Value? What value?

Are the investment style benchmarks launched in the 1980s and 1990s still fit for purpose?

I am often assumed to be a value investor. In most instances over the past decade or so, the association has been presented in a sympathetic fashion, with a “your time will come” coda. By others, it is posed as a “why would you do that” accusation in the age of all things growth. From my end, four books, dozens of podcasts, and too many social media posts, have tried to define what being a Dividend Investor in a Stock Market means. Yet the confusion lingers, in large part because it is far from clear—in fact utterly unclear—what value investing actually means in this day and age.

Now to be fair, some terminological overlap is justified. Packaged investment products (SMAs, funds, ETFs, even institutional accounts) are distributed or at least pass through intermediaries who categorize them according to existing semantic and actual product silos. There is one called large value, but no major (or minor) platforms have a category called Dividend Investor in a Stock Market.

And institutional gatekeepers still operate in the peculiar world of relative total return benchmarks dating from 40 years ago when factor-based investing was coming to the fore. The dividend factor was and still is considered just a subset of the broader value approach. So Dividend Investors in a Stock Market are usually measured from a total return perspective against the leading value index introduced in 1987. But using a total return benchmark in a market yielding 1% (and a value benchmark with a yield below 2%) just doesn’t do much beyond measuring near-term share price movements. These tools do not even purport to capture dividend growth as a driver of long-term share-price appreciation. For dividend-focused investors, not incorporating the dividend growth component, and its total return consequences, is a critical failure.

Similarly, the semantic landscape in which many investors still operate is largely a function of a classification system created at a very particular moment in institutional finance. With the rise of the style box in the early 1990s (introduced in 1992 by another market observer), investment products were increasingly understood as leaning toward value, toward growth, or toward some blend of the two, with size layered on top. There were, and still are, no other style options in the box.

So while the common association with value investing is fully understandable, as a practical matter it’s just not true, not least because I can state (and have, repeatedly) what being a Dividend Investor in a Stock Market means circa 2026, or any other year. And that approach has minimal overlap with what currently passes for value investing. It’s time for the latter’s ambiguity to be resolved. Forty years is long enough.

One thing current value investing is not is what Ben Graham outlined initially in the 1930s. While one can still have notions of intrinsic value above current market value, having an investment framework based on that maxim is far different today from what it might have been in Graham’s day. When you go back and read the 1934 edition of Security Analysis, much of it, when translated into modern parlance, reads like a special situations handbook, with specific circumstances applying to each security. For some stated reason, the net asset value of Acme Widget is greater than its market value. Or the price for Roadrunner Contraption in a particular context did not reflect its core earnings power recovering from the Great Depression. In both instances, the opportunity was linked to a stated line on the balance sheet (more often) or the income statement (increasingly in later editions).

Graham’s approach was not a “style;” it was not a factor. It was a company-specific circumstance, with an implied or articulated catalyst or resolution in the near or intermediate term. Graham did not even use the term value investing. For him it was just sensible security analysis versus outright speculation. The term was applied decades later and applied retroactively.

Keep in mind that at that time, there were no valuation-insensitive index funds dominating investment flows, no style boxes, no 500- or 1000-stock benchmarks and gatekeepers looking over your shoulders every 90 days asking what “worked” and what didn’t against the index. Indeed, systematic, regular calculation of total return emerged only in the post-war period when the data gathering technology became widely available. While there were rudimentary sector indices and a weekly (from 1926) broader market cap-weighted index, there was nothing that we would recognize today as a value ecosystem, and certainly not the two rigid notions of growth and value as exclusive investment styles, with their own total return indices, toggling back and forth as if they were opposite ends of a single spectrum, and competing for institutional support.

Fast forward a few decades, and the “factor-based” value investing that emerged in the 1970s and 1980s is as remote from Graham in the 1930s as it is from being a Dividend Investor in a Stock Market in 2026. At least Graham knew the names of the companies he considered good values and what line of business they were in. But that was not what value investing became after the tsunami of quantitative analysis and then quant investing itself flooded the investment landscape in the decades after the Second World War.

The development of the Capital Asset Pricing Model (CAPM) in the mid 1960s identified a security’s market sensitivity (beta) as a key driver in explaining investment returns. Merrill Lynch’s printed beta books were the result. The increasing availability of fundamental data about companies—the Compustat database was launched in 1962—allowed Barr Rosenberg and other academics to go further and propose that total return and (their definitions of) risk could be decomposed into numerous components beyond just market sensitivity. Modern factor investing was born. The shift from investing in businesses through the stock market to investing in disembodied stock-market factors is the triumph of modern finance as told enthusiastically by Peter Bernstein in Capital Ideas from 1992. It is contextualized and critiqued in my own Getting Back to Business from 2018.

