EE156 Midterm Cheatsheet (v2)
Lectures 1B–8A plus HW1–HW7. Organized by lecture block, with formulas, lists, definitions and traps. Lecture conventions win where they differ from BMA (see the study guide, Part 3).
A — Time Value of Money, Interest Rates & Bonds, Yield Curve (L1B, L2A, L2B)
TVM core (L1B)
|
|
| FV / PV |
\(FV = PV(1+r)^t\); \(DF_t = 1/(1+r)^t\) (0 < DF < 1) |
| NPV (DCF) |
\(NPV = C_0 + \sum_{t\ge1} C_t/(1+r)^t\); >0 do it |
| Rate of return |
profit / investment; accept if > OCC (⇔ NPV > 0) |
| Rule of 72 |
\(T \approx 72/r\) (r in %); exact \(\ln 2/\ln(1+r)\); within 0.25 yr for 5–33%; exact near 8% |
| Total cost |
direct + opportunity cost (MBA: ~$160K + 2 yrs salary/experience/advancement) |
- Why $1 today > $1 tomorrow: invest earlier, impatience (liquidity), uncertainty. Holds in deflation.
- Safe $ > risky $. \(r\) = risk-free + risk premium. Discount at OCC = return on similar-risk investments; in %; estimated via comps.
Perpetuities & annuities (L2A, HW2 Q1) — first payment at t = 1
|
|
| Perpetuity |
\(C/r\) (also: endowment size to pay \(C\)/yr) |
| Perpetuity-due (pay now) |
\(C + C/r = C(1+r)/r\) |
| Growing perpetuity |
\(C_1/(r-g)\), need \(r > g\) |
| Annuity |
\(C\left[\frac1r - \frac1{r(1+r)^t}\right]\) |
| Growing annuity |
\(\frac{C_1}{r-g}\left[1-\left(\frac{1+g}{1+r}\right)^t\right]\) |
| Mortgage pmt |
\(C = PV\big/\left[\frac1r - \frac1{r(1+r)^n}\right]\), \(r\) = APR/12, \(n\) = 12·yrs |
- Growing annuity = growing perp from 1 − growing perp from \(t+1\) (worth \(C_1(1+g)^t/(r-g)\) at \(t\), discount \(t\)).
- Delayed first payment at \(T\): \(\frac{C/r}{(1+r)^{T-1}}\).
- Amortization: interest = \(r\) × balance; principal = pmt − interest (grows over time). Balance = PV of remaining payments.
- $1M, 7.5%, 30y → $6,992.15/mo; month 1: $6,250 interest, $742.15 principal.
Compounding (L2A)
- APR = stated, no compounding; per-period = APR/\(n\). EAR \(= (1+r/n)^n - 1\).
- 10% APR semiannual → EAR 10.25%. EAR ↑ with \(n\).
- APR from EAR: \(n[(1+EAR)^{1/n}-1]\). Quarterly rate from EAR: \((1+EAR)^{1/4}-1\) — not EAR/4.
- Continuous: \(EAR = e^r - 1\); growth \(e^{rt}\); from \(dA/dt = rA\). ≈ daily. Used in quant finance.
- 1 bp = 0.01%; 150 bp = 1.5%.
IRR (L2A)
- Rate where NPV = 0. Bond's IRR held to maturity = YTM.
- Caveats: assumes reinvestment at IRR; needs numerical methods; multiple IRRs if signs change more than once; prefer NPV.
- HW2: −15M, −5M, +25M at 11% → NPV $786,056; IRR 13.50%.
Bonds (L2A, L2B)
- Securities: stocks, bonds. Not: currencies, commodities (Howey Test).
- Issuers: Treasuries, JGBs, Gilts, OATs; FNMA (buys mortgages), GNMA (guarantees); munis; corporates — IG = BBB+, junk = BB and below.
- Quote as % of face: $1M at 99.5 = $995K (discount); 100 = par; 100.5 = $1.005M (premium).
- Zero = single "bullet" at maturity, trades at discount. Coupon bond = portfolio of zeros (STRIPS).
| Price |
Coupon vs YTM |
| Premium (>100) |
coupon > YTM |
| Par (=100) |
coupon = YTM |
| Discount (<100) |
coupon < YTM |
- \(P = \sum_{k=1}^{2T}\frac{c/2}{(1+y/2)^k} + \frac{100}{(1+y/2)^{2T}}\) = annuity + bullet. Compounding = coupon frequency.
