When we launched the Research Affiliates Fundamental Index (RAFI) in 2005, we could not have imagined how such a remarkably simple concept would inspire the ingenuity of the indexing world in the subsequent decades. RAFI challenged cap weighting’s primacy by offering a fundamentals-weighted and hence value-tilted broad-market index. Live results¹ show RAFI has performed well against CW value indices all over the world. As we demonstrate here, the same idea extends naturally beyond value investing and can reshape how we think about CW growth and CW core indices as well.

That the CW index is the best proxy for all types of equity markets is predicated on efficient markets, among other strong assumptions. But what if markets are not perfectly efficient? Then breaking the connection between weight and price may create opportunities for better investment outcomes. Our innovation seeks to retain indexing’s benefits while redefining how we conceive of selection and weighting.² Fundamental measures can help select companies, weight them, or both for value, core, and growth indices. If we can develop indices that have historically outperformed³ in all three domains – value, core and growth, all over the world – the term “Trifecta” is hardly an exaggeration.

The essential elements of RAFI are fundamental selection and fundamental weighting. Fundamental selection means that we determine index membership based on objective measures of the underlying companies’ economic influence – such as sales, cash flow, dividends, and book value – while deliberately ignoring share price and market value, which already reflect consensus expectations of future growth. Fundamental weighting means that we size each holding according to those same measures of economic scale. Periodically rebalancing our portfolio to these fundamental weights can deliver an additional rebalancing alpha over time, when price movements are not supported by changes in underlying fundamentals.

The Research Affiliates “Trifecta” thus counters indexing orthodoxy by severing the tie between price and portfolio construction across the value, core, and growth segments. Its reliance on fundamental selection and fundamental weighting produces three distinct strategies: Research Affiliates Fundamental Index (RAFI), a value index that uses fundamentals for both selection and weighting; the Research Affiliates Cap-Weighted Index (RACWI), a core index that applies fundamental selection but retains cap weighting; and the RAFI Growth Index (RAFIG), a reimagined growth index that selects and weights based on the degree of actual business expansion rather than valuation multiples.

In RAFI’s early days, we observed that CW core indices studiously mirror the composition of the stock market, while RAFI studiously mirrors the composition of the publicly traded macroeconomy. RAFI therefore serves as an economy-weighted alternative to CW indexing. It can also complement conventional cap weighting, as Schwab⁴ and others have advocated, and provide an alternative value strategy. While RAFI introduces a stark value tilt, it differs from CW value indices by down-weighting growth stocks to their economic footprint rather than excluding them outright and by systematically contra-trading price movements not supported by changes in fundamentals.

RAFI is sometimes criticized as backward-looking because it pays no explicit attention to a company’s expected future growth. While directionally correct, this assessment misses the point. Share prices already embed the consensus market expectations about a company’s future prospects. Precisely because those expectations are reflected in prices, they offer no reliable source of future return unless they prove incorrect. By building portfolios based on fundamentals rather than forecasts, fundamental selection and weighting systematically lean against speculative narratives and can potentially harvest a reliable risk- and style-adjusted alpha when expectations are wrong.

The intuitive logic that inspired RAFI can also inform the way we manage growth and core investing. When we buy a stock, we are buying a stake in a tangible business with observable sales, R&D, profit, net worth, dividends, and distributions to shareholders. We are not speculating on the future price appreciation of a set of factors or a ticker symbol. By moving beyond market value data to focus on the observable characteristics of a company’s real business, we create better-informed indices in all segments of the market.

Growth in this sense is not a valuation multiple or a narrative. It is an observable economic phenomenon measured by the rate and magnitude at which a company expands its contribution to the macroeconomy. When we apply these ideas systematically, something striking happens. A single guiding principle gives rise to three distinct expressions that each address a different area of the equity market. In value investing, it offers a disciplined alternative to price-based indexing. In core investing, the same selection logic – combined with cap weighting – yields an index that reflects the evolving structure of the economy rather than the market’s shifting enthusiasms. When viewed through the lens of business growth, the result is an entirely new way to think about growth investing.⁵

These strategies constitute the Trifecta, our unified framework that reshapes value, core, and growth investing into a potentially better total equity experience. As Table 1 shows, across nearly three decades, all three strategies have performed well, with remarkable statistical significance,⁶ and offer a robust alternative to traditional CW indexing.

