Cash, Accruals, Intangibles and Sector Effects: What Really Separates Avantis and Dimensional?

Avantis and Dimensional are often grouped together for good reason. Both firms sit firmly within the evidence-based investing space, both draw heavily on the academic asset-pricing literature, and both reject discretionary stock picking in favour of systematic exposure to well-documented sources of expected return. This shared philosophy is not accidental. Avantis was founded by former Dimensional professionals, and the firm’s approach reflects the same intellectual lineage: factor investing grounded in theory, long-run empirical evidence, and an emphasis on implementability rather than storytelling.

Yet despite this shared school of thought, Avantis and Dimensional do not construct identical portfolios. The differences are subtle rather than ideological, but they matter. A large part of the divergence comes down to how each firm interprets accounting data, how far they are willing to adjust it, and how they weigh the trade-off between economic precision and robustness in live portfolios.

Two debates dominate these differences. The first concerns how profitability should be measured, specifically whether reported earnings should be adjusted to be more cash-like by stripping out accruals. The second concerns how book equity should be defined in an economy dominated by intangible assets, and how that definition interacts with value metrics and profitability scaling. These debates are related, but they are not the same. One is primarily about the numerator in profitability ratios, the other about the denominator. As Eduardo Repetto puts it, ‘what you care [about] is these two things (profitability and book value) together’ (Felix and Passmore 2022). The valuation of a company considers the equity position and the expected cash flows that the company is generating. These two things together, combined with the discount rate, form the stock’s price.

Taken together, these two debates explain a large share of the philosophical and portfolio-construction differences between Avantis and Dimensional.

What Does it Mean to ‘Scale Profitability by Book Equity’?

Before getting into where the firms disagree, it is worth clarifying what both are actually doing when they talk about profitability. In factor models, profitability is almost never used in absolute terms. Instead, profits are scaled by some measure of invested capital, most commonly book equity. Conceptually, this converts profits into a return-on-capital-type measure.

A firm earning £100 million in profits tells us very little on its own. A firm earning £100 million on £500 million of book equity looks very different from one earning the same amount on £5 billion. Scaling profits by book equity allows comparisons across firms of different sizes and embeds an economic intuition: firms that are able to generate higher profits per unit of capital have historically been associated with higher expected returns, once other characteristics are controlled for.

The difficulty is that both the numerator (profits) and the denominator (book equity) are accounting constructs rather than direct economic quantities. If either is distorted, the resulting ratio may give a misleading picture of a firm’s true economic profitability. The disagreements between Avantis and Dimensional can therefore be understood as disagreements about how much correction is warranted on each side of that ratio.

Profitability and Accruals: Cash versus Accounting Earnings

The first major point of contention concerns the definition of profits themselves. Dimensional largely follows the operating profitability measure introduced by Fama and French, which uses operating income before depreciation and amortisation, minus interest expense, scaled by book equity. This measure stays relatively close to standard accounting earnings and avoids extensive adjustments. Dimensional’s view is that whilst accounting numbers are imperfect, they are broadly informative, and that adding further layers of correction risks increasing complexity and turnover without delivering a clear and robust improvement in expected-return targeting.

Avantis takes a slightly more interventionist approach and argues for cash-based operating profitability. Starting from the same point as Dimensional, Avantis also takes off accruals to make profitability more closely resemble cash flows from operations. This involves explicitly penalising accruals and other non-cash components of earnings. To understand why this matters, it is important to be clear about what accruals actually are.

Accruals arise because accounting is designed to reflect economic activity, not cash movements. Revenues are recognised when they are earned, not when cash is received. Expenses are matched to the periods in which they help generate revenue, not necessarily when cash is paid. As a result, profits can move independently of cash flows, sometimes for perfectly sensible reasons, but sometimes in ways that later reverse. For example, it makes sense to include revenues from credit-worthy debtors or to recognise non-cash charges like depreciation and amortisation in a measure of profitability. Equally, though, it doesn’t make sense to recognise one-off accounting items that boost earnings in the short-term but are not a sustainable source of profit.

Common examples of accruals include recognising revenue before customers pay (creating trade debtors), spreading the cost of assets over time through depreciation or amortisation, creating provisions for future costs, or adjusting inventories and working capital. In very simple terms, accruals can be thought of as the gap between accounting profit and cash profit. High accruals mean that a large portion of reported earnings is not backed by contemporaneous cash flows.

