代做Project 1: Long-Only Multi-Asset Portfolio with Tilts代写留学生Matlab语言程序

Project 1: Long-Only Multi-Asset Portfolio with Tilts

Your team has to submit a report of 3 to 5 pages documenting the project and an excel spreadsheet containing key results and formulas, following the guidelines below.

This project has to do with creating and back-testing a long-only multi-asset portfolio with tilts. Your investment universe should comprise a diversified set of assets including US and international equity indexes,  US and international bond indexes, etc. You can use the returns and market values of global assets, posted on Courseworks, or download your own. At the minimum, you have to establish 2 tilts. Your backtest should    be at least 10 years long.

Your goals are as follows:

1. Each year, produce the expected returns for assets in your investment universe. Apply the Black-

Litterman method to blend the market (or a benchmark) with views.  You have to introduce at least two view portfolios, which can be as simple as a positive weight for at least one asset, such as US Small Cap Value equities, and a negative weight for another asset, such as US Small Cap Growth equities for one view portfolio and a positive weight for commodities and negative weights for the rest of the assets (market value-weighted) for the second view portfolio, for example. The view portfolios can be the same in each year of the backtest. Use the lecture notes and the He-Litterman    paper for guidance as to the choice of any parameters and make the level of confidence in your views sufficiently weak. The deviations from the market of your final weights should be with the -5%

and +5% range.

2.    In your spreadsheet, for the final portfolio only, provide:

a.    The market capitalization weights and the view portfolio weights [tab name: “Weights”]

b.   Your chosen risk aversion and the measures of your confidence in the market-implied expected returns and your views [tab name: “Parameters”]

c.    The market-implied expected returns, and the Black-Litterman blended expected returns [tab name: “Expected Returns”]

3.    Estimate your covariance matrix and re-estimate it every year by using at least 36 months of prior

returns in each estimation. In your excel spreadsheet, provide the annualized covariance matrix for the last year of your backtest in a tab labeled “Covariances” . In addition, provide volatility of each asset and provide the correlation matrix of the assets, extracting them from the covariance matrix that you used for the last year of your backtest.

4.    Produce optimized long-only portfolio for each year in your backtest.  Use mean-variance

optimization with the same risk aversion that you used to derive the expected returns, the covariance matrices computed in item 3 and the Black-Litterman expected returns computed in item 1. Impose  the following constraints in the optimization, if needed:

a.    The portfolio is long only (no negative weights).

b.   Portfolio weights add to 1.

5.    In your spreadsheet, in a tab named “Backtest weights”, present a chart of the market capitalization weights and the optimal portfolio weights, for every year of your backtest.

6.    Backtest your portfolio and produce the following set of analytics for the full back-test period. Formulas must be given in your spreadsheet for at least the last month:

a.    Cumulative payoffs (growth of $1) of your portfolio and the market, also presented as a

chart [tab name: “Payoffs”]. To calculate monthly returns on the market, you can multiply the market capitalization weights from the prior month (calculated form the provided data or downloaded by yourselves) by the respective asset index returns and sum them up for all assets.

b.   Annualized volatility and geometrically annualized return of your portfolio and the market [tab name: “Total Stats”]

c.    Annualized active risk (tracking error) and arithmetically annualized active return of your portfolio, based on the difference each month between your portfolio’s return and the market return. From these statistics compute the information ratio of your portfolio [tab name: “Active Stats”]

Please avoid a look-ahead bias: for each year in the backtest you can use only the data that existed prior to that year. The purpose of the backtest is to recreate real-life conditions and assess how well your strategy would have performed by applying it to historical data. In real life, obtaining data from the future is impossible.

Please describe your work. Your report may contain some of the following points:

.    Brief description of the data sources, if you used your own

.    Description of the investment universe, if you used your own

.    Brief discussion of the tilts that you established

.    Black-Litterman: brief description of parameters you used

Project grade will not depend on the active returns that your portfolio produced.