A new approach to portfolio construction

At Risk Based Investment Solutions Ltd (RBIS), a wholly-owned subsidiary of the Rothschild Group, there is a firm belief that smart beta has a promising future. In fact, Herve Foucault (pictured), Head of Business and Product Development, goes so far as to say: “This is a real paradigm shift from the traditional cap-weighted model.”

For decades, the approach to traditional portfolio construction has been anchored in using market cap-weighted models. But in Foucault’s opinion, there is a growing recognition among academics and asset managers that this approach is too disconnected from reality.
 
“Historical data demonstrates that a cap-weighted approach for equities is not the most efficient way to achieve an optimal risk/return profile for portfolio construction. There is room in the market to explore alternatives to traditional portfolio construction, such as smart beta. It started with minimum volatility and minimum correlation, and now we have equally weighted risk contribution models,” says Foucault.
 
At RBIS, they are pushing the next evolution of portfolio construction under the ‘smart beta’ moniker by using a multi-factor approach, which Foucault refers to as Equally-weighted Risk Contribution (ERC).
 
Think about a standard portfolio of 100 stocks tracking the FTSE 100 index. There are a number of different ways to weight the portfolio: one could allocate 100 per cent to the single largest stock, 100 per cent to the single smallest stock, and all other options in between including the cap-weighted one.
 
The ultimate stock selection, using this defined approach, would be to have a future view on which stocks are the best performers. This would create the most possible performing portfolio on a day-to-day basis. But nobody has that view into the future. The best one can hope for is to destroy as minimal an amount of value as possible compared to this hypothetical portfolio. At RBIS, rather than focusing on portfolio weighting – and subsequent risk output – they first focus on the risk parameters.
 
“We have moved from a ‘weight-to-risk’ process to a ‘risk-to-weight’ process. Once yoy have more than one component in the portfolio you have what we call ‘hidden assets’; these are basically the correlation between each stock and their individual volatility,” says Foucault.
 
As the name suggests, it is an asset and an asset has a value. “We focus on these hidden assets of the portfolio. For the 100 stocks in the portfolio, there are 100 different volatilities, which we measure. Then we use a 100 x 100 correlation matrix to measure inter-correlation between stocks; this is the core of our portfolio construction. Taking into account the “hidden assets” in the portfolio construction process is the most efficient way to extract value from the hidden assets.”
 
Volatility and correlation are the two primary risk factors that RBIS focuses on in order to then create an equally-weighted risk contribution portfolio where each line in the portfolio has the same risk budget.
 
“ERC is the next stage in the smart beta approach. We are transforming the hidden assets into real-life assets. We’ve used this approach to backtest simulations on every global market,” says Foucault.
 
The approach is systematic. Once the index components are defined the formula is applied on a monthly basis.
 
So far, the methodology is yielding attractive results. For a typical equity portfolio, volatility in an ERC portfolio is reduced by 20 per cent, the Sharpe ratio increases by a third and maximum drawdowns are reduced by 25 per cent. 

Author Profile