But notice what was lost in translation. Graham’s intrinsic value was a conclusion drawn from an investigation of a specific business. In contrast, the modern value factor was the aggregation of a screening exercise of thousands of securities. The two approaches share the name value, but they were (and remain) fundamentally different exercises. One asks: what is this company really worth and is there a path to realizing that figure? The other asks whether a portfolio of low price-to-book stocks systematically produces a higher total return than a different basket of stocks or a benchmark of a large number of them when calculated on a yearly basis. The first is a business analysis; the second is a statistical observation.

The idea of style-specific benchmarks appeared in this context. In 1984, the Frank Russell Company, an institutional and pension consulting firm, introduced the Russell 3000 and its large‑cap subset, the Russell 1000 benchmark. They did so because retirement plan and endowment sponsors were dissatisfied with the available benchmarks. The S&P 500 Index was too narrow for managers who invested outside its membership while the much larger Wilshire 5000 was full of illiquid, hard‑to‑track companies. Russell’s answer was a realistic, investable “whole market” universe of roughly 3,000 U.S. stocks and a carve out of the largest 1,000 as the natural arena for mainstream managers.

Once that size architecture was in place, splitting the Russell 1000 into style components in 1987 was a natural refinement to measure practitioners who had already long fallen into undefined, small-g growth and undefined small-v, value investing. Dividing the main benchmark into two market-cap equal sleeves based on price‑to‑book—the cheaper half of market cap as value, the other half as growth—worked well enough at the time because a low price‑to‑book ratio was in the mid‑1980s a tolerable shorthand for “cheap.” The large‑cap universe was still dominated by businesses whose book values roughly represented tangible assets, not just capitalized software development and intangible capital. If you were a value investor in 1987, the new value benchmark was far from a perfect mirror of your opportunity set, but it was close enough to be used. It offered a single metric, two sleeves, and a tidy 50/50 partition of the parent index by market cap. Importantly, at the time, the market was not 1/3 concentrated in a handful of individual names, so each of the two indices could have roughly similar n-counts to get to their respective 50% market cap of the overall Russell 1000 index.

The Russell 1000 Value and Russell 1000 Growth (and their smaller-sized cousins) have become, in the subsequent decades, the leading measurement tools for institutional investors and their gatekeepers. Although there are now other vendors of value and growth indices (defined as they see fit), Russell dominates the market for measuring institutional portfolios run according to these two styles. One decade ago, a Russell press release claimed that 99%--yes, 99%--of institutional equity portfolios managed to the growth or value style used a Russell benchmark. (“U.S. Equity Indexes: Institutional Benchmark Survey,” Russell Research, Russell Investments, January 2015, https://www.lseg.com/content/dam/ftse-russell/en_us/documents/research/russell-pure-style-indexes.pdf ) While that dominance may have diminished in the past decade somewhat, the R1V and R1G still rule the investing roost.

Despite that great success, almost everything that has occurred since has diminished the utility of this definition of value and this mechanism of measuring it. In particular, the U.S. economy’s migration from smokestacks to code necessarily diminished book value as a credible measure of growth, value, or much of anything else. The benchmark’s designers responded to the emerging challenge in the mid 1990s by adding two additional measures: a stock’s long-term (5 year) earnings growth estimate, and a trailing per share sales figure. These three rankings then formed a composite score used to allocate stocks—they had ceased being businesses—into the style benchmarks based on their scores.

The book-to-price value metric has remained unchanged. That means it has become doubly diluted over time, first by the declining relevance of book value, and second, by the addition of two additional factors that say little if anything about the genuine attraction of an investment from a valuation perspective. (The earnings growth metric was changed from 5 years to 2 in 2011 due to data issues.)

The consequences for investors of these choices made in the 1980s and 1990s were manageable so long as the benchmarks were “bigger” than the market, in the sense of not driven by a handful of constituents. But in the last decade, that math has been reversed. When only a few companies represent an outsized share of the benchmark capitalization, and when their valuations swing dramatically, any style definition that must assign roughly half of benchmark market cap to “value” and half to “growth” cannot but distort the meaning of those words, not to mention make a hash of any reasonable effort to measure investor outcomes.

