Alternative Investment Portfolio construction methods — Risk Budgeting using a Data Science based Risk Analytics platform
Portolio Management comprises of at a minimum the following steps:
- Put together a Policy Statement that documents the investor’s objectives, goals and constraints
- Pick assets that can be considered for inclusion in the target portfolio. Below story covers this step where one can analyze risk/return characteristics, correlations with macro-economic indicators etc… for mutual funds and stocks in India on a data-driven risk analytics platform.
3. Come up with asset allocation strategies that determine how much weight in the overall portfolio should go towards each asset class. Strategic Asset allocation (SAA) is a more static long-term target asset allocation mix that is appropriate for a specific investor based on their policy statement goals, objectives and constraints. Tactical Asset Allocation (TAA) introduces small biases to the SAA mix based on the asset manager’s insights and experience of current market and economic conditions and how best to position a portfolio to benefit from such short-term conditions. Mean-Variance portfolio optimization is a proven quantitative optimization approach that has been widely used across the industry for decades now. It helps by coming up with a target portfolio allocation mix based on the desired risk appetite of an investor. Below story covers a DIY (Do it Yourself) process that a retail investor can undertake to construct their desired target portfolios.
4. Periodic monitoring and rebalancing. From time to time, the investor’s portfolio performance must be reviewed and effort must be made to validate its suitability to the investor’s goals and constraints. Necessary changes must be made based on the outcome of this process.
In this story, let’s focus more on step 3 using Vesp Analytics. Specifically, are there other approaches to portfolio construction that may be appropriate for certain “edge cases”?
As a continuation of the above two referenced stories, let’s review the example portfolio “Diversified Portfolio” more in detail.
On the “Efficient Frontier” optimized portfolios curve, if one were to select points on the top right side (with high risk and return properties), the suggested portfolio mix will be heavily tilted towards the most riskiest asset in the portfolio.
As we can see, the resulting portfolio is pretty much all of “Maruti Suzuki” stock and everything else drops close to zero. In the end, this is no better than buying just the stock with no diversification at all. To be fair, Mean-Variance portfolio optimization does what it is told to do, it optimizes for the combination that gets us the most return for the selected risk appetite and there are no other constraints. Since we chose the riskiest portfolio, it ended up tilting heavily towards the component that has the highest Risk/Return value (we can confirm that by clicking on the top right most dot on the “Risk Return Distribution” chart).
But, is there an alternative approach to portfolio construction that gives us a more fine grained control over how much maximum risk each portfolio component can contribute, rather than just selecting an aggregate portfolio level maximum risk. Risk Budgeting is one such approach that allows us to specify additional optimization constraints where a maximum risk contribution weight for each of the portfolio component can be specified upfront. Thereby, it controls for excessive leverage and risk concentration in the constructed portfolios. Risk Budgeting is one of the sophisticated portfolio construction approaches used by some of the top hedge funds. It has gained in prominence after 2007–2008 global financial crisis.
Now, let’s walk-through an example portfolio construction process using Risk Budgeting.
- Follow first half of this story till the point where we can add stocks and/or mutual funds to a portfolio and perform its initial correlation analysis.
- Select “Risk Budgeting” option as the portfolio optimization method and click on “Calculate Optimal Portfolio Allocations”.
3. On the “Optimized Portfolios” chart, click on any of the points (each represents a different portfolio mix) to see its corresponding risk budgets (each component gets a contribution budget, represented in %. All of them add up to 1). In the same table, we can see the suggested allocation mix to achieve the desired risk budgets. Feel free to try selecting each of the points and review their risk budget mix and identify the point with the desired risk budget mix. Suggested portfolio allocation mix is available in the next column.
4. Rest of the portfolio risk analytics for the selected point can be reviewed in the other charts
Feel free to try out different asset combinations, construct portfolios using both mean-variance and risk-budgeting approaches and analyze their risk/return characteristics to get a sense of how they map to one’s financial goals, objectives and risk tolerance levels.
Vesp Analytics provides a platform not just to perform an in-depth analysis of individual stocks and mutual funds, but also gives the ability for regular retail investors to construct portfolios using modern sophisticated approaches that were not easily available until now. More importantly, its an ideal platform to build upon by adding:
- New ways to analyze individual funds or stocks
- Add more sophisticated portfolio construction methods like sector rotation, Recurrent Neural Network based macro-economic prediction models etc…
- Investor portfolio construction, on-going tracking and periodic re-balancing