Products

Quantitative Management - Seeking alpha where it lies, controlling risk where it hides

In equity allocations, quantitative management holds a special place: it aims to generate alpha while tightly controlling the risk taken relative to the benchmark index.

Published on 19 November 2026

quant

Cyrille Collet,

Head of Quantitative Equity Management at CPRAM

Based on systematic processes applied to broad investment universes, it makes it possible to diversify sources of performance, adapt stock selection to market regimes and incorporate each investor’s specific constraints. It thus offers fully active yet rigorously framed management, with a cost structure that is generally lower than that of traditional conviction-based equity management.

Active management built on a very broad universe

Quantitative management relies on the systematic analysis of a very broad universe of stocks. At CPRAM, more than 8,000 stocks can be monitored and analyzed through a combination of selection factors whose weighting changes with market conditions.

This depth of analysis means that the search for outperformance does not rest on a few highly concentrated convictions. Portfolios generally hold between 100 and 200 stocks: the search for alpha therefore relies on a large number of selection decisions, rather than on a small number of positions that could, on their own, determine performance.

This diversification does not, however, rule out tight risk control. Depending on the strategy, tracking error generally ranges between 0.5% and 3%. The objective is to preserve the active nature of the management while remaining close to the reference universe. This makes it a particularly well-suited tool for core allocations: an exposure to European, U.S., global or emerging market equities can remain close to its benchmark while benefiting from active portfolio management.

Managing unintended biases

Quantitative management relies on the systematic analysis of a very broad universe of stocks. At CPRAM, more than 8,000 stocks can be monitored and analyzed through a combination of selection factors whose weighting changes with market conditions.

This depth of analysis means that the search for outperformance does not rest on a few highly concentrated convictions. Portfolios generally hold between 100 and 200 stocks: the search for alpha therefore relies on a large number of selection decisions, rather than on a small number of positions that could, on their own, determine performance.

This diversification does not, however, rule out tight risk control. Depending on the strategy, tracking error generally ranges between 0.5% and 3%. The objective is to preserve the active nature of the management while remaining close to the reference universe. This makes it a particularly well-suited tool for core allocations: an exposure to European, U.S., global or emerging market equities can remain close to its benchmark while benefiting from active portfolio management.

Adapting selection to market regimes

That said, the discipline inherent in quantitative management does not mean that the same criteria must be applied unchanged in every market environment. Factors are not rewarded in the same way across the phases of the cycle: some may contribute more to performance in a supportive environment, while more defensive characteristics become more relevant when tensions rise.

At CPRAM, the process relies on a regime signal calculated daily, combining trend indicators and machine learning tools. It distinguishes between “stressed” and “non-stressed” market phases in order to adjust the combination of factors used in stock selection.

In a non-stressed regime, the multi-factor blend approach combines, among others, valuation, quality, growth and momentum. When the market enters a more stressed phase, the model gives greater weight to criteria such as low volatility, yield, financial strength or earnings momentum.

This adaptation does not mean constantly changing the portfolio’s exposure in line with tactical views: the process remains focused on stock selection. What changes with the market regime are the criteria used to decide between stocks. 

Better risk readability

One of the benefits of this approach also lies in the clarity it brings to portfolio risks. Because the construction rules are formalized, it becomes possible to identify more precisely the sources of differences in behavior relative to the benchmark index, and to distinguish what stems from stock selection from what results from other exposures.

This clarity is particularly useful for institutional investors and fund selectors, who need to understand how a strategy fits into an overall allocation. It also comes into its own during more difficult periods: an underperformance can be analyzed to determine whether it stems from a factor, from stock selection or from a residual market exposure.

Thus, while the risk of underperformance remains, the benefit of the quantitative approach is that its origin can be identified more precisely, and that one can check that the portfolio stays within the risk budget set at the outset.

Incorporating each investor’s specific constraints

Finally, this same discipline can be applied to very different sets of specifications. Sector or geographic exclusions, regulatory requirements, ESG criteria, ethical or Sharia-compliant universes, concentration limits or a specific benchmark index can be incorporated directly into portfolio construction.

These constraints are never neutral, however. By reducing the investable universe, they can reduce diversification, create new concentrations or significantly alter certain exposures. It is therefore not simply a matter of adding an extra exclusion or limit, but of measuring its consequences on the overall balance of the portfolio.

The same selection engine can thus be applied to an MSCI World (developed markets) or All Country (developed and emerging markets) universe, as well as to a universe defined according to ESG or Sharia criteria, or to an investor’s own exclusions. The resulting portfolios may differ significantly, precisely because the starting constraints are not the same.

The issue is particularly important in certain emerging markets, where regulatory, geographic or concentration restrictions can sharply reduce the number of accessible stocks. In this case, weighting rules, concentration caps or geographic limits make it possible to organize the exposure consistently, while taking into account the risks arising from a narrower universe.

Ultimately, the value of the quantitative approach lies in this combination of discipline, adaptability and risk control. Models feed into the management process: they provide a framework for broadening the potential sources of alpha, adapting selection as the environment changes, incorporating each investor’s specific constraints and, above all, maintaining a precise understanding of the risks actually taken.