- STRIP: \(P = 100/(1+y/2)^{2T}\) (3y@5% = 86.23; 30y@6% = 16.97). 6%/5y@4.25% = 107.81.
| Yield measure |
Definition |
| YTM |
IRR held to maturity |
| Current yield |
annual coupon / price |
| Nominal (coupon rate) |
annual coupon / par |
| YTC |
YTM assuming called at earliest date |
| YTW |
lowest of all scenarios |
- Bond ROI = coupon income + capital gain/loss.
- Duration = weighted-avg time to payments; price sensitivity to rates. Long bond: more rate + default risk.
- Rates up after issue: YTM ↑ ~same amount, price ↓.
Market language (L2B)
- Price ↑ ⇔ yield ↓. "Rally"/bull = rates ↓ prices ↑; "sell-off"/bear = rates ↑ prices ↓. Bondholder wants rates down.
- Treasuries: Bills <1y zero-coupon (4, 8, 13, 26, 52 wk); Notes 2, 3, 5, 7, 10y; Bonds 20, 30y.
- On-the-run (newest) = most liquid, tighter bid-ask → slight preference.
- Bid = buyer, offer/ask = seller. Liquidity = trade without moving price; measured by bid-ask tightness + depth. Thin retail markets → high prices/low yields.
- Money market = fixed income < 1 yr. ABS = auto/student/card loans. FX = largest by volume.
- Equity issuance ≪ bonds: bonds must be rolled; equity never matures.
- Yields mean-revert: 10y long-run avg ≈ 6%, last 20 yrs ≈ 3%.
Short selling (L2B)
- Borrow + sell; owe the asset; "cover" by buying back. Post margin; price ↑ → margin call → forced buyback; mass buyback → short squeeze. Shorting increases float.
Yield curve (L2B, HW2 Q4)
- Single discount rate = oversimplification; on-the-run Treasuries → term structure.
- Normal (up): term premium (a risk premium) for rising rates, unexpected inflation, default risk; + liquidity preference (HW sol).
- Drivers: monetary policy (overnight rate, expected policy, QE), rate expectations, market forces (long bond vs rolling shorts).
- Roll-vs-long breakeven: \((1+r_2)^2 = (1+r_1)(1+f)\) (not on slides — implied by "rolling" argument).
- Inverted: expected rate cuts (slowdown/recession), after aggressive tightening; expectations overwhelm term premium.
- France vs Germany: credit risk + liquidity → spread. Euro yields < US yields: consider FX.
Traps
- Semiannual: \(c/2\), \(y/2\), \(2T\) — also for STRIPS.
- Quarterly from EAR: 4th root, not /4.
- Perpetuity values one period before first payment.
- Coupon rate never changes after issuance.
- Rent-vs-buy: include opportunity cost of the cash.
B — Credit Spreads, Duration & Convexity; Real Rates, Inflation & TIPS (L3A, L3B)
Credit & Spreads (L3A)
- Credit = funds now for promise of repayment. "Credit" in markets = credit-sensitive debt (corporates, loans). Treasuries = "rates," not credit.
- Credit spread = corporate yield − same-maturity Treasury yield. Lower rating ⇒ wider.
- Spread curves usually upward sloping (default uncertainty grows with horizon) but can invert (expectations, liquidity).
- Other spread curves: MBS, munis, ABS, EM.
- Flight to safety (recession fear): sell corporates → buy Treasuries ⇒ Tsy yld ↓, corp yld ↑ ⇒ spread widens.
| View |
Trade (duration-matched) |
| Spread will widen / recession underpriced |
Short corp/MBS, long Treasuries |
| Spread will narrow |
Long corp/MBS, short Treasuries |
| Curve steepens (10y ↑ vs 2y) |
Long 2y, short 10y |
| Curve flattens |
Long 10y, short 2y |
| 3 legs |
Butterfly |
- Hedge notional: \(N_{T} = N_{c}\cdot D_c/D_T\). DV01 \(= N \cdot D_{mod} \cdot 0.0001\). Spread-trade P&L ≈ DV01 × Δspread (bp).
- CFO wants own bond price ↑ (yield ↓): tighter spread or lower Treasuries ⇒ cheaper new issuance.
- Quotes per 100: 99.5 on $1M = $995,000 (discount); 100.5 = premium. Zero = one bullet payment, trades at discount.
Duration
|
Formula |
Units |
| Macaulay |
\(D_M = \sum t\,w_t\), \(w_t = \frac{C_t/(1+r)^t}{P}\) |
years (periods ÷ 2 for semiannual) |
| Modified |
\(D_{mod} = \frac{D_M}{1+r} = -\frac{1}{P}\frac{dP}{dr}\) |
% per 1% (unitless) |
| Zero |
\(D_M = T\); \(D_{mod} = T/(1+r)\) |
|
| DV01 |
\(P\cdot D_{mod}\cdot 0.0001\) |
$ per bp |
- \(r\) = rate per period (YTM/2 for Treasuries).
