WEALTHWYSE

“I have X dollars, how should I invest them in the listed markets?”

Over the years people have tried numerous approaches for finding an all-weather solution to this question. To be able to fully discuss this we need to slightly rephrase the question – “I want to structure the portfolio to achieve the highest risk adjusted return, which means I want to maximize my returns, while limiting my risk to a level which I am comfortable with”. While the concept of return needs no definition, we must define what we mean by risk. Risk, specifically should mean the potential drawdown, but the portfolio volatility is a good approximation. So now we can refine the question further – “We want to maximize returns for a given level of volatility”.  

The traditional approach

Traditionally, financial advisors suggest an allocation between equity and debt, e.g. 60/40 and use mutual funds to achieve the required exposure (may use a questionnaire to assess the investor’s risk appetite) and typically suggest higher equity allocation for young investors and lower for older one. Once allocation is finalized, they deploy the funds in a number of equity/debt/mixed mutual funds, with a bias towards large caps (deemed less risky). And that’s where it ends. From time to time, based on their hunches and based on the flavor of the month, they may suggest replacing one mutual fund with another, but there is really no method to this. For most investors in India, this is how "investing in the market" goes.  Don't get me wrong, for many investors, this works and if you are a busy soul with little time to attend to market gyrations, this is a good way. We, however, want some thing that's more logical and depends less on subjectivity. 

Markowitz optimal portfolio

The foundation stone of pretty much all investing today, the mean variance optimization approach earned Harry Markovitz a Nobel Prize in 1990. He showed that one can find a portfolio of uncorrelated assets which can generate the highest returns for each level of risk. With a given risk free rate, his model derives a “Market Portfolio”, the optimum combination of risky and risk free assets.  After he published his findings in 1952, the concept of market portfolio has been wildly adopted by money managers all over the world. Refined in numerous ways by asset managers all over the world, the central concept remains the same – maximize return per unit of risk.

Endowment Model

Popularized by David Swensen, the revered manager of Yale’s endowment fund, this approach applies the mean variance approach across several “non-traditional” asset classes such as international equities, real estate etc. A model similar to the endowment model is the All Weather Fund Model which allocates equal amount of capital across several non-correlated assets and uses periodic balancing to ensure the allocation stays the same across asset classes (so if one asset class went up and another went down, the portfolio manager will sell some of asset 1 and buy more of asset 2 to ensure equal allocation).

Risk Parity

Probably the newest entrant in the portfolio construction methods, first proposed by Ray Dalio who currently runs Bridgewater Associates, world’s largest hedge fund. This approach “equalizes risk” (hence the name risk parity), which means at first it allocates capital to asset classes in such a way that each asset will contribute same risk to the portfolio. After this, it then uses leverage to enhance the returns of the overall portfolio.

In traditional models, the only way to achieve higher returns is by increasing exposure to risky assets, so to get equity like returns, one needs a portfolio which has higher exposure to equities, which also increases risk.  Risk Parity finds a novel way to achieve higher returns without proportionate exposure to risky assets. To give an example, say S&P has an annual volatility of 20% and treasury has an annualized volatility of 5%. This means equities are 4 times as risky as bonds. To allocate $100 across these two asset classes such that each contributes same risk to the portfolio will mean allocating 20% to S&P and 80% to bonds. Now let’s say S&P is expected to generate 5% returns per annum and bonds are expected to generate 2% returns per annum. This means the portfolio with 20% S&P and 80% bond has an expected return of 1% + 1.6% = 2.6%, and volatility of 4% (simplistic, this actually depends on the correlation between S&P and Bonds). At this stage, one can choose to use leverage and increase total exposure to increase the returns, without increasing undue exposure to the risky asset classes i.e. S&P. In the last few years, Risk Parity as a tool has become hugely popular and is a hot favorite of quantitative funds.

Once an allocation model has been selected, the next step is to find methods which can be used to identify securities for each component of the portfolio. Let’s say, you have decided to allocate 70% of the capital in equities, how should one go about it? That’s what we will discuss in our next post, using the framework of Alpha and Beta of portfolios.