DeMiguel, Garlappi and Uppal (2007) find that the naive diversification or the 1/N rule is generally no worse than many more sophisticated strategies suggesting that “there are still many miles to go before the gains promised by optimal portfolio choice can actually be realized out of sample."

The objective of this assignment is to understand why optimal portfolios are not well performing particularly when compared against the equally weighted portfolio and learn how academics and practitioners have been addressing this issue. By replicating some of the portfolio models, students will have an opportunity to practise Matlab and understand potential issues that might arise during implementation.

Students are encouraged to use Matlab for empirical analysis, but other programs are also allowed. For example, possible alternatives to Matlab are Scilab and Octave, which are freely distributed and very similar to Matlab.

The report must follow the structure:

1.   Introduction (5%)

Give a brief introduction and overview of your report.

2.   Literature Review (40%)

Choose three papers that are related to the subject; that is, papers comparing portfolio models, and provide a detailed summary of each paper. The summary may include the portfolio models, datasets, evaluation method, and main findings of each paper. Please use Harvard style to cite references. Do not copy and paste long sentences from references. You can re-write and/or paraphrase the important points from the  papers  in your own words. The  report should  have an easy structure to follow. Critically evaluate each paper. A focused review on a small number of papers is preferred to a brief review of many papers.

3.   Empirical Analysis (40%)

Choose one paper and replicate it using a different dataset. The papers must be published in high ranked journals including the ABS 3* journals and above. There will be also a sample of papers available for you as a guide. The dataset will be uploaded in Ultra in due course. You can also use other relevant datasets from reliable sources. If the paper contains several models, choose one for replication.

o Provide a description of the model with implementation details.

o Evaluate the model using the dataset. Evaluation should include comparison with the equally-weight portfolio. Discuss your findings: compare them with the results presented in the original paper and critically evaluate them.

4.   Conclusion (15%)

Reflect on your findings and provide your view on the effectiveness of optimal portfolio models. Are we better off with the equally-weight portfolio?

Overall word limit: 2500 WORDS MAXIMUM

SUBMISSION INSTRUCTIONS

Assignments should be typed, using 1.5 spacing and an easy-to-read 12-point font. Assignments and dissertations/business projects must not exceed the word count indicated in the module handbook/assessment brief.

The word count should:

  Include all the text, including title, preface, introduction, in-text citations, quotations,

footnotes, and any other items not specifically excluded below.

  Exclude diagrams, tables (including tables/lists of contents and figures), equations,

executive summary/abstract, acknowledgements, declaration, bibliography/list of references and appendices. However, it is not appropriate to use diagrams or tables merely as a way of circumventing the word limit. If a student uses a table or figure as a means of presenting his/her own words, then this is included in the word count.

Examiners will stop reading once the word limit has been reached, and work beyond this point will not be assessed. Checks of word counts will be carried out on submitted work, including any assignments or dissertations/business projects that appear to be clearly over-length. Checks may take place manually and/or with the aid of the word count  provided  via  an  electronic  submission.  Where  a  student  has  intentionally misrepresented their word count, the school may treat this as an offence under Section IV of the General Regulations of the University. Extreme cases may be viewed as dishonest practice under Section IV, 5 (a) (x) of the General Regulations.