- Proof: \(\frac{dP}{dr} = -\sum tC_t(1+r)^{-t-1}\) ⇒ factor \(\frac{1}{1+r}\) ⇒ \(\frac{D_M}{1+r}\).
- Coupon ↑ ⇒ duration ↓. Maturity ↑ ⇒ duration ↑. Yield ↑ ⇒ duration ↓.
- Duration never negative (price, yield always opposite).
Convexity
- \(C = \frac{1}{P}\frac{d^2P}{dr^2}\); discrete (annual): \(C = \frac{1}{P}\sum \frac{t(t+1)C_t}{(1+y)^{t+2}}\). Semiannual: compute in periods, ÷4 for years².
- Taylor: \(\frac{\Delta P}{P} \approx -D_{mod}\Delta y + \tfrac12 C(\Delta y)^2\).
- Positive convexity always helps holder (rise > fall for same |Δy|). Same yield & duration ⇒ hold the more convex bond.
- Convexity can be negative: MBS/ABS, callables (embedded options) — never standard bonds.
- Long bonds more convex (distant cash flows). Same-coupon price–yield curves cross at par where yield = coupon.
Spot curve → bond (HW3 Q1)
- \(P = \sum \frac{C_k}{(1+s_k/2)^{k}}\), \(k\) = half-years. Then solve YTM, then \(D_M\).
- HW3 official: 5y 4% bond → 96.77, YTM 4.73%, \(D_M\) 4.57; zero at 4.57y interpolated 4.63%. (Partial-solutions labels shifted one row.)
- Upward curve ⇒ coupon YTM > duration-matched zero (final CF dominates).
- HW3 Q2.1: 4.5% bond, 5y left, 103.50 → YTM 3.73%, \(D_M\) 4.54, \(D_{mod}\) 4.46. Q2.2 OAT 3%/3y/3.5% → €985.99.
Yield curve tidbits (L3A)
- 10y long-run avg ≈ 6%; last 20y ≈ 3%; yields mean-revert.
- Normal curve = term premium (rate risk, inflation risk, default risk).
- Inverted = expected cuts/recession; steepness = 2s10s.
- Spot→par: PV each CF, sum, → YTM. Par→spot: bootstrap from short end.
- Fed (monetary, independent) vs Treasury (fiscal, political). Türkiye 85% inflation 2022.
Inflation & Fisher (L3B)
- BLS: CPI, Core CPI (ex food & energy), PPI.
- Inflation good for net borrowers; bad for net positive financial assets. Real assets hold value.
- Deflation bad: delayed spending; real debt burden ↑. CPI −27% Great Depression. Fed target ≈ 2%.
- Exact: \(1+r = (1+r^*)(1+\pi)\) ⇒ \(\pi = \frac{1+r}{1+r^*}-1\), \(r^* = \frac{1+r}{1+\pi}-1\).
- Approx: \(r \approx r^* + \pi\); error = \(r^*\pi\). HW3: 5%/3% ⇒ exact 1.94%, approx 2.00%, ~6 bp.
TIPS
- Since 1997. Principal indexed to CPI; fixed real coupon × adjusted principal.
- Maturity: max(original, adjusted principal) — floor on principal only, not coupons.
- TIPS yield = real yield; can be negative (demand for inflation protection).
- Breakeven = nominal Tsy yield − TIPS yield = market's CPI forecast.
- Expect inflation > BE ⇒ long TIPS / short nominals; < BE ⇒ short TIPS / long nominals (duration-matched).
- Hold-to-maturity nominal IRR (constant π): \((1+y^*_{per})(1+\pi_{per})-1\). HW3 Q3.1: \(1.015\times1.02-1 = 3.53\%\)/half ⇒ 7.06% BEY (7.18% EAR).
- Purchasing power: TIPS, real estate, commodities, equities w/ pricing power.
Traps
- Macaulay (years) ≠ modified (%/1%); semiannual divide by \(1+y/2\).
- Price change uses modified duration, Δy in decimals.
- Convexity term sign is + regardless of Δy direction.
- Spot rates on Treasuries compound semiannually — don't use annual.
- Spread trades: match duration, not face, or you're betting on rates.
- "Credit" ≠ Treasuries.
- Breakeven: exact < approx when both positive.
