Factor investing has been a fixture of academic research for decades, but do the premia it promises actually hold up under the weight of real-world markets? In the second instalment of the Factor Academy series, Maarten Smit, CFA, Senior Portfolio Manager at Northern Trust Asset Management (NTAM) Quantitative Strategies, took advisers on a deep dive through 150 years of evidence, inflation regimes, economic shocks and implementation realities to answer that question with some conviction.
What is a factor?
For those joining the series for the first time, Maarten opened with a concise refresher. A factor is a specific characteristic or attribute of a security that helps explain its risk and return. The four most established factors are value, quality, momentum, and small size, and each is grounded in intuitive investment logic.
Identifying companies trading at a discount to their fair value, measured through earnings multiples, book value and other fundamental metrics. Think of it like a property investor assessing rental yield against purchase price.
Identifying companies with a sustainable competitive advantage, high profitability, low leverage and strong management. Return on equity is the classic starting metric, but quality is inherently multi-dimensional.
Identifying stocks with strong recent sentiment and price trends. One of the most pervasive anomalies in finance, documented across asset classes, regions and cap sizes. It challenges efficient market theory but the empirical evidence is hard to argue with.
Smaller companies tend to be less efficiently priced, covered by fewer analysts and carry a risk premium. The size effect is most powerful when combined with the other three factors rather than used in isolation.
Used together, these two factors are particularly complementary. Value can identify cheap stocks, but cheap stocks often have poor recent sentiment. Momentum filters out the ones that are cheap for a reason, surfacing those that have already turned a corner.
No single factor leads in every regime. Combining them in a diversified multi-factor approach smooths out the cyclicality that any individual factor will inevitably experience over time.
150 years of persistence
The headline claim of the session was a bold one: factor premia have persisted across 150 years of global market history. That dataset, constructed in collaboration with Erasmus University Rotterdam and led by NTAM's global head of quantitative strategies Guido Baltussen, spans the period from 1875 to 2024 and includes two world wars, stagflation, the Great Depression, multiple inflation shocks and the entire age of globalisation.
Crucially, this extended dataset goes well beyond the data available in most academic research, which tends to begin in the 1960s with the CRSP database. Going back further allows NTAM to test whether factor behaviour is a product of the modern financial era or something more structural and enduring.
"Factors are not a tactical trade. They have not been here only for the last decade. They are essentially an eternal feature of financial markets."
Maarten Smit, NTAM Quantitative Strategies
Across the full 150-year sample, each of the core factors delivered positive excess returns above equities, with a diversified multi-factor combination averaging around 2% premium over the market. Momentum showed the strongest raw premium across the long sample, while low volatility offered meaningful risk-adjusted outperformance given its lower market beta. Crucially, value and quality also delivered across both the 1875-2024 window and the more recent 1995-2024 period, countering the narrative that the value premium is dead.
Factors across every inflation regime
One of the most practically useful elements of the session was how Maarten sliced the 150-year dataset by inflation environment. The data was divided into four regimes: deflation (below 0%), moderate inflation (0-2%), mild overshoot (2-4%) and high inflation (above 4%). Across all four, every factor outperformed the market on average.
This is a meaningful finding for advisers currently navigating a period where inflation has moderated but remains above central bank targets. The data shows that where we are today, in a mild overshoot regime, has historically been a supportive environment for momentum in particular, with quality and the multi-factor combination also performing well above the market.
Low volatility led the way (9.1% real return), followed by momentum (8.5%) and the multi-factor combination (8.0%). The market returned 5.5%. Defensive quality characteristics proved most durable when prices were falling.
The most historically common regime and the most broadly supportive. Quality, momentum and the multi-factor approach all delivered real returns of 12-13%, versus a market return of 9.8%.
The regime most relevant to today's environment. Momentum led (12.2%) followed by low volatility (11.2%) and quality (10.6%), with the market returning 8.5%. All factors outperformed.
The most challenging regime for equities, with the market returning -1.8% in real terms. Yet quality (3.3%), momentum (1.9%) and low volatility (1.3%) all stayed positive. The multi-factor combination returned 0.6% versus a negative market. Size was the only factor to dip below zero.
"In the most feared scenario of stagflation, quality, low volatility and momentum are again helping to cushion the loss. A multi-factor combination shows less severe losses than the market in that environment."
Maarten Smit, NTAM Quantitative Strategies
Maarten also decomposed the high inflation bucket into rising and falling inflation sub-scenarios, and further into stagflationary and non-stagflationary environments. In genuine stagflation, equities returned -20% per year on average. Quality and low volatility still outperformed. In high inflation with sustained growth, by contrast, all factors delivered positive real returns and broad-based outperformance.
