bump
Build up method for estimating required return on assets does not rely on beta’s.
Accounting risk — Financial statement and disclosures, not clear, misleading, incomplete, mistated Asset risk – excessive compensation or other perquisties Liability risk – excessive obligations, off-balance sheet obligations Strategic policy risk – risk of entering M&A, incur other business risks, not best long term interest of shareholder. resulting in large payoff of management
I’m surprised no one summarizes the GDP topic in Econ yet. This is very testable and easily overlooked (just like BOP last year). GDP + Foreign income = GNP GNP - Depr = NNI GDP is measured before taxes. You can convert it into factor GDP by subtracting indirect taxes and adding subsidiaries. Three ways to measure GDP: output, expenditure and income. For expenditure method, we measure GDP = TFE (Total final expenditure) - imports, which is calculated by: Total consumption + investment = Total domestic expenditure (TDE) + export = Total final expenditure (TFE) - import = GDP Output is measured in both constant and current price, expenditure measured in current price (but is remeasured using GDP deflator) and income is measure in current price (but is remeasured using inflation index). GDP deflator can be used to measure inflation (but doesn’t work in hyperinflationary environment).
Type II Error: OJ Simpson, “OJ” has two letters… Null hypothesis is innocence. Fail to reject that he’s innocent. Breusch sounds like a White guy who is Heterosexual: Testing for Heteroskedasticity: Breusch-Pagan test, use White method to correct for it.
SS18: In Developing Portfolio Constraints, just remember to be like a Turtle T = Time Horizon U = Unique Needs R = Regulatory/Legal Constraints T = Taxes L = Liquidity
For Type I and Type II errors, this is how I remember it: [Is H1 true?] and [Fail to reject H0?] True and True => Type I False and False => Type II
Treynor Black Optimal Risky Portfolio “Super Shortcut”: if you blank on this formula, the last step requires taking the square root. therefore, they **might** give a misleading answer that forgets to take the square root (the last step) so if it is a) .7160 b) .5147 c) .0479 The answer would be (a) because the root of .0479 is nothing the root of .5147 is .7160 therefore, the answer is (a)
eltia Wrote: ------------------------------------------------------- > I’m surprised no one summarizes the GDP topic in > Econ yet. This is very testable and easily > overlooked (just like BOP last year). > > GDP + Foreign income = GNP > GNP - Depr = NNI > > GDP is measured before taxes. You can convert it > into factor GDP by subtracting indirect taxes and > adding subsidiaries. > > Three ways to measure GDP: output, expenditure and > income. For expenditure method, we measure GDP = > TFE (Total final expenditure) - imports, which is > calculated by: > > Total consumption > + investment > = Total domestic expenditure (TDE) > + export > = Total final expenditure (TFE) > - import > = GDP > > Output is measured in both constant and current > price, expenditure measured in current price (but > is remeasured using GDP deflator) and income is > measure in current price (but is remeasured using > inflation index). > > GDP deflator can be used to measure inflation (but > doesn’t work in hyperinflationary environment). econ is what? 5-10? i spent all my time on tri arb and exch rate. and then the class, neoclass, and new growth/endoes//// gotta gamble on some topics. pleeeez let this not be tested, but i may review now that you brought it up
GDP Deflator showing up on AM = me not showing up for PM
Also, don’t forget to look over the various scatter plots in Quant. e.g. how do you detect unit root / non-covariance stability by look at scatter plots, why upward sloping is not always a good thing, etc.
eltia Wrote: ------------------------------------------------------- > Also, don’t forget to look over the various > scatter plots in Quant. e.g. how do you detect > unit root / non-covariance stability by look at > scatter plots, why upward sloping is not always a > good thing, etc. why upward sloping is not always a good thing ???
Upward / downward sloping is bad for time series, because it implies non-constant mean (meaning it’s not covariance stationary). Contracting or expanding movement around the mean isn’t a good sign neither, as it implies non-constant variance (again, it’s not covariance stationary). When I did the CFAI text questions, I was like “What? I thought an up market is a good thing…”
Forward looking estimates for equity premium. Three approaches: macroeconomic (Ibbotsen-Chen model), survey estimates (expert opinions) or GGM related approach. For GGM related approaches, there are single stage and multi-stage models. For the former, EPremium = delta_{1} + g + R_{LT,0}, where delta_{1} is the leading dividend payout ratio, g is estimated growth and R_{LT,0} is yield on long term Government bond. For the latter, we estimate the IRR from equity index price = PV_{rapid}(IRR) + PV_{transition}(IRR) + PV_{stable}(IRR), and then EPremium = IRR - Government bond yield.
Purchase Method under US GAAP and IFRS: 1) Fair allocation at acquisition date: US GAAP: (% ownership * fair market value) + ((1-% ownership) * book value)’ IAS: 100% fair value 2) Acquired in-process R&D (IPR&D) US GAAP: fair value for goodwill computation only; expense at Income Stmt immediately IFRS: Amortized 3) Contingent consideration US GAAP: No effects now; Record in future when assets/equity are transferred IFRS: Estimated and include in purchase price now Goodwill Impairment US GAAP: 2-step i) if carry value > fair value, go to next step (carrying value include goodwill) ii) loss (on Income Stmt) = carry value - implied fair value (implied fair value: same calculation as goodwill at acquisition date) IFRS: 1-step: loss = carry value - recoverable amount Don’t worry about the terms: they should be provided. (hopefully)
Just realized that we should be calculating the risk neutral probalities for valuing option using binomial model. I thought I would use 50% in either direction… That would have been wrong.
As far as binomial models are concerned, there are two application: bond valuation in FI and option valuation in Derivatives. There are two major favours for option valuation: bond or stock. Stock option valuation is evaluated by propagating stock values from root to leaves and then propagate the option values back to the root. Bond option valuation are done via backward propagation (leaves to root), and then the option values are propagated from leaves to root. Bond valuation are always backward propagated. For stock option valuation, we assume risk neutral probabilities and constant interest rate. For bond and its option valuation, the probabilities are always 0.5 and the interest rate is specific at each node. Note that for American call option valuation on bonds, you must compare the calculated option value to the intrinsic option value at each node (and take the maximum value).
<> I totally forgot this. Thanks for the reminder
Yield volatility measurement in FI. X_{i} = ln(Y_{i}/Y_{i-1}) daily std. dev. = std. dev. of X annual std. dev. = std. dev. of X x (# of trading days)^{1/2} Confident level for Yield = Yield +/- [k x (annual std. dev. x Yield)], where k is the number of standard deviation given a particular confident level. e.g. k = 1 when confident level is 68% under normal distribution. For forecast volatility, assume mean of X = 0 when computing std. dev. of X.
FCFE = NI - (1-DR) * (Fcinv - Dep) -(1-DR) * Wcinv