Novus Methodology: How We Calculate Alpha
The holy grail of our analyses is the Novus Framework, representing the next generation tools to help investors isolate and measure alpha.
The investment management industry is going through an irreversible transition. As investors start understanding the need to move from an absolute-performance model to a more holistic, alpha-based one, and as more financial research is developing in the academic and professional worlds, the theories on how to generate the best risk-adjusted returns are being re-designed.
Novus is a product-oriented portfolio intelligence company. As such, we deliver the most comprehensive insights in the market. The holy grail of our analyses and methodology is the Novus Framework. This framework represents the next generation in tools to help investors isolate and measure alpha.
What is alpha?
The Novus Framework is very straightforward. It decomposes the contribution of a specific position, or the overall portfolio, into 4 different factors:
Market and sector contribution together determine the relative contribution, or beta, of the portfolio, while security selection and trading contribution are considered the absolute contribution, or alpha.
Skill in Alpha
Each factor, besides the market, represents a different skill. If you are an allocator you would want to evaluate the skill set of the manager before committing to investing. Along the same line, if you are a manager you would want to evaluate yourself so as to improve future performance in the long term. This is indeed one of the added values of the Novus Framework: it not only quantifies alpha, but it also shows what skills a specific manager is good at. You can think of these skills in the following way:
For Novus clients, alpha generation is neither the risk-adjusted excess return on the risk free rate or benchmark, commonly known as Sharpe ratio, nor the Jensen index. Alpha is far more granular.
Novus Framework and Performance
In the following example, we are going to draw on the public data of Viking Global Investors, an equity long/short manager. For the period from June 2014 to June 2015, this is what the Novus Framework shows:
How can we interpret these results?
Can we calculate alpha at a security level?
Yes. After all, security selection alpha at the portfolio level is the alpha in every security added together
Let’s look at an example. Viking was invested in Allergan (and 148 other positions) during the selected period. The contribution of this equity stock to Viking’s public portfolio is summarized below:
Out of the 4.45% of security selection alpha, 0.45% came from Allegan alone. In other words, 14% of the Viking’s alpha was derived by a single name. In comparison, for the average Tiger Cub fund, the top winner contributed 37.5% to the total alpha generated over time.
What if a manager is interested in the TTM rolling alpha of Allergan in Viking’s portfolio? To get a more detailed view, it helps to look at the historical contribution of these factors.
During June 2014-June 2015, in addition to the overall outperformance of the health care sector, Allergan itself added increasing returns over time, beyond market returns.
Positive returns but negative alpha?
On the opposite side of the spectrum, Viking owned Thermo Fisher Scientific, also a healthcare name. During the same time range, while the market and sector contributed to the portfolio, the specific security actually detracted alpha. This would be an immediate red flag if it wasn’t because Profit/Loss was actually positive for the period. Sometimes, when there is positive returns, be it at the security or at the portfolio level, there is not necessarily positive alpha. Understanding this concept and how to derive it is what the Novus framework excels at.
The Novus Framework is the best way to derive alpha insights for a portfolio. Not all managers that enjoy positive returns generate alpha, and not all managers that have negative returns generate negative alpha. Skill should not solely be measured in absolute terms.