- TIPS deflation floor only at maturity, only on principal.
C — Stocks & Markets, Valuation, Calculating ROI (L4A, L4B, L5A)
Stocks & Markets (L4A)
|
Common |
Preferred |
Bond |
| Claim |
Residual owner |
Owner, senior to common |
Lender |
| Votes |
Yes |
Usually no (VC exceptions) |
No — covenants instead |
| Payout |
Variable dividends + cap gains |
Fixed div (VC pref: often none) |
Fixed coupon + principal |
| Liquidation |
Last |
Before common |
Before all equity |
| Maturity |
None |
None |
Fixed |
- Hybrids (JP): corporate bonds (credit risk ↔ company performance) and preferred stock (fixed div + priority).
- Covenants (on corporate bonds/loans): D/E caps, interest-coverage minimums, dividend restrictions, change-of-control limits ⇒ protect creditors from actions raising default risk.
- Issuance: Seed → Series A/B/C (private) → IPO (primary) → FPO (primary; more common, less publicized) → trades on secondary (NYSE/Nasdaq). Company gets $0 from secondary trades.
- Capital structure = RHS of balance sheet. VC startup ≈ all (preferred) equity; PE ≈ debt-heavy (LBO).
- Value vs growth — no fixed line; use P/E (E/P), dividend yield, earnings growth. Value: low P/E, high div, stable (KO, JPM, XOM). Growth: high P/E, reinvests, volatile (AMZN, NVDA, TSLA).
- Malkiel: firm-foundation = intrinsic value (PV of CFs); castle-in-the-air = what next buyer pays (res tantum valet quantum vendi potest) ⇒ bubbles.
- Buffett: beat S&P 27/31 yrs (1976–2006), +11.1%/yr, t > 3 ⇒ skill.
Indices
| Index |
N |
Weighting |
Note |
| DJIA |
30 |
Price |
"Worst, most quoted"; hand-picked |
| S&P 500 |
500 |
Cap |
Pros' benchmark; ~80% US cap; rules-based |
| Nasdaq Comp. |
~3,000 |
Cap |
Tech-skewed |
| Russell 2000 |
2,000 |
Cap |
Small-cap benchmark |
Price-weight: weight\(_i = P_i/\sum P\) (share count ignored). Cap-weight: weight\(_i = P_iN_i/\sum P N\).
Strategies (L4A/L4B — 8 in summary)
| Strategy |
How it earns |
Names |
| Fundamental |
Intrinsic value (DCF) vs price, long horizon |
Graham, Buffett/Munger |
| Technical |
Patterns, momentum, mean reversion, support/resistance; candlesticks (body = open/close, wicks = hi/lo) |
Day/swing traders |
| Sentiment/behavioral |
Crowd mood (short), mispricing (long); contrarian, FOMO, biases |
Soros (reflexivity); Kahneman & Tversky |
| Quant/algo |
Stat arb, HFT, AI |
Simons: Medallion 66% gross / 39% net |
| Macro/event |
Rates, FX, geopolitics, M&A, commodities |
Dalio; Paulson ($4B MBS short '08); Soros (BoE '92) |
| Activist |
Stakes → force change; short activism; LBOs |
Icahn vs Ackman (Herbalife 2013 squeeze) |
| Market making |
Bid-ask spread; inventory risk |
HFTs (retail), IBs (institutions) |
| Derivatives |
Option spreads, vol, structured products |
Big IBs |
- Traders: short horizon, trading gains, higher short-term risk. Investors: long horizon, income + appreciation.
- Short-report activism: legal if accurate, illegal if false.
SEC Filings
10-K annual (audited) · 10-Q quarterly (unaudited) · 8-K current/major event · 13D >5% voting stake w/ intent to influence (13D/A amendment; 13G passive). TTM = trailing 12 months.
Valuation
- Earnings yield \(= E/P = 1/(P/E)\); \(EPS = P/(P/E)\).
- Comps: \(P_{target} = \text{multiple}_{comp} \times \text{metric}_{target}\) (P/E × EPS, P/S × sales/share, P/B × book/share). Average comps' multiples. P/E for mature/profitable; P/S for unprofitable/early growth.
- Precedent transactions: same, with M&A multiples.
- DDM: \(P_0=\sum \frac{DIV_t}{(1+r)^t} = \sum_{t=1}^{H}\frac{DIV_t}{(1+r)^t}+\frac{P_H}{(1+r)^H}\). Gordon: \(P_0=\frac{DIV_1}{r-g}\) ⇒ \(r=\frac{DIV_1}{P_0}+g\).