In short, the basic insights that gave rise to RAFI have far more general applications than we initially imagined. Reframing index construction around the real economy – rather than market prices – reveals a simple but surprisingly powerful way to think about equity investing. Our analysis of the underlying dynamics begins by examining how these ideas first took root in RAFI before exploring how they extend naturally across the equity landscape and, ultimately, why they have the potential to work so well.

Transformative ideas are often the result of challenging conventional thinking. RAFI is a testament to that insight. For years, much of the investment industry accepted the premise that markets are efficient, and as a consequence, that the market-clearing CW portfolio offers the highest risk-adjusted returns and maximizes investor utility in the context of mean-variance optimality.⁷ Moreover, if the market is efficient, by extension, alternatives to cap weighting are inherently sub-optimal and therefore wastes of time. This is what academic finance has been teaching for 60 years with varying degrees of stridency depending on the era and institution. It’s no wonder then that the RAFI concept angered so many people when we first proposed it 20 years ago.

The original Fundamental Index transformed our theoretical departure from cap weighting into a practical application through its four measures of company size: sales, cash flow, dividends, and book value. Subsequent variants have applied different metrics, but the core principle, that portfolio construction should be driven by a company’s measured role in the real economy, not by fluctuations in its market price, is unchanged.

This has translated into statistically meaningful outperformance across markets and regions. Table 2 shows the track records of various RAFI strategies from April 1998 through December 2025.⁸ Over this span, the RAFI US Large Index beat the Russell 1000 Value Index by an annualized 2.5%. With only 2.4% tracking error relative to Russell 1000 Value, this index provided an information ratio of 1.01, and a t-Statistic of 5.

Figure 1 shows the wealth advantage that investing in RAFI has produced over the CW value indices in Table 2. Live RAFI index results span roughly two-thirds of this period, with launch dates ranging from late 2005 to late 2007.⁹

Several features in Figure 1 stand out. The most obvious is the sheer economic magnitude of the effect. In emerging markets, RAFI compounds to more than three times the total wealth of the cap-weighted value benchmark over the sample. Even in developed markets, with presumably more efficient pricing, the cumulative gains remain substantial, reaching about 60% to 80% in both the US and developed ex-US markets. These results are not incremental; they are the long-run consequence of modest but persistent differences in annual returns and outperformance in roughly three of every four years. Further, their cross-regional ordering feels entirely intuitive. The strongest effects occur in emerging markets, where price signals are noisier, index reconstitutions are costlier, and pricing errors tend to be more pronounced. Developed markets offer fewer such opportunities yet still reward the same discipline over time.

CW indices were a remarkable innovation and have continued to improve over the years. But they retain an Achilles heel: While selection rules differ by index provider, the typical catalyst for adding a stock is a substantial run-up in price, leading to a market value that's too large to ignore, while a tumbling price is the proximate cause for most discretionary deletions.¹⁰ Traditional CW index additions perform brilliantly before being added to the index but often fall behind in the years just after their inclusion. Deletions perform abysmally before being dropped from the index but on average outperform after elimination.¹¹ RACWI seeks to exclude the performance drag of buying high and selling low when stocks are added or nixed while retaining the market exposure, liquidity, and low turnover of the legacy CW indices.

We construct RACWI by selecting stocks according to measures of the fundamental size of the underlying companies rather than their much more volatile market prices and then weighting the stocks based on their market capitalization or float. In the US market, this CW index has historically delivered long-term outperformance. As both RACWI and legacy CW indices are market-cap weighted, this is entirely attributable to the "difference portfolio," the roughly 5% to 10% of the index where RACWI diverges from its conventional CW peers.

Table 3 shows RACWI’s performance characteristics from April 1998 through December 2025.¹³ Across all regions, RACWI delivered higher returns than its traditional CW counterparts with virtually indistinguishable volatility.