From academic definition to real-world implementation
A recurring theme in adviser conversations around factor investing is the gap between what the academic literature shows and what a real investor can actually capture. Maarten addressed this directly, outlining the six criteria NTAM applies when evaluating any factor for inclusion in live client portfolios.
A defensible mechanism must exist for why the signal earns a return, whether behavioural, structural or informational. Theory comes first.
Significant, persistent long-short returns across time horizons, market universes and sub-samples. Not just a statistical curiosity in one narrow period.
Uncompensated risks must be diagnosed and neutralised. For momentum, this means implementing in a beta-neutral, risk-scaled manner to avoid the market timing exposures that can cause academic momentum to crash.
A new signal must add genuine explanatory power beyond factors already in the portfolio. Repackaged signals that are simply a linear combination of existing ones do not qualify.
Works across regions, cap segments and market conditions. Not reliant on a single environment or a narrow data window. Macro robustness, across rate and inflation regimes, matters too.
A factor must survive realistic transaction costs, long-only constraints and capacity considerations at scale. If the premium only exists before costs, it does not make the cut.
On the question of implementation gap, Maarten was candid: for slow-moving factors like value and quality, investors can capture a very significant share of the theoretical premium. For highly transient signals, such as post-earnings announcement drift, which may exist for only a few days before evaporating, NTAM simply does not include them in live portfolios. The framework is designed to ensure only factors that survive all six tests reach client allocations.
Factor definitions have also evolved considerably since the original academic papers. NTAM now combines financial statements and market data with alternative data, network effects, analyst reports and AI-driven natural language processing to answer the same underlying questions through a richer set of lenses. The question of what makes a quality company has not changed. The tools available to answer it have.
Questions from the session
Will AI erode factor premiums? Maarten's view is that, so far, it has not and that AI operating in a standalone capacity is unlikely to degrade returns. Hedge funds using AI exclusively have not yet outperformed those combining human judgement with AI-assisted tools. The way forward, in NTAM's assessment, is humans using AI to generate and refine investment strategies, which is already embedded in their process.
How do factors behave in market drawdowns? It depends on the factor. Quality and low volatility tend to be defensive: they may lag in strong rising markets but offer meaningful protection when markets fall. Value and momentum are broadly direction-neutral. Size tends to be pro-cyclical, declining more sharply in down markets and rising more in bull runs.
How does NTAM's approach compare to providers like Dimensional? Maarten noted that Dimensional's approach tends to stay closer to classic academic factor definitions, reflecting its academic founders. NTAM takes a more multi-dimensional and evolving approach, particularly around momentum, which plays a larger role in NTAM's portfolios given the strength of the empirical evidence, and around quality, where NTAM incorporates a wider range of signals beyond pure profitability.
Which stress period gives the most confidence? Maarten's answer was telling: not the 19th century data, but the current decade. The 2010s saw some question the future of value as it underperformed through the large cap growth rally. Since then, value has staged a strong recovery. Seeing the last five years deliver positive premiums across all three core factors, in a period when ETFs and institutional mandates are actively tracking them, is the most meaningful validation of all.
Key takeaways for advisers
150 years of evidence spanning wars, deflation, stagflation and globalisation shows these premia are a persistent feature of markets, not a product of any single era or regime.
No one factor leads in every environment. Combining value, quality, momentum and low volatility smooths cyclicality and has historically produced around 2% premium above passive equity across all inflation regimes.
Academic factor definitions are a starting point, not an endpoint. Risk efficiency, transaction cost awareness and multi-dimensional signal construction are what separate theoretical premia from real-world returns.
Research mentioned in the session
Three publications were highlighted as further reading for those who want to explore the evidence in more depth:
An accessible NTAM white paper on factor, equity, bond and gold performance across different inflation and tariff regimes, drawing on the 150-year dataset.
Download the white paper →Baltussen, van Vliet and van Vliet. The academic paper behind the 150-year dataset, examining factor premia using a novel database of US stocks predating CRSP data. Available free via SSRN.
Read on SSRN →Baltussen, Dom, van Vliet and Vidojevic. A peer-reviewed paper tracing the evolution of momentum from simple price signals through to industry, factor and network momentum. Available free via SSRN.
Read on SSRN →This content is intended for UK financial advisers only and should not be relied upon by retail clients. The views and opinions expressed are those of the presenter and are provided for informational purposes only. Past performance is not a reliable indicator of future results. Factor returns referenced are based on simulated data and do not represent actual investment returns. Nothing in this recap constitutes investment advice or a recommendation to buy or sell any security. Please refer to Northern Trust Asset Management's full disclaimers.