- DCF = DDM in theory: retained cash eventually reaches owners.
- Why use (uncertain!) methods: public ⇒ find mispricing; private ⇒ only option, weighted average of all four.
- Cash-neutral pair trade: $X per 1 pt outperformance ⇒ notional \(N = 100X\) each side; shares \(= N/P\). P&L \(= N(R_{long}-R_{short})\).
- Ex-dividend: \(P_{ex} = P_{cum} - D\). Ex-date = first day trading without the dividend.
ROI (L5A)
- ROI = income (dividends) + capital gains/losses (realized vs unrealized). Gains taxed lower; taxed only on sale.
- \(TR = \dfrac{P_1 - P_0 + D}{P_0}\) — SEC-required for funds, but ignores dividend TVM.
- Annualize: \((1+TR)^{1/T}-1\).
- CAGR \(=\left[\prod_{i=1}^{N}(1+R_i)\right]^{1/(NT)}-1\); price-only \((P_{end}/P_{beg})^{1/T}-1\).
- TR with reinvestment (dividend-aligned periods): \(\prod \frac{P_i+D_i}{P_{i-1}}\), then annualize.
- 10% & 20% ⇒ 14.89%, not 15%.
- Malkiel market return ≈ div yield + earnings growth + \(\left[(PE_{end}/PE_0)^{1/n}-1\right]\). HW5: 1.5 + 8 − 4.74 = 4.76% < 5.20% T-note.
- Long bond sold early: return = IRR(coupons + sale price at new yield). Rates ↓ ⇒ gain.
Capital Markets & Payouts
- Sell side (Goldman, MS, JPM, Deutsche) underwrites/sells; buy side (pensions, mutual funds, HFs, PE, insurers, SWFs) buys.
- Penny stocks: fail listing rules ⇒ OTC. Clearinghouse guarantees settlement.
- Block trade below screen: seller buys immediacy; dealer bears inventory risk, no guaranteed profit.
- Excess cash: keep only if project return > opportunity cost of capital. Buyback ⇒ fewer shares ⇒ EPS ↑, price holds (vs drops by div). Execs: pay tied to price/EPS. Holders: lower, deferred cap-gains tax. Illegal 1934–82; 1% excise tax now.
Traps
- Quarterly \(DIV\) with annual \(r,g\) — match periods.
- Arithmetic mean ≠ CAGR.
- Simple TR understates the reinvested TR.
- DJIA weights by price, not size.
- Ex-date: buyer does not get the dividend.
- IPO and FPO are both primary market.
- Pair trade: profit depends only on the spread of returns.
- HW5 official P/E comp value (\(506) ≠ recomputed **\)509.35** (28.615 × $1.78B / 100M).
D — Risk & Diversification, CAPM, Market Efficiency & Corporate Finance (L5B, L6A, L6B)
Risk & Return (L5B)
- Risk = (1) uncertainty of outcomes (volatility σ) or (2) probability of loss (VaR).
- VaR (99%, 1-day) \(= 2.33 \times \sigma_{daily} \times \text{Value}\) (normal, mean ≈ 0). HW6: \(2.33(1.5\%)(\$12M) = \$419{,}400\). Implied \(\sigma = \text{VaR}/(2.33 \cdot V)\).
- CAGR \(= \left(\prod(1+r_t)\right)^{1/T} - 1\), the geometric mean. +10%, +20% gives 14.89%, not 15%. Arithmetic mean ≥ CAGR.
- Total return = (div + ΔP)/P₀. It ignores dividend timing, so match periods to the dividend cycle. Reinvested ⇒ TR = IRR.
- US stocks: mean ≈ 10%, σ ≈ 20%.
- Same cash flows + more risk ⇒ lower price ⇒ higher expected return (risk premium). Junk vs Treasury.
| Market / systematic |
Idiosyncratic |
| economic cycles, rate changes, inflation/deflation, geopolitics, disasters, pandemics |
credit/default, liquidity, product failure, regulatory, management change, competition |
| not diversifiable; can be hedged (short the market) |
diversified away |
- Can't rank E[R] by σ for individual stocks; only for diversified portfolios (all risk systematic) does more σ mean more E.
- Risk-free rate = zero-coupon Treasury matching the horizon (STRIPS ≤ 30y). Coupon Treasuries carry reinvestment risk.
- Buybacks: anti-dilution raises the price. Pros: capital gains taxed lower and deferred until sale; exec pay tied to price/EPS; signals that management thinks the stock is cheap. Illegal 1934–1982; 1% excise tax.