The mismatch between definition and calculation has become all too apparent the last few years. We saw an early version of it in the summer 2025 reconstitution, when Amazon, Alphabet and Meta all entered the value index with a combined weight of more than five percent, while their combined weight in the growth sleeve declined only modestly. The rules were working as designed: according to the measurement metrics, these firms had slid away from their prior growth characterization, and the style score reclassified a slice of their capitalization as “value.” But one does not have to be a Graham disciple to see the disconnect between that particular logic and any intuitive notion of value investing, defining the value opportunity set, and measuring investment outcomes.

The just-completed 2026 reconstitution raised the stakes, with massive style shifts as mega‑caps ping‑ponged between the growth and value sleeves. High-profile, world-beating tech stocks are now regularly pushed into the value bucket, not because a Graham‑style analyst would necessarily recognize them as bargains (though they might be!), but because the scoring algorithm must find enough “value” to fill a quota in a universe dominated (in terms of market cap) by an unusually small number of issues. This isn’t drift or inconsistency as much as it is oscillation due to a dated methodology at odds with today’s market. The resulting churn in style exposure is now so large that it is visible not just in attribution reports, but in flows and trading volumes around the annual rebalance.

The widening gap in n-count (the number of securities in each sleeve) in recent years shows the extent of the pendulum swing. At the next major market correction, the gap will narrow, but this flaw is a feature, not a bug, of a benchmark that must be 50/50 in terms of market cap at each annual reconstitution.

To be fair, Russell is in the business of maintaining a complete, consistent map of the U.S. equity market. It necessarily adjusts that map to reflect the changing geography of capitalization. Russell’s ground rules state plainly how each stock’s market cap is allocated between growth and value so that, in the aggregate, the two style indices sum exactly to the Russell 1000. The composite style score is disclosed. The reconstitution timetable is known in advance. There is no concealment here. The Russell benchmarks do exactly what it says on the tin.

This is where the historian of the stock market asks whether the current rules, created in a prior time and place, are still fit for purpose now. And here the answer is far from clear. On one hand, Russell does not guard Value Valhalla. Not its job. It is not a named trustee for the intellectual work of Ben Graham, Barr Rosenberg, Eugene Fama, Ken French or anyone else. It does not take direction from Warren Buffett or Seth Klarman or a handful of other superstars who get to define value investing however they see fit.

But while there remains some coherence in both the definition and popular understanding of the growth sleeve (through the admittedly very limited prism of the two growth-oriented composition measures), the value half of the benchmark appears to be a residual of stocks that just don’t score as growth, rather than as an independently defined standard for bargain hunters. Once the constituents of the value sleeve become, as a practical matter, anything that is not in the growth one, it ceases to be a positive description of an investment style or a means of measuring said approach. The result has been several decades of intellectual muddle, definitional complexity, and, not coincidentally, real-world investor frustration as the market and the measure drifted in different directions.

The prior outperformance of low price-to-book stocks has been hard to spot for many years now and has made portfolios committed to investing as defined by this benchmark look pointless. Market observers and value-oriented investors will have their explanation about how this state of affairs came about. Many will point to the evolution of the economy, the changing nature of investing, the glut of investment information, and quantitative investing itself. As one grizzled value manager commented to me (about the style box system), “I think this basically ruined investment creativity and the value seeking opportunities of investors. It can be and has been very challenging to stay in a Value box when you have a different view of “Value” than the guardians do. I understand that there is a need for a benchmark or reference index, but you miss out on opportunities just to stay in a box. That doesn’t seem right for clients.”

I would also point to the observer paradox. After forty years of talking about, slicing and dicing, and allocating capital directly into something called a value benchmark—through index funds, ETFs, consultant screens, and manager evaluation systems—it would be surprising if the original premium remained untouched. Once a metric becomes the target of large, rules-based flows and annual trading around a well-publicized rebalance, it ceases to be a neutral measure of outcomes. It becomes part of the mechanism shaping prices. The overlap of observer and participant is perhaps more significant in regard to the S&P 500 Index, a phenomenon I’ve taken up elsewhere ( https://americanaffairsjournal.org/2025/08/dr-frankensteins-benchmark-the-sp-500-index-and-the-observer-paradox/ ), but it can also be seen in the style indices.

I would also assert a large degree of original sin in the semantic and practical juxtaposition of value versus growth. Yes, investors answering to those names existed prior to the creation of dedicated style indices, and, yes, they appeared to be juxtaposed to one another. But do these two exercises really belong on a single, linear calculation spectrum? I don’t think so.