RACWI US outperformed the Russell 1000 by 0.7% per year. With a tracking error of 1.5%, this excess return delivers an information ratio of 0.47 – a strong result for a strategy with an overwhelmingly market-like construction. Very similar international results provide independent validation, and emerging markets again show the strongest results. Fundamental selection beats market value selection for cap-weighted indices all over the world.

These results are especially compelling because they are achieved without introducing systematic factor tilts or materially altering portfolio risk. The gains do not come from loading on value, size, or momentum. Instead, they arise from a quiet source: avoiding the inherent buy-high, sell-low bias of cap-weighted index reconstitutions. Again, selecting stocks based on fundamental size rather than market capitalization helps RACWI sidestep this drag while otherwise behaving like a conventional market index.

Figure 2 shows how RACWI’s excess returns over Table 3’s benchmark indices compound over time. In the US and developed ex-US markets, RACWI delivers about 20% more cumulative wealth than conventional CW benchmarks over the sample. Emerging markets exhibit an even larger effect, consistent with the greater frictions associated with cap-weighted index reconstitutions.

RACWI’s live performance has been especially impressive. Since its September 2021 launch, RACWI US has gained roughly 167 basis points (bps) per annum in excess return relative to the Russell 1000 (and by half as much relative to the S&P 500).¹⁴ For a passive CW index, changing only the rules by which stocks enter or leave the index highlights how a small tweak in index construction can produce economically meaningful gains. For investors seeking market-like performance with a modest but persistent edge, RACWI offers a compelling refinement of traditional cap-weighted indexing.

RAFIG corrects what we believe is a critical flaw in traditional growth investing, which conflates “expensive” with “growth.” The binary duality of growth versus value – where cheap means value and expensive means growth – is nonsense. Expensive does not equate to growth, it equates to expensive. We may be the first to have created a growth index that selects stocks on their percentage growth rates and then weights them on the dollar magnitude of that growth. This is RAFI’s fundamental selection and fundamental weighting reimagined into the growth arena. Big growth stocks get more weight, jointly proportioned to the size of their business and their percentage growth rates, without any regard for valuation multiples or price. With this simple concept, RAFIG may serve growth investors in the next 20 years as well as its RAFI predecessor has served value investors for the previous 20.

Standard growth indices select stocks partly based on price-to-earnings (P/E) ratios or other high valuation multiples as well as projected growth, which is already reflected in share price. This effectively defines growth as high prices. RAFIG selects stocks based on the observed historical growth of the underlying companies; it chooses companies that are really growing rather than those priced for future growth. We construct RAFIG by selecting stocks based on the growth rate of the sales, profitability, and R&D spending of their underlying companies and then weighting by the dollar value of that growth.

Table 4 shows how RAFIG strategies fared by region from April 1998 through December 2025.¹⁵ The results are striking. Across all regions, RAFIG substantially outpaced traditional cap-weighted growth benchmarks, delivering large excess returns with only moderately higher volatility and superior risk-adjusted performance.

RAFIG US provided a 14.2% annualized return, beating Russell 1000 Growth by 4.7%. Although its volatility is somewhat higher, 19.6% to 17.7%, the improvement in returns more than compensates, with a 0.62 Sharpe ratio to Russell 1000 Growth’s 0.42. The 7.1% tracking error reflects RAFIG’s deliberate departure from valuation-based growth definitions. RAFIG’s 0.67 information ratio is also impressive.

International markets reinforce these findings. Developed ex-US outperformed the MSCI World ex-US Growth index by 2.9% per year, while the excess return of RAFIG Emerging Markets relative to MSCI Emerging Markets Growth reached 3.1% per year. In both cases, the improved returns came with higher – but not disproportionate – volatility and produced consistently better Sharpe ratios than those of legacy growth benchmarks.¹⁶

Figure 3 shows the cumulative relative wealth of RAFIG strategies versus the conventional growth benchmarks in Table 4. The patterns are unmistakable. In all regions, RAFIG steadily outperformed cap-weighted growth indices, with no extended reversal periods. In the US and globally, RAFIG compounded to roughly three times the relative wealth of the Russell 1000 Growth benchmark over the full sample. Developed ex-US and emerging markets show similarly persistent advantages, effectively doubling the wealth gains of their conventional peers. These are not short-lived episodes tied to a particular market cycle or style rotation; they reflect the long-run consequence of anchoring growth exposure to actual business expansion rather than to stocks priced for growth.

