Why conviction matters in factor investing

Key takeaways:

• Factor models can tell you what a portfolio looks like, but they rarely tell you what the manager believes will happen next.
• Conviction isn’t just a bigger overweight, because position sizes are also shaped by risk, correlations, diversification and constraints.
• Combine multiple active managers and the strongest insights can get washed out, leaving a portfolio that behaves closer to the benchmark than you expect.

Over the past two decades, investors have become very good at breaking markets into signals such as - value, quality, momentum or sentiment.

If a stock behaves a certain way, there is usually a factor that helps explain it.

That shift has reshaped how portfolios are built. Factor investing introduced a more systematic way to target drivers of return, bringing consistency, transparency and scalability to equity investing. It works by identifying shared characteristics across stocks and applying those signals in a disciplined, repeatable way.

In many ways, that has been a success story.

But in actively managed portfolios, there’s something else, something those signals don’t directly show.

More than a list of holdings

A factor looks at stocks. A portfolio reflects a set of decisions.

That distinction is easy to overlook, because portfolios are often presented as lists;  lists of holdings, weights and exposures. But what sits behind that list is more interesting.

Every portfolio is built through a series of choices: what to own, what to avoid, and how much to allocate to each position. Those choices are not just expressions of characteristics but they also reflect expectations about future returns, balanced against risk, diversification and constraints.

Two managers can hold the same stock for very different reasons. One may see it as a high-conviction opportunity. Another may hold it in a smaller size to manage risk or maintain balance. The name is the same. The intent is not.

In other words, a portfolio is not just a collection of exposures. A portfolio is a structured expression of belief.

That belief is what we mean by conviction.

Conviction shows up in how strongly a manager backs an idea. Not just whether a stock appears in a portfolio, but how capital is allocated across positions. If conviction is already embedded in portfolios, you might expect it to be easy to measure. But in practice, it isn’t.

A common shortcut is to treat active weights as a proxy for conviction. A larger overweight is assumed to signal stronger belief. A smaller position is assumed to signal weaker conviction.

That seems intuitive but it is incomplete.

Position size reflects more than just expected return. Position size is shaped by volatility, correlation with other holdings, diversification needs and portfolio constraints.

A 1% position in one stock does not mean the same thing as a 1% position in another. One may carry more risk. One may overlap with other holdings. One may be there to support the structure of the portfolio rather than express a strong view.

So, the signal we often rely on is not wrong but it may not tell the full story.

Where conviction starts to get diluted

This becomes more pronounced when investors allocate across multiple managers.

Each manager brings a set of ideas, expressed through their portfolio construction decisions. On their own, those portfolios reflect clear views and deliberate trade-offs.

But when they are brought together, something changes.

The underlying insights remain, but their intensity becomes harder to see. As a result, conviction can be diluted, unintended exposures can build up, and the overall portfolio can begin to resemble the benchmark more closely than expected.

This is a familiar frustration. A portfolio that looks active on paper can feel surprisingly average in practice. Strong ideas are present, but they are harder to distinguish. The connection between insight and outcome becomes less direct.

Over time, that has a practical impact. Portfolios can become efficient—but less expressive. The clarity of where returns are expected to come from starts to fade.

What factor investing does (and doesn’t) do

None of this is a criticism of factor investing.

Factor investing captures systematic characteristics of stocks in a consistent and scalable way. That is precisely why it has become such a central part of modern portfolio construction.

But factor investing operates on observable attributes.

Factor investing does not explicitly extract the expectations embedded in how portfolios are constructed.

This is where the idea of 'implied expectations' becomes useful. When a portfolio is built, the combination of positions, sizes and trade-offs reflects a set of return assumptions: what the manager believes each stock will deliver, after accounting for risk.

In more technical terms, these are often called 'implied alphas'. In simple language, they are the return expectations that must be true for the portfolio to make sense given how it is constructed.

That distinction matters.

Two stocks can share similar characteristics. From a factor perspective, they may look almost identical but a manager may have very different expectations for each.

One idea may be held with strong conviction, based on deep research and forward-looking insight, while another may play a supporting role, helping to manage risk or maintain balance within the portfolio.

Factors treat those signals consistently. Portfolios do not.

A different way of looking at portfolios

Once you see it, the perspective shifts.

Markets give you data and portfolios give you decisions.

Factors organise the data but decisions reveal something else, what a manager believes.

Those beliefs are already there, expressed through how risk is allocated across positions.

The question is whether those beliefs can be isolated and used more deliberately.

What comes next?

If conviction is embedded in portfolios but difficult to observe, it raises a more practical question.

What happens to portfolio outcomes when that conviction is not fully captured?

And if it can be extracted more clearly, how might that change the way portfolios are built?

There is a way to approach this problem.  One that starts by working backwards from the portfolio itself, bringing together the discipline of factor investing with the insight embedded in active portfolios.

In the next article, we explore how that works in practice and how conviction can be translated into a usable signal.

For a deeper dive into how this works in practice, you can download the Capturing Conviction white paper.