Diversification (L5B, L6A)
- 2 assets: \(E_p = w_AE_A + w_BE_B\); \(\sigma_p = \sqrt{w_A^2\sigma_A^2 + w_B^2\sigma_B^2 + 2w_Aw_B\rho\sigma_A\sigma_B}\); \(\text{Cov} = \rho\sigma_A\sigma_B\).
- \(\rho = +1\): \(\sigma_p = w_A\sigma_A + w_B\sigma_B\) gives a straight line, no benefit.
- \(\rho = -1\): \(\sigma_p = |w_A\sigma_A - w_B\sigma_B|\) gives two lines meeting at σ = 0; \(w_A = \sigma_B/(\sigma_A+\sigma_B)\). Its return must be ≤ \(r_f\) (no arbitrage).
- Min-variance weight: \(w_A^* = \dfrac{\sigma_B^2 - \rho\sigma_A\sigma_B}{\sigma_A^2+\sigma_B^2-2\rho\sigma_A\sigma_B}\).
- N equal-weight stocks: \(\sigma_p^2 = \sigma^2/N + (1-1/N)\rho\sigma^2 \to \rho\sigma^2\) (floor = market risk). Higher ρ means a higher floor, flatter curve, less benefit.
- Anti-correlated asset = hedge. International diversification helps, but correlations are mostly positive (globalization).
- Bullet (Southwest/Amazon): risk-averse picks the min-variance point; risk-tolerant picks 100% of the high-return asset.
- Efficient frontier = max E per σ. Only one stock sits on it: the highest-E stock. Markowitz = father of MPT.
MPT, Sharpe, CML (L6A)
- Sharpe \(= (E_p - r_f)/\sigma_p\) = slope of the line from \(r_f\). Higher is better. It is the extra E per 1% of σ.
- Risk-free asset + any portfolio gives a straight line (σ_rf = 0, Cov = 0). The best line is tangent at T.
- Left of T = lending; right of T = borrowing (leverage), i.e. short the zero. In theory E has no cap.
- Everyone holds T + a long or short zero; preference only sets the leverage (indifference curve tangent to the line).
- Equilibrium: T = market portfolio (NVDA over-demanded ⇒ price up ⇒ weight up, E down ⇒ balance).
- CML: \(E_p = r_f + \dfrac{E_m - r_f}{\sigma_m}\sigma_p\) (efficient portfolios, uses σ). Weight in market \(= \sigma_{target}/\sigma_m\).
- HW6 Q2 (ABC 9%/27%, XYZ 15%/33%, ρ = 0.2, \(r_f\) = 4.5%): max Sharpe 0.3351 at 30% ABC / 70% XYZ (verified; exact optimum 30.7%). Betas (not in solutions text) via \(\beta = (E - r_f)/\text{MRP}\): ABC 0.9, XYZ 2.1.
Beta & CAPM (L6A–6B)
- \(\beta_i = \dfrac{\text{Cov}(R_i,R_m)}{\sigma_m^2} = \rho_{im}\dfrac{\sigma_i}{\sigma_m} = \dfrac{E_i - r_f}{E_m - r_f}\)
- Market β = 1. T-bill β = 0. Uncorrelated stock (any σ) has β = 0 ⇒ \(E = r_f\). \(\beta_p = \sum w_i\beta_i\).
- HW6: ρ = 0.8, σ = 2σ_m ⇒ β = 1.6. NVDA β = 2.3, \(r_f\) = 4.25%, MRP 5% ⇒ 15.75%.
- SML / CAPM: \(E_i = r_f + \beta_i(E_m - r_f)\). The x-axis is β; it applies to every asset; slope = MRP > 0 (risk aversion). A flat or downward SML contradicts CAPM.
- Alpha: realized − CAPM, \(\alpha = R - [r_f + \beta(R_m - r_f)]\). Regression: \(r_{i,t} - r_f = \alpha + \beta(r_{m,t} - r_f) + \varepsilon\). Positive α = beat CAPM.
- Only systematic risk is priced: the A (idiosyncratic) vs B (systematic) argument says investors buy and diversify A, its price rises, and its premium goes to 0. Before MPT, return was thought to track total risk.
- CAPM assumptions (lecture list): (1) rational, on average risk-averse investors; (2) only systematic risk matters (idiosyncratic diversified away); (3) frictionless (no costs, borrowing limits, taxes) and efficient markets. "Highly idealized."
- CAPM = single-factor. Fama-French 3-factor: beta, market cap, value (P/B).
- High β = high E with high market risk; low β = low market exposure. High CAPM E (US Steel) ≠ "better" than Newmont.