Being in the bottom half of a single value-growth continuum doesn’t necessarily make a stock a candidate for “cheap” or “discounted” or “bargain status. Not in 1987 and not now. And being in the top half of the same ranking doesn’t necessarily mean that a stock is poised for expansion ahead. The former is about current valuation; the latter is about business prospects. In mixing a valuation measure (book to price), and a growth measure (sales per share), the benchmark conflates two separate categories of analysis. The third, expected EPS growth, is about investor expectations, and comes from a less than neutral source. That metric only muddies the waters. This mix of measures may have made some sense in the 1980s, but decades later?

In short, traditional benchmark-oriented value investing is a mess. But the issue is not just poor total return in an extended period of technology-driven productivity gains and in an environment of (until recently) declining interest rates. It is how this exercise in measurement came to be defined in the 1980s and how it has aged since. Forty years is a long time for any taxonomy to go unchallenged. It is time to refresh how we define and measure investment communities. (Russell and other index providers have in recent decades introduced “pure” style indices where securities must fall into one or another camp, but they are not, as of yet, widely used in the marketplace.)

From my perspective, two changes above all are required. First, the styles deserve distinct treatment so that investors who seek bargains and those who seek high growth companies are not forced, as they are now, to share a combined measurement framework based on a composite score that doesn’t really say much about either current valuation or the serious future growth prospects of the companies being measured. A value benchmark should be structured around valuation, with a continuum running from cheap to expensive. A growth benchmark should be structured around business outlook, with a continuum running from stagnant to explosive. While it was once innovative and convenient to measure these two separate communities with a single yardstick, that is no longer the case.

Second, the tidy solution of having 50% of the large cap universe be characterized as value and 50% as growth might have been clever in 1987, an innocuous way to validate the Russell 1000 as the large cap institutional benchmark, but now it creates unnecessary chaos. That constraint needs to be abandoned. Well-defined value and growth benchmarks shouldn’t have to split the market’s value evenly. That’s an artificial constraint, not a requirement. The large-cap equity universe is not limited to the value-growth binary. There are investors focused on current cash returns, on dividend durability and growth, on balance-sheet conservatism, on governance and capital allocation, and on a range of other business characteristics that do not fit naturally onto a single line running from “value” to “growth.” (A Dividend Investor in a Stock Market belongs in that broader field, not as a subset of value by default, but as a distinct discipline with its own objectives, opportunity sets, and standards of success.)

Freed from these dated assumptions, self-defined value investors can chime in on how they would have Russell define the opportunity set and measure investor outcomes for those seeking bargains, discounts, and potential good values. Straight up valuation metrics relevant for the economy as it is now, not a half-century ago, would be a good start. And echoing Ben Graham’s original instinct about getting a dollar for 80 cents due to a temporary dislocation, some mechanism for value realization is a must.

Second, any honest reformulation of measuring value in the present environment should contend seriously with cash returns. Graham’s original intrinsic value calculation placed considerable weight on dividends and real distributable earnings. Instead, during the past forty years, buybacks have supplanted dividends, and capital gains have become the dominant currency of equity return. Value investors can determine the proper role of the buyback phenomenon in a new system of measurement.

Third, and perhaps most radically, I continue to question total return as the right measure for success in today’s the stock market. Like benchmarks, the concept and measure of total return emerged at a particular place and post-war time to meet a pressing need. But times have changed and it is appropriate to ask whether the approach is still fit for purpose. In a prior piece, I suggest breaking the measurement into a cash component (income payments and realized capital gains/losses) and contingent returns (unrealized capital gains/losses). (https://danielxperis.substack.com/p/when-total-return-isnt). While index providers are unlikely to endorse such a radical step, having total return defined almost entirely by contingent, theoretical gains—as it is today—cannot pass unquestioned.

Path dependency and inertia are obvious obstacles to this long-overdue revision. As we approach the centenary of Security Analysis, and forty years since professional investors were shoe-horned into a one-size-fits-all measurement system, it’s time to revisit how we as investors define our goals and measure our outcomes

----

Comments, criticisms, and suggestions welcome. This is a work in progress, not a completed piece. It is part of a broader, on-going assessment of definitions, frameworks, and measurements widely used in the US stock market, including MPT (Getting Back to Business, 2018), declining interest rates (The Ownership Dividend, 2024), total return (Substack, 2024), and the S&P 500 Index (American Affairs, 2025). The theme running through all the pieces is that much of the financial architecture we take for granted was designed in the post-war decades and is aging. There is much to review.