There are four sections in our report body. After the general conception understanding and discussion about requirement, the first problem we confront is the data collection as we explain in Part 1 of our report. We finally achieve the agreement to use 4 years monthly data via a lot of article reading and group meetings. The second part in our report is about the more accurate method for return calculation Continuous Compound formula which refers to “Research design issues in the estimation of Beta” (Bradford, Tim, 1997).
In the third section we present the detail of calculation and outcomes following the two approaches. It insists of tables and graphs with formula and some comments. The last section of our report is the analysis of two approaches’ assumption and the outcomes, ending with our conclusion as the recommendation. The reference, appendixes and coversheet are attached in the end as well. 1.
Data Explanation 1) Stocks we choose: ANZA and AX 2) Sample interval: Our raw-data is based on month both in Mean-variance and CAMP: Against Yearly data, quarterly data and Daily data, monthly data is preferred: Yearly data was not accurate to analysis the return of a share price for the share price was announced regency, the standard approach is to match macro data with monthly or quarterly aggregates of financial series to build prediction models. (Should macroeconomic forecasters use daily financial data and how? By Elena Androgen, Eric Eggshells and Android Courteous).
However, data of intra-day intervals would result in unstable and unreliable estimate of beta, while using of quarterly data requires us to cover price histories of at least ten years to yield adequate data points (Bradford, Tim, 1997). Consequently, we’d prefer to use the monthly data. 3) Sample period: We choose four years data which from Mar 31, 2006 to Mar 31, 2010 in both of Mean Variance approach and CAMP approach. To get reliable beta, around 50 data points are required to be derived from 4 to 5 years’ time, in which beta estimate appears reasonable stable(Bradford, Tim,1997).
The time period we chose is a large enough sample to ensure us get a reliable statistical estimate, and because we will use the estimate beta to study the risk and return of current stock portfolio, the recent data is more preferred because it has a larger relevance(Bradford, Tim,1997). 4) Sample resources (refer to Appendix) ANZA stock prices & dividends prices & dividends Mornings Admiralty’s AX stock Mornings Admiralty’s S/ASX 200 Accumulation Index Reserve Bank of Australia Capital Market Yields Reserve Bank of Australia 2. Return calculation formula (refer to Appendix. ) In both Mean-variance approach and CAMP approach, refer to the monograph (Bradford, Tim, 1997), we calculate return by continuous compound formula. Art – stock/Market return in period t, Apt stock price/Market Index at the end of measurement interval t, and EDT- stock/market dividend on period t (ASSESS Scum. Index including dividends) Continuous compound can reduce the effect of outlier data or the data error because the continuous compound data series is more likely to follow a normal distribution which is the assumption of both Mean-variance and CAMP.