What Is the Size Factor and Why Is It More Than Investing in Small Caps?

The size factor systematically tilts an equity portfolio towards smaller companies. In a market-capitalisation-weighted benchmark, the largest companies have the highest weights. A size-factor portfolio reduces these weights and increases the weights of smaller companies.

It is important to distinguish between a position and positioning. A large company can be included in the portfolio and still be significantly underweighted. In factor investing, what matters is not the factor label attached to an individual stock, but the positioning of the overall portfolio.

Small caps represent a market segment. A small-cap index contains a different equity universe from a standard index and may therefore reflect different countries, sectors and company characteristics. Its performance is also influenced by valuation, profitability, leverage, sensitivity to the economic cycle, idiosyncratic volatility, beta and market liquidity.

The same applies to an equal-weight index. Equal weighting reduces the influence of very large companies, but often also changes the country and sector structure, quality exposure, valuation, momentum, beta, liquidity and single-stock risks.

A small-cap index or equal-weight index may have strong size exposure, but it is not automatically a Pure Size portfolio. Moreover, the largest factor exposure does not necessarily have to be the strongest return driver in a particular market environment. Even a smaller exposure can have a major impact if the respective factor moves particularly strongly.

Size, value, quality and momentum tilts can also be analysed within a small-cap universe. It is problematic to attribute the entire return differential versus a standard index to the size factor without examining the other exposures.

Whether a company is classified as a large cap, mid cap, small cap or micro cap should not be determined using fixed thresholds in EUR or USD. Inflation, interest rates, market valuations and broad price movements change market values without necessarily altering the relative size relationships to the same extent.

For example, if a company's market capitalisation falls from EUR 10 bn to EUR 7 bn during a broad market decline, it could cross a fixed threshold and be classified as a small cap. Its relative position may hardly have changed, however, because comparable companies have also fallen.

A relative classification within the respective market or investment universe is more meaningful. The key measure is free-float-adjusted market capitalisation. Market capitalisation refers to a company's stock-market value, while free float refers to the proportion of its shares that is generally available for public trading. This means that only the investable proportion is taken into account.

The MSCI methodology provides an illustrative example. In simplified terms, companies are ranked by size and classified according to the cumulative proportion of free-float-adjusted market capitalisation they cover:

• The Large Cap Index targets approximately 70% coverage of an investable market.
• Large and mid caps together form the Standard Index, with target coverage of approximately 85%.
• Large, mid and small caps together form the Investable Market Index, with target coverage of approximately 99%.
• This results arithmetically in approximately 15% for mid caps and a further 14% for small caps.
• Micro caps sit below the small-cap segment and are subject to additional size, liquidity and investability requirements.

If 100 companies have a combined free-float-adjusted market capitalisation of EUR 100 bn, approximately EUR 70 bn would be attributable to large caps, a further EUR 15 bn to mid caps and EUR 14 bn to small caps. These figures do not refer to the number of companies. A small number of large companies may already account for a significant proportion of total market capitalisation.

The remaining 1% does not automatically belong entirely to the micro-cap segment. MSCI includes eligible micro caps in developed markets in separate All Cap Indexes. The size segments are determined at the individual-market level and then combined to form regional and global indexes.

The figures mentioned are target ranges within a methodology, not fixed thresholds for individual companies. Other index providers may therefore arrive at different classifications.

Revenue, earnings, total assets or enterprise value can also be used to describe company size, but they do not necessarily produce the same result. Defining size based on revenue or earnings may create additional quality or value exposure. A more complex metric is therefore not automatically a purer measure of size. A simple size metric may be sufficient if other return and risk dimensions are controlled separately.

In simplified terms, companies are classified in three steps:

• Step 1: All companies are ranked from largest to smallest based on their free-float-adjusted market capitalisation.
• Step 2: The market values of the companies are then added together cumulatively.
• Step 3: All companies that together account for approximately 70% of total free-float-adjusted market capitalisation form the large-cap segment. They are followed by the mid caps and then the small caps.

The important point is that these percentages do not refer to the number of companies. They do not mean that 70% of all companies are large caps. Because the largest companies have particularly high market values, a comparatively small number of companies may already account for approximately 70% of total market capitalisation.

The size premium refers to a potential return differential between smaller and larger companies. One central explanation is a risk premium: Smaller companies may offer greater growth opportunities, but may also involve additional risks.

A company with only a few products and a small customer base may, in theory, be able to multiply its revenue or market position more substantially than an established corporation. This does not mean, however, that smaller companies automatically grow faster. Many are unable to scale their business model successfully.

Smaller companies are often less diversified and more dependent on a single product, a few major customers, individual suppliers or specific end markets. Financing may be more difficult or expensive, and their shares may be less liquid. Information asymmetries may additionally facilitate mispricing, but they are not the central explanation.

The size premium can therefore be understood as potential compensation for taking additional risks. It is not certain: Not every smaller company grows successfully, and not every risk is rewarded with a higher return.

Small Minus Big, or SMB, is often used in academic research to measure the size premium. In the Fama-French three-factor model, the average return of three portfolios of smaller companies is compared with that of three portfolios of larger companies. The stocks are also sorted according to their book-to-market ratio.