EMH (Fama, L6B)
| Form |
Reflects |
Useless |
Evidence |
| Weak |
past prices/volume |
technical |
generally supported (momentum, mean reversion seen) |
| Semi-strong |
+ all public info |
fundamental |
partially (PEAD, over/under-reaction, irrational exuberance) |
| Strong |
+ private/insider |
everything |
not supported (insider trading, privileged funds earn α) |
| - Paradox: markets must be slightly inefficient to attract the traders who make them efficient. |
|
|
|
Corporate finance (L6B, HW6 Q6)
- Corporation = separate legal entity (state law, often Delaware), shares. Limited liability, perpetual existence, access to capital. Purpose: risk management, growth, liquidity.
- Sole prop / partnership: no liability shield. S-corp: pass-through (no double tax), ≤100 owners, so VCs avoid it. C-corp: unlimited shareholders, the VC default (double taxed).
- Investment banks (GS, MS, JPM) underwrite (primary market). Commercial banks (Chase, BofA, Wells) handle deposits and loans. Big Four auditors: Deloitte, PwC, EY, KPMG.
- Secondary: exchange (clearinghouse guarantees settlement) vs OTC (dealers, counterparty risk; FX, penny stocks, corporate/muni bonds). The firm gets no secondary proceeds. Dark pools ≈ 40% of US volume.
Traps
- σ on the CML, β on the SML. 2. Don't rank individual stocks' E by σ. 3. ρ = −1 zero-vol portfolio can't beat \(r_f\). 4. β = 0 ≠ σ = 0. 5. CAGR, not the arithmetic mean. 6. The risk-free proxy is a zero matching the horizon, not a T-bill. 7. Market risk can't be diversified but can be hedged.
E — Governance, Accounting, Financial Statements, Cash Flows & Double Entry (L7A, L7B, L8A)
Governance and markets (L7A, L7B)
- Corporation: limited liability, perpetual existence, access to capital.
- "Hourglass": Shareholders → elect Board → appoints/supervises CEO → C-suite. Chairman sets agenda, runs votes; board acts by majority. CEO can fire CTO (reports to CEO).
- Shareholder levers: director votes, proxy fights, proposals, selling. Activists (Ariel → Mattel): "undervalued," sell to strategic buyer / break up, hire independent advisor.
- Primary market via investment banks; buyback = primary (JP). Secondary: exchange (clearinghouse) vs OTC (counterparty risk). Market makers = dealers regularly quoting bids/offers.
Accounting branches
| Branch |
Audience |
Rules |
| Financial |
external |
SEC + GAAP (public required; private often voluntary) |
| Tax |
external (IRS) |
IRS/local, ≠ GAAP |
| Managerial & cost |
internal |
none; decision support |
| Big Four: Deloitte, PwC, EY, KPMG. |
|
|
Three statements
|
BS |
IS |
SoCF |
| Time |
point |
period |
period |
| Equation |
\(A - L = E\) |
Rev − Exp = Profit |
Begin + In − Out = End |
| Tells you |
health |
performance |
liquidity / sustainability |
Balance sheet
- Current = within 1 yr. Assets in descending liquidity: cash → AR → inventory → prepaid → PP&E (cost − accum. dep.) → intangibles/other.
- Accrued = incurred, not yet paid. Equity = book value. Capital stock = paid in at issuance. \(RE_{end} = RE_{beg} + NI - Div\).
- Working capital = CA − CL.
| Ratio |
Formula |
| Current |
CA / CL |
| Quick (acid) |
(CA − Inventory) / CL |
| Cash |
Cash & equivalents / CL (no marketable securities) |
| D/E |
interest-bearing debt / equity |
| L/A |
TL / TA |
| Leverage (L8A reading) |
Debt / EBITDA |
| CR < 1 → can't cover near-term bills from near-term assets. High D/E → fixed claims, default risk, volatile equity. Cash equivalents < 3 mo maturity; marketable securities < 1 yr. |
|
Goodwill
- Goodwill = Price − FV(identifiable assets − liabilities). Use fair value, not book.
- Only from an acquisition; can't self-create (unmeasurable without a transaction).
- Impairment test annually; write-down hits goodwill and earnings (lecture: below-the-line one-time charge), non-cash.
Income statement
Net sales (net of refunds, discounts) [+ fees, rent, royalties = Revenue]
− COGS (direct; can include allocated fixed costs)
= Gross profit → Gross margin = GP / Rev
− Opex (S&M, R&D, G&A)
= EBIT = operating income/profit/earnings → Operating margin = EBIT / Rev
± Below the line: interest, one-time charges (impairments), non-operating gains (property sale)
− Taxes
= Net income = P&L = bottom line → Profit margin = NI / Rev
- COGS ≠ variable, opex ≠ fixed. Mfg overhead is in COGS. Commissions are variable opex.