This distinction has proven empirically important. Some academic evidence shows that firms with high accruals tend to earn lower future stock returns, even after controlling for other characteristics (Sloan 1996). One interpretation is that accrual-heavy earnings are of lower quality and more prone to reversal, either because managers are using accounting discretion aggressively or because current profits reflect temporary timing effects rather than sustainable cash generation.

Avantis leans directly into this evidence. By rewarding firms whose earnings are strongly supported by cash and penalising firms where accruals dominate reported profits, it aims to target a form of profitability that is closer to ‘economic reality’.

Dimensional, by contrast, acknowledges the accruals evidence but argues that the effect is a US phenomenon and has gotten weaker over time. Furthermore, Dimensional argues that cash-based operating profitability is a lot less persistent in terms of its ability to predict future cash flows (Felix and Passmore 2025). Mamdouh Medhat even states that it ‘actually exhibits mean reversion in the sense that companies that look high on cash profitability today, actually tend to grow negatively or stagnate in terms of their cash profitability going forward’ (Felix and Passmore 2025). He goes on, ‘we're talking about earnings growth over long periods of time, five years, seven years. [Therefore], we think that's a pretty compelling argument for sticking with operating profitability’ (Felix and Passmore 2025). Consequently, Dimensional prefers a profitability measure that is simpler and arguably more predictive going forward, possibly due to being less sensitive to accounting-based noise.

Overall, there are clearly good arguments on both sides of the debate and it is hard to decipher who is actually ‘right’.

Book Value, Intangibles and Sector Effects

The second major debate between Avantis and Dimensional concerns the denominator of profitability ratios and the construction of value metrics such as book-to-price. In principle, book equity is meant to capture the capital backing a business, providing a scaling base for profits and a reference point for assessing how cheaply or expensively a firm is priced. In practice, however, modern accounting treats intangible assets inconsistently, and those inconsistencies have become increasingly important as intangible investment has grown.

When a firm acquires another company and pays more than the target’s book value, the excess is recorded on the balance sheet as goodwill or other acquired intangibles. By contrast, when a firm builds intangible assets internally, through spending on research and development, software, brand building or customer acquisition, those expenditures are typically expensed immediately through the profit and loss account and never appear in book equity. Two firms with very similar economic assets can therefore report radically different book values purely because of how those assets were acquired. One firm looks asset-rich, the other asset-light, even if their underlying businesses are economically comparable.

Sector effects magnify this problem. Some industries naturally rely much more on intangible capital than others. Technology, healthcare and communications firms invest heavily in R&D, data and software, whilst utilities, materials and traditional manufacturing depend far more on physical capital. As a result, unadjusted book-to-price ratios vary systematically across sectors. Intangible-heavy sectors tend to look expensive, whilst tangible-heavy sectors tend to look cheap, even when this reflects industry structure rather than genuine mispricing. A naïve value screen can therefore end up embedding large, unintended sector bets.

Dimensional’s research suggests that once these sector effects are dealt with directly, for example by controlling sector weights or examining expected returns within sectors, the incremental benefit of adjusting book value for intangibles is modest. From this perspective, much of what appears to be an ‘intangibles problem’ is actually a sector composition problem. If value and profitability are assessed in a way that already accounts for industry structure, the case for aggressive balance-sheet adjustments weakens considerably.

Avantis takes a different view, particularly with respect to goodwill, and this is where the disagreement becomes more subtle. As Eduardo Repetto explains, firms with very high market valuations relative to fundamentals effectively have a very low implied discount rate. For such firms, equity becomes a form of hard currency. From a corporate-finance perspective, it can be entirely rational for management to issue shares and acquire other companies, even at a premium, because the transaction can still have positive net present value for existing shareholders (Felix and Passmore 2022). In that sense, using a firm’s own expensive equity to buy other businesses can genuinely create value. Think of it as using £1 to buy £2.

The problem Avantis is concerned with is not whether these acquisitions are smart, but how they interact with value signals. When a firm repeatedly acquires other companies using high-priced equity, the premiums paid are capitalised as goodwill, mechanically inflating book equity. The firm can begin to look cheaper on a book-to-price basis, not because its expected return has risen, but because future expectations have been rolled into the balance sheet through accounting. In other words, goodwill reflects the price paid for expected future profits, yet those same expectations are already embedded in market prices that the acquiring firm will pay for. Thus, from Avantis’ perspective, to include that goodwill in the acquiring firm’s book equity risks double-counting future expected profits. Consequently, serial acquirers with low discount rates can drift into value portfolios simply because of how accounting records their capital allocation strategy, not because they offer higher expected returns.

