Using multi-factor allocation in portfolio construction

Post with image

MFA – The new model for more intuitive, reliable and understandable results

  • For a more optimal allocation of core assets, we are introducing factor analysis to map our investment views to multi-asset exposures.
  • Our new multi-factor allocation (MFA) model is less sensitive to changes in inputs compared to traditional optimisers. Its results are more intuitive, reliable and understandable.
  • This approach can be run in a bespoke manner, taking into account client specifications and their wishes as regards portfolio outcomes [1][2].

Based on the latest academic insights, MFA provides us with an efficient and precise tool to translate our views on the drivers and factors affecting the core asset classes into multi-asset portfolios.

In one single step, it takes into account our active views on the core asset classes, correlations, risks and any clients’ constraints, for example, on the acceptable tracking error and the exposure to a particular asset class.

Its objective is to maximise the portfolio’s risk/return and its outcome-based targets, while linking the size of positions in the portfolio to the strength of the conviction scores on the core asset classes.

Exhibit 1: The principles underlying our proprietary multi-factor allocation (MFA) model

Exhibit 1: The principles underlying our proprietary MFA model

Source: BNPP AM, as of 11/10/2019

Multi-factor allocation works with an uncertainty term on expected returns. Accordingly, it takes into account alternative outcomes where reality deviates from the initial forecasts – i.e. when it is more negative or more positive than our investment view. We believe this helps produce robust portfolios.

The MFA algorithm finds the optimal allocation combining our multi-asset investment views, portfolio structure and the client’s risk profile.

We use six factors in our model.

  • Market risk – equity and equity-like (e.g. real estate) assets as well as the riskier parts of the fixed income universe (high-yield credit, emerging market debt) have a large exposure to this factor.
  • Duration – government bonds and the safer parts of the fixed income universe (investment-grade debt) are exposed to this factor. Equities have a negative exposure to duration.
  • Emerging markets & commodities – this factor mainly shows up with assets such as EM debt and equites.
  • Corporate spreads – this can be seen in credit markets; other major asset classes have a slightly negative exposure.
  • For the US factor, broad major asset classes show little exposure, but US equities and bonds are more exposed to it.
  • Similar to EM & commodities, the Asia & China factor mainly shows up with assets such as EM debt and equites.

Exhibit 2 shows the factor exposures of the main asset classes. What is evident is the vast difference in factor loadings from asset class to asset class.

Exhibit 2: Factor exposures of main asset classes

Exhibit 2: Factor exposures of main asset classes

Source: BNPP AM, as of 11/10/2019

So how does MFA work in practice?

Let’s use high-yield credit as an example. It sits halfway between equities and bonds in a company’s capital structure, so multiple factors are at work. European high-yield credit is exposed to corporate spreads and market risk. If you were to overweight European HY in portfolios, they would be impacted along these factor exposures. To be complete, equities would be primarily driven by market risk, while bonds would be mainly driven by the duration factor.

Example 1: Factor exposures of EUR high-yield credit

Example 1: Factor exposures of EUR high-yield

Source: BNPP AM, as of 10/10/2019

MFA’s most powerful feature is the simultaneous mapping of core views across asset classes onto a single set of factor exposures, while taking into account the constraints for each portfolio.

The optimiser will try to achieve the same factor exposures for every portfolio, using the portfolio’s tradeable assets. Of course, in practice, achieving identical factor exposures with only a limited set of assets is impossible, but as the example shows, factor exposures in real-life portfolios will largely mimic unconstrained exposures.

Example 2: Unconstrained vs. a Europe-centric multi-asset portfolio – comparing the factor exposures of our current views

Example 2: Unconstrained vs. a Europe-centric multi-asset portfolio – comparing the factor exposures of our current views

Source: BNPP AM, as of 11/10/2019


[1] This is a shortened version of the paper entitled INTRODUCING A PROPRIETARY PORTFOLIO CONSTRUCTION SYSTEM. For the full version, click here.

[2] Watch the video on our new multi-factor allocation here


For more articles by Maximilian Moldaschl, click here >

To discover our funds and select the ones that meet your requirements, click here >


The views expressed are those of the Investment Committee of MAQS, as of October 2019. Individual portfolio management teams outside of MAQS may hold different views and may make different investment decisions for different clients.

The value of investments and the income they generate may go down as well as up and it is possible that investors will not recover their initial outlay.

Investing in emerging markets, or specialised or restricted sectors is likely to be subject to a higher than average volatility due to a high degree of concentration, greater uncertainty because less information is available, there is less liquidity, or due to greater sensitivity to changes in market conditions (social, political and economic conditions).

Some emerging markets offer less security than the majority of international developed markets. For this reason, services for portfolio transactions, liquidation and conservation on behalf of funds invested in emerging markets may carry greater risk.


Maximilian Moldaschl

Senior Global Multi-Asset Strategist

Leave a reply

Your email adress will not be published. Required fields are marked*