Concentrated funds with less than 50 holdings may not be adequately assessed for risks by the current approach to risk modelling. The volume of concentrated funds has more than tripled to $600 billion over the past decade but risk modelling tools have not kept pace with the risks involved.

Laura Dew argues for better risk modelling of concentrated strategies. Her report was published by Money Management:

According to Franklin Templeton Global Growth portfolio manager, Don Huber, the volume of concentrated funds were under-served by risk modelling tools, which analysed volatility, tracking error and permanent loss of capital. This meant many concentrated funds were flagged by risk modelling tools as risky due to ‘unexplained’ or ‘stock specific’ risks within the portfolio.

“For concentrated, unconstrained global equity managers who tend to run portfolios of less than 50 stocks across the market cap spectrum and typically have higher active share, the risk ‘information’ for these models is often of limited use.”

A second problem was the use of sector or geographic restrictions, which Huber said gave investors a ‘false sense of comfort’ that they were managing risk. This was because, while companies may sit in the same sector or region, they were often very different to each other and carried different risks.

“While such limits may aid client comfort levels and prompt associations with strong risk management, this approach presumes there is a high level of commonality and correlation between securities in a country or sector.

“The assumption of high levels of correlation become particularly problematic with concentrated portfolios and is further challenged when companies are not mega-cap companies that can act as benchmark proxies for capitalisation-weighted indices.

“In a more concentrated portfolio, with fewer holdings in any sector or country, those company-specific characteristics are less likely to be meaningfully ‘averaged away’, driving lower correlation between a smaller group of holdings and benchmark returns.”

Instead of using these traditional methods, Huber suggested an approach of fundamentals-driven risk management which would look at the individual fundamentals, taking a more granular approach than relying on constraints.