- Semi-variable: utilities (fixed fee + usage), salary + commission.
- Operating loss → tax loss carryforward.
- Cash accounting: revenue when cash received. Accrual: when earned/delivered, matching principle; GAAP-required for public cos.
Statement of cash flows (L8A)
- "Companies fail when they run out of cash." Cash is King.
- Expenses → IS; payments → SoCF. Cash accounting: IS ≈ SoCF.
- Non-cash: shipping product, receiving supplies, depreciation, impairments, stock comp (expensed since 2005).
- Uses of cash: operations, financing, CapEx, taxes.
| CFO |
CFI |
CFF |
| customer receipts, supplier/employee payments, interest & taxes paid (GAAP) |
CapEx (−), asset sales (+) |
stock issue/buyback, borrow/repay, dividends |
Indirect CFO = NI + depreciation/amortization + impairments + stock comp − ΔAR − ΔInventory − ΔPrepaid + ΔAP + ΔAccrued + ΔTaxes payable.
Rule: operating asset ↑ = cash ↓; operating liability ↑ = cash ↑.
Direct CFO: (Rev − ΔAR) − (COGS + ΔInv − ΔAP) − cash opex − interest paid − (Tax − ΔTaxes payable). Same answer.
CapEx = ΔNet PP&E + depreciation + BV of assets sold.
Double-entry (L8A)
- \(A = L + CC + RE_0 + Rev - Exp - Div\).
- DR = left, CR = right; total DR = total CR every entry.
| Account |
Increase |
Decrease |
Normal balance |
| Asset, Expense, Dividends |
DR |
CR |
DR |
| Liability, Equity, Revenue |
CR |
DR |
CR |
| Contra (accum. dep., sales discount) |
opposite of pair |
|
|
Templates:
- Buy asset part on credit: DR Asset / CR Cash, CR AP.
- Credit sale w/ tax: DR AR (price × (1+t)) / CR Sales, CR Sales tax payable.
- Early pay, discount d: DR Cash, DR Sales discount (d × price), DR Sales tax payable (t × d × price) / CR AR. (L8A: 10,584 + 200 + 16 = 10,800.)
- Sale also: DR COGS / CR Inventory. Depreciation: DR Dep exp / CR Accum. dep.
Links
- NI − Div → RE (IS → BS). Δcash on SoCF = Δcash on BS.
- +$10 depreciation, tax 25%: NI −7.5, CFO +2.5, PP&E −10, assets −7.5 = RE −7.5.
- \(P = E \times P/E\); P/E "turns" = points of multiple.
Traps
- D/E excludes AP/accruals. 2. Cash ratio excludes marketable securities. 3. Goodwill uses fair value. 4. Buyback = primary (JP). 5. Collecting AR: total assets unchanged. 6. Depreciation tax shield raises CFO. 7. Interest paid: CFO (GAAP) vs financing use (JP's slide) — check convention.
Top exam traps (all blocks)
- Timing: "just paid" \(D_0\) vs next \(D_1\). A perpetuity's value lands one period before its first payment, and a perpetuity-due adds back one \(C\).
- Frequency: semiannual means halve the coupon and yield and double the periods. A quarterly rate from an EAR is \((1+EAR)^{1/4}-1\), not EAR/4.
- Par logic: coupon > YTM ⇒ premium. A bond's coupon never changes after issue; only price and YTM move.
- Duration: use modified (not Macaulay) duration for ΔP. Duration ≥ 0 always, while convexity can be negative (MBS, callables).
- Fisher: exact vs approximate. Check which one the question wants.
- bp vs %: 100 bp = 1%.
- Comps: average the multiples first, then multiply by the target's metric. Check the arithmetic, since the HW5 official P/E answer slipped.
- CAGR is a geometric mean. Never average returns arithmetically for growth.
- CML uses σ (efficient portfolios only); SML uses β (any asset). Market risk can't be diversified but can be hedged.
- D/E uses interest-bearing debt only. Goodwill comes only from acquisitions and is tested for impairment annually.
- Cash flow statement: depreciation is added back; ↑AR and ↑inventory are uses of cash; ↑AP is a source. Interest paid is in CFO under GAAP (JP's slide lists it with financing).
- Debits = credits always. Assets and expenses increase with a debit; liabilities, equity and revenue increase with a credit.