Economically, SMB is a theoretical long-short factor portfolio: Long small and short big. It must be distinguished both from an investable long-short portfolio and from a long-only small-cap portfolio. The latter remains fully exposed to general equity market risk.

The theoretical SMB return is not directly attainable by investors. In practice, transaction costs, bid-ask spreads, financing costs, securities lending costs, rebalancing costs and liquidity costs arise. These costs can be particularly significant on the short side.

The assessment depends heavily on the measurement method, definition, investment universe, period, breadth of the data and costs. A small-cap index, an equal-weight index and a simple SMB portfolio do not necessarily measure the same thing. Long periods and broad universes are important because factor premiums are cyclical and small regional datasets can be shaped by individual stock outcomes.

A measurement designed to isolate size more purely can make the influence of company size clearer, but it provides neither a guarantee nor conclusive evidence of future returns. Different broad equity universes may also exhibit a similar risk/return profile if their factor exposures are comparable.

The size factor is a distinct driver of returns and risks, but it does not exist in isolation from value, quality, momentum or low risk. Factors can overlap, reinforce one another, partially offset one another or mask one another. Because a market-capitalisation-weighted benchmark concentrates high weights in large companies, size positioning can deviate particularly strongly from the benchmark and dominate other intended factor exposures.

A conventional small-cap index or equal-weight index may also exhibit a "junk bias" – a tendency towards less profitable, more highly leveraged or more idiosyncratically risky stocks. These characteristics can increase default risk, but they are not necessary in order to give smaller companies higher weights. A Pure Size approach can control profitability, leverage, beta and idiosyncratic volatility separately, without implying an automatic return or risk advantage.

Business model risk often results from having fewer products, customers and sources of revenue. Financing risk may increase when capital providers demand higher credit spreads. In a broad small-cap universe, higher average leverage and more limited financing options can therefore increase sensitivity to rising refinancing costs. However, this does not apply to every smaller company.

When market liquidity is low, larger orders may have a greater impact on the price, bid-ask spreads may be wider and a sale during a period of market stress may only be possible at a discount. Particularly among micro caps, transaction costs, market impact, limited capacity and rebalancing costs can reduce or even eliminate a theoretical premium. A smaller company therefore does not automatically offer a greater risk-adjusted benefit.

When analysing market environments, two meanings of cyclicality must be distinguished. From a benchmark perspective, size tends to have a contrarian character because particularly large and heavily weighted market segments are underweighted. From an economic perspective, smaller companies may at the same time be more sensitive to the economic cycle than broadly diversified corporations. Favourable credit conditions may support them, while a weaker economy or rising refinancing costs may weigh more heavily on them.

These relationships are not mechanical. Beta is not stable either: A portfolio may appear to have low market sensitivity during calm periods and react more strongly in periods of stress. Similarly, a rising overall index is not necessarily positive for the size factor if the market's performance is driven by only a few large caps, mega caps or individual sectors.

Broad diversification can significantly reduce single-stock risk, but it cannot eliminate all risks. A long-only portfolio remains exposed to general equity market risk, factor risk, economic risks, liquidity risks as well as model and implementation risks.

A Pure Size portfolio begins with a broad equity universe and a clearly defined benchmark. For each company, other return and risk characteristics are considered alongside market capitalisation, including country, region, sector, valuation, profitability, leverage, momentum, beta, idiosyncratic volatility and liquidity.

The desired size exposure is then increased through targeted overweighting of smaller companies and underweighting of larger companies. At the same time, deviations in other known risk dimensions and at single-stock level are limited. The size positioning should arise from many controlled changes in weights rather than from a small number of pronounced single-stock bets.

A large company can therefore remain in the portfolio while being significantly underweighted relative to the benchmark. Its position can help control country allocation, sector allocation or beta. This again shows that what matters is the relative positioning of the overall portfolio.

Technically, this can be achieved through portfolio optimisation. The optimisation seeks weights that combine as much of the desired size exposure as possible with as few unintended bets as possible. Risk management thereby becomes an integral part of portfolio construction. However, only characteristics that are known, measurable and included in the model can be controlled.

A Pure Size approach does not automatically lead to higher returns or lower risks. Its objective is to attribute the drivers of returns and risks more clearly. This is the decisive difference between a simple small-cap filter and a Pure Factor approach.

The size factor means giving smaller and, where appropriate, mid-sized companies higher weights and very large companies lower weights relative to a relevant benchmark. A small-cap index or equal-weight index may have significant size exposure, but it does not automatically provide pure exposure to the factor because it can also introduce country, sector, factor and liquidity effects.

The measurement method, investment universe, period, implementation costs and, above all, portfolio construction are crucial to the assessment. A Pure Size portfolio seeks to make the intended exposure to company size visible while limiting other known risk dimensions. The size premium remains a potential, cyclical risk premium and not a guarantee.

The most important practical insight is that factor interactions must be identified, measured and managed. Only then is it possible to determine whether a portfolio genuinely has clear size exposure or whether other drivers of returns and risks dominate the outcome.