CFA CFA Level 1 z scores.

z scores.

  • This topic has 4 replies, 2 voices, and was last updated Apr-18 by googs1484.
  • Author
    Posts
    • emmanuel.ch
      Participant
      Up
      14
      Down

      I’ve ran into a few problems that use different formulas to compute the z score. 

      Some solutions show z = (observation – population mean)/ standard deviation

      Some solutions show z = (observation – population mean)/(Std error) 
      where std error = (std deviation/Sqrt(population size))

      the following question used the second equation to compute the z score while I used the first equation [(z = observation – population mean)/ std deviation]

       How would I know which one to use? they gave significantly different answers.

      Please refer to question B. I am fine with question A. 

      Ex: Assume that the equity risk premium is normally distributed with a population mean of 6% and a population std. deviation of 18%. Over the last four years, equity returns (relative to the risk free rate) have averaged -2.0%. You have a large client who is very upset and claims that results this poor should never occur. Evaluate your client’s concerns. 
      A. Construct a 95% confidence interval around the population mean for a sample of four year returns. 
      B. What is the probability of a -2.0 percent or lower average return over a four year period? 

    • googs1484
      Participant
      Up
      2
      Down

      Your not testing the hypothesis of a SAMPLE here. 95% confidence interval is 1.96 plus/minus 6 percent. The probability of would be (-.02 – .06)/.18. Take that value and look it up on z table.

    • googs1484
      Participant
      Up
      4
      Down

      Make sure you get the probability for LEFT outter portion of tail. For example 2.5% on a z value of 1.96

    • googs1484
      Participant
      Up
      2
      Down

      Some solutions show z = (observation – population mean)/(Std error) 
      where std error = (std deviation/Sqrt(population size))

      Hint: use this one when you see a sample size in the question. Only use this hint if you’re really strapped on time though. Could throw a curveball obviously on the test. 

    • googs1484
      Participant
      Up
      0
      Down
      googs1484 said:
      Your not testing the hypothesis of a SAMPLE here. 95% confidence interval is 1.96 plus/minus 6 percent. The probability of would be (-.02 – .06)/.18. Take that value and look it up on z table.

      Oops sry this was wrong but I’m hoping you caught it. Forgot to multiply 1.96 by 18 percent 

Viewing 4 reply threads
  • You must be logged in to reply to this topic.