Step 1 Building the stock ranks

Over 1000 stocks are monitored tracking value, growth, estimate revisions, price trends, and quality amongst others. The output of the models is combined with market knowledge to build a portfolio with low risk characteristics and good return potential. Our processes steer us towards stocks with either superior long or inferior short characteristics, so that we can benefit from either upward or downward stock price movement.

Long positions are ideally stocks that are cheap, show good profitability & growth, are trending up, beat earnings estimates, earn cash instead of paper earnings, use their money & resources conservatively and show year-on-year improvement in margins, asset turns, dividends and buybacks.

  • Model enhancements
  • Several enhancements have been developed in-house that makes our model special compared to others.
    The enhancements include a Value/Growth Score, that allows us to move away from the categorical concept of “pure Value” or “pure Growth”, and consider each stock as a mix of the two styles instead. Additionally, co-movement between stocks, as well as the correlation between factors and stocks, are taken explicitly into account.
    Factor asymmetry is utilized by allowing factor distribution and threshold to vary. The model is adaptive, as emphasized by the management of the dynamic factor weights. Since factors are not linear and factor’s behavior is not constant over time, we determine a strategic set of factor weights that maximizes the long term predicting power of the model, conditioned on volatility and a minimum/maximum weight of individual factors and quadrants. More information about some of the enhancements is available for registered users in our Whitepaper on Quantitative Stock Selection.

As a result, every day 1000 European stocks get a rank from 0 to 100. Stocks with a rank from 80 to 100 are the long candidates, from 0 to 20 are the short candidates.

Step 2 From stock ranks to portfolio construction


Subsequent to alpha generation process by the multifactor model, we proceed to a portfolio optimizer that calculates risk characteristics for each stock in our investable universe. The optimizer constructs a portfolio that maximizes the expected alpha, subject to a custom level of volatility, gross and net exposure, and additional diversification and holdings constraints. Holding restrictions are implemented to deal with stocks that are experiencing litigation risk, political risk, impending earnings reports, and other idiosyncratic conditions that might be overlooked by a quantitative model.

This is done systematically using fundamental scores assigned by portfolio managers. Ultimately, the portfolio should represent alpha signals generated by the multifactor model while minimizing the unintentional risk.
The Fund can take both long and short positions. The Fund is run with low beta-adjusted net exposure, and will be predominantly market-neutral over time.

Step 3 Risk Management

Portfolio risk management is an integrated and central part of the investment process. The portfolio construction follows a strict framework with restrictions on VaR, beta, leverage and liquidity. Stock, sector, country and market cap exposures are tightly controlled. We implement holding restrictions for stocks that are exposed to M&A, litigation risk, or other conditions that might be overlooked by a quantitative model. 

Companies that represent high risk in terms of integrity of their data or management are eliminated from our investable universe. The Fund operates a stop-loss policy. We use various risk models to fully understand the nature of the portfolio exposures and manage them accordingly.