Subtracting goodwill from book value is therefore an attempt to prevent book-to-price ratios from being distorted by acquisition activity and to keep the value signal focused on differences in discount rates and thus expected returns rather than differences in corporate behaviour.

Estimation Noise and the Limits of Adjustment

A central concern for Dimensional is estimation noise. Intangibles that are not recorded on the balance sheet are not directly observable. To add them back into book value, an investor must build a model. That typically involves capitalising several years of R&D or other expenditures and making assumptions about useful lives, amortisation patterns and depreciation rates.

Some of what comes out of this process may be genuine signal, a better reflection of true underlying intangible capital. But some of it is noise: variation driven purely by modelling choices, data limitations and arbitrary parameters rather than real economic differences. Data availability is also uneven. As Dimensional has noted, R&D data are only available for about half of the US market, meaning that any adjustment would be applied to some firms but not others, potentially reducing comparability rather than improving it (Felix and Passmore 2025).

The key point is that once sector effects are already addressed directly, the marginal benefit of these adjustments appears small, whilst the estimation noise can be large. In practical terms, the modest improvement in measuring value or profitability can easily be cancelled out, or even overwhelmed, by the wobble introduced by the estimation process itself. This trade-off sits at the heart of Dimensional’s scepticism.

The disagreement, then, is not about whether intangibles or goodwill are ‘real’. It is about whether including them in book equity improves or degrades the usefulness of value metrics once sector effects, profitability and robustness concerns are taken into account.

Related Debates, Different Levers

It is tempting to treat accruals, cash-based profitability and intangibles as part of a single debate about ‘better accounting’. In reality, they operate on different levers. The cash versus operating profitability debate is mainly about the numerator: how profits are defined and how aggressively non-cash components are stripped out. The intangibles debate is mainly about the denominator: how book equity is defined and whether it adequately captures the capital invested in the business.

They feel related because Avantis adjusts both sides of the ratio. By combining cash-based profitability with an adjusted book value, Avantis ends up with a stronger profitability emphasis and somewhat different sector tilts. Dimensional, by contrast, treats value, profitability and investment as related but distinct dimensions and remains cautious about layering on adjustments that may not be robust over long horizons and across markets.

Are These the Main Points of Contention?

On the technical design side, these are among the most important differences. Dimensional prefers operating profitability and largely unadjusted book value, prioritising robustness, breadth and low sensitivity to modelling assumptions. Avantis prefers accrual-adjusted, cash-like profitability and a cleaner measure of book equity that removes goodwill distortions, prioritising accounting precision and economic interpretation.

They are not the only differences. The firms also differ in how strongly they tilt away from the market, how they balance value, profitability and investment signals, how concentrated their portfolios are, and how they implement trading, momentum screens and tax management. But conceptually, the debates around accruals, intangibles, sector effects and estimation noise explain a large share of why two firms drawing from the same academic literature can arrive at meaningfully different portfolios.

To Conclude

The accruals and cash-profitability debate and the intangibles and book-value debate are separate but connected. One concerns how cash-like profits should be; the other concerns how complete and comparable book equity is in an economy dominated by intangible capital and strong sector effects. Dimensional’s response emphasises robustness and parsimony, favouring measures that are simple, stable and resilient to estimation error. Avantis’ response emphasises accounting realism and economic precision, seeking to correct known distortions even at the cost of additional complexity.

At the end of the day, however, these differences amount to splitting hairs within the same evidence-based framework. Both firms are drawing from the same academic literature, targeting the same underlying sources of expected return, and making thoughtful but defensible trade-offs between signal purity and robustness. For investors, the more consequential distinction is not between Avantis and Dimensional, but between systematic, research-driven implementation and traditional discretionary fund management. Relative to the latter, either approach is far more likely to deliver better, more sensible and repeatable outcomes over the long run.

References

Felix, Benjamin, and Cameron Passmore, hosts. 2022. ‘Episode 228: Eduardo Repetto: Deep Dive with Avantis Investors' CIO.’ The Rational Reminder Podcast, November 24, 2022 (audio podcast).

Felix, Benjamin, and Cameron Passmore, hosts. 2025. ‘Episode 384: Mamdouh Medhat – A Profitability Retrospective, and Private Fund Performance.’ The Rational Reminder Podcast, November 20, 2025 (audio podcast).

Sloan, Richard G. 1996. ‘Do Stock Prices Fully Reflect Information in Accruals and Cash Flows about Future Earnings?’ The Accounting Review 71 (3): 289–315.