P1.T2 Quantitative Analysis (Chpt 3 of Stock and Watson)

[skoh]

About standard deviation, on question 209.1, the answer is C. The standard error is derived by (0.15x 0.85÷60)^0.5. Is this an alternative method to calculate standard error besides S÷√60? is 0.15x0.85 the standard deviation?

On question 210.2, I'm still confused why the answer is c instead of d? How can we determine if 1.65 or 1.96 should be used?

Thanks

 

[skoh]

I have found this link to compute duration http://www.bionicturtle.com/how-to/...lus_ti_ba_ii_essential_functions_for_the_frm/ using Texas BA II Plus. How do we determine the amount of basis shock if we were to solve question 25.1 of Tuckman's Fixed Income Market risk?

Thanks

 

[David Harper CFA FRM]

Hi skoh,

It's a little easier if you append to the discussion at the source questions.

• source 209.1 is here, with discussion: http://forum.bionicturtle.com/threads/p1-t2-209-t-statistic-and-confidence-interval.5318/
  the variance of a Bernoulli (e.g., PD) = p*(1-p) = PD*(1-PD); and the standard deviation is therefore SQRT[PD*(1-PD)]
  o Standard Error = SQRT(15%*85%/60) is SE = SQRT(variance/n) or, equivalently, SE = Standard Deviation/SQRT(n).
• source 210.2 is here http://forum.bionicturtle.com/threads/p1-t2-210-hypothesis-testing-stock-watson.5336/
  VaR is always one-tailed; i.e., 1.645 @ 95% and 2.33 @ 99%
  But significance tests can be either one or two-sided. The default assumption is two-sided, however the questions says:
  "Her one-sided alternative hypothesis, stated with 95.0% confidence, is that DSO is greater than 27.0."
  The answer would be (d) if the question read something like:
  "Her alternative hypothesis, stated with 95.0% confidence, is that DSO is different than 27.0."
• Your choice for basis shock is somewhat arbitrary: it will not make much of a difference if you select 10, 20 or 50 bps. Technically, since we are just computing the slope of a secant, the smaller the shock, the more accurate the effective duration should be w.r.t the modified duration. So, I think 10 bps is a good choice. Thanks,