Taken together, Table 4 and Figure 3 demonstrate that redefining growth as an observable economic phenomenon – rather than a collection of expensive stocks – could lead to a comprehensively different and far more effective approach to growth investing. RAFIG seeks to capture growth when it actually occurs, which is often before the market fully recognizes or prices it.

Our fundamental selection and fundamental weighting methodology introduces three alphas:

• RAFI earns one alpha from its dynamic value tilt – its value exposure changes over time. To be sure, since 2007, value has had its longest and deepest span of underperformance in at least the past half-century. From its relative performance peak in early 2007 to its nadir in summer 2020, Russell 1000 Value lagged the Russell 1000 by 38%, or roughly 3.5% per annum compounded. That means Russell Growth beat Russell Value by 7% per year over these 13.5 years.¹⁷ RAFI’s value tilt tends to be lower when the relative valuation gap between value and growth stocks is narrower than usual and deeper when value stocks are abnormally cheap. When valuations mean revert, this dynamic tilt can help RAFI outperform traditional value indices.
• RAFI and RAFIG each have a rebalancing alpha, trimming stocks that have soared and topping up stocks that have tanked whenever the underlying fundamentals haven’t validated the price moves.
• Each of the three - RAFI, RACWI, and RAFIG - has a selection alpha. Legacy CW indices add stocks as they rise into the top tier of market capitalization and drop them as they fall off. This means that stocks are added as a direct consequence of soaring market valuation and are dropped in response to a plunge in share prices. As Arnott, Brightman, Kalesnik, and Wu detail,¹⁸ this has consequences. RAFI waits for the fundamentals to validate that soaring or crashing price. As such, it may be late to the next Nvidia or Tesla, but it is designed to entirely miss the Lucid-like free falls.

Summary Musings . . .

Much of the evidence presented in this paper is necessarily historical and backward-looking. But the results are not fragile curiosities produced by finely tuned parameters or clever modeling. The philosophy that informed RAFI’s creation two decades ago is not a niche, data-mined alternative to capitalization weighting or a better way to build a value index. It is a general framework for equity indexing.

Across value, core, and growth and among regions with very different market structures, the same organizing principle produces consistent and significant improvements. When index construction is a function of the scale and growth of real businesses rather than their market prices, portfolios behave differently. They add companies because their economic footprints have grown, not because their share prices have soared. They reduce exposure when fundamentals falter, not when prices collapse. Prices are inherently volatile and reflect changing sentiments, random asset flows unrelated to company fundamentals, and possible investor miscalculations. By refocusing the indexing process on fundamentals, we can systematically resist the buy-high, sell-low dynamics embedded in price-driven index reconstitutions.

What is striking is not only that this framework works, but that it also works in three distinct ways. In RAFI, it delivers a disciplined and dynamic approach to value investing. In RACWI, it refines cap-weighted indexing by improving stock selection while preserving market-like characteristics. In RAFIG, it redefines growth itself – away from valuation multiples and narratives and toward observable, realized business expansion. These three expressions are not separate ideas. This is the Research Affiliates “Trifecta”: a single insight expressed coherently across value, core, and growth.

These results do not depend on forecasting, factor timing, or increasingly elaborate models. They emerge from the deliberate choice to distinguish between prices, which reflect expectations, and fundamentals, which reflect economic reality. That distinction proves powerful across markets, across decades, and across investment styles.

As index investing continues to evolve, it does not have to be towards ever-expanding complexity. Sometimes progress comes from asking simpler questions and answering them consistently. One such answer is to anchor portfolio selection, weighting, or both to the real economy rather than to market prices. This simple principle appears to work far better and far more broadly than conventional wisdom would suggest.