A DSCR of 1.5x means the company earns $1.50 for every $1.00 of debt service. Lenders require minimums: investment-grade corporates often need 2x+, real estate projects 1.25x+, leveraged buyouts accept 1.1x-1.3x. Think of it as a paycheck-to-rent ratio for companies: just as landlords want rent ≤ 30% of income, lenders want debt service well below EBITDA. DSCR < 1.0 means the company can't cover debt from operations — it must refinance, sell assets, or default. Unlike interest coverage (which ignores principal), DSCR includes the full debt burden, making it a more conservative credit metric.

The Efficient Market Hypothesis (EMH) is the cornerstone of modern finance. It claims that stock prices always incorporate all available information, making it impossible to 'beat the market' consistently through analysis, market timing, or insider knowledge (strong form). Three forms: Weak: Prices reflect all past trading data (technical analysis doesn't work). Semi-strong: Prices reflect all public information (fundamental analysis doesn't work). Strong: Prices reflect all information, public and private (insider trading doesn't work — empirically false). Implications: If markets are efficient, active management is futile — just buy index funds. Criticisms: Behavioral biases (overreaction, momentum), market anomalies (value premium, size effect), and bubbles/crashes challenge EMH. Most evidence supports semi-strong efficiency for liquid large-caps, but inefficiencies exist in small-caps and emerging markets.

CAPM is the foundation of modern finance: it says the expected return on an asset equals the risk-free rate plus a risk premium proportional to the asset's systematic risk (beta). Formula: E[R] = Rf + β×(Rm − Rf). Example: If Rf = 3%, market return (Rm) = 10%, and a stock's beta = 1.3, expected return = 3% + 1.3×(10%−3%) = 12.1%. Key insight: Only systematic risk (beta) is rewarded — idiosyncratic risk can be diversified away. Uses: Estimating cost of equity for DCF, evaluating risk-adjusted performance (alpha), portfolio construction. Criticisms: Assumes investors hold the market portfolio, single-period model, beta is stable (empirically false). Despite flaws, CAPM remains the dominant framework for cost of equity.

Modern Portfolio Theory (MPT), developed by Harry Markowitz (1952), is the mathematical foundation for portfolio construction. Core insight: Combining assets with imperfect correlation reduces portfolio volatility below the weighted average of individual volatilities — this is diversification benefit. MPT defines the efficient frontier: the set of portfolios offering maximum return for a given risk level (or minimum risk for a given return). Rational investors should only hold efficient portfolios. Inputs: Expected returns, volatilities, and correlations for all assets. Outputs: Optimal weights (min variance portfolio, max Sharpe portfolio, etc.). Limitations: Garbage in, garbage out — small changes in expected return assumptions cause massive weight shifts. Assumes normal distributions (ignores tail risk). Despite limitations, MPT underpins modern asset allocation.

Random Walk Theory states that stock price changes are independent and identically distributed—tomorrow's return is unrelated to today's (or any past) return. This is the weak form of the Efficient Market Hypothesis. Implication: Technical analysis is useless — chart patterns, momentum, and trends have no predictive power. Mathematical model: P_t = P_{t-1} + ε_t, where ε is random noise. Empirical evidence: Mixed. Short-term returns do exhibit slight autocorrelation (momentum effect), violating pure random walk. Long-term returns show mean reversion. Practical: While not perfectly true, random walk is a useful baseline assumption — beating the market consistently is very difficult.

Behavioral finance challenges the Efficient Market Hypothesis by documenting systematic psychological biases that cause irrational investment decisions. Key biases: Overconfidence (investors overestimate their skill), Loss aversion (losses hurt 2x more than gains feel good), Herding (following the crowd), Anchoring (fixating on irrelevant reference points), Recency bias (overweighting recent events). Market implications: These biases create anomalies: momentum (trends persist due to underreaction), value premium (mean reversion after overreaction), bubbles and crashes. Contrast with EMH: EMH assumes rational investors; behavioral finance assumes humans are predictably irrational. Practical: Understanding biases helps avoid costly mistakes (panic selling, chasing performance, overtrading).