Strategies · Long-term investing
Value investing as a strategy: the rules and a 26-year test
Value as a rule set: how an index picks cheap stocks, and IWD against SPY from 2000 to 2026, with seven strong years, thirteen weak ones and a near tie overall.
On this page
Value investing, as a strategy that can be written down, is simple: buy the stocks that are cheapest against what their companies own or earn, and replace them as they stop being cheap. It is the oldest systematic idea in stock picking and one of the most studied. It is also the one whose record in this century has disappointed the most people.
Growth vs value investing: styles, P/E and the long record explains the two styles and the ratios behind them, and Benjamin Graham and value investing: margin of safety explained covers where the idea came from. This lesson treats value as a rule set and tests it: the largest US value index fund with a long history, IWD, against SPY from June 2000 to October 2026. The 26 years divide into three stretches so different that the average of them describes none.
The idea
A stock's price is a forecast. A high price against today's profits says investors expect growth; a low one says they expect trouble. The value idea is that these forecasts are too extreme on both sides. Companies priced for trouble turn out, on average, less bad than feared, and companies priced for brilliance turn out less brilliant.
Fama and French measured it in 1992. Sorting US stocks from 1963 to 1990 by the ratio of book value to market price, they found that the cheapest group had earned far more than the most expensive, and that this ratio explained returns better than beta did. Two explanations have been argued over ever since. One says cheap stocks are riskier, so the extra return is payment for risk. The other, from Lakonishok, Shleifer and Vishny, says investors project recent bad news too far forward, and the extra return is the correction of that mistake.
The distinction matters for what to expect. A payment for risk should persist, and should come with bad stretches. A mistake can be learned from and traded away. In 2021 Fama and French revisited their own finding: the value premium in US stocks from 1991 to 2019 was much smaller than from 1963 to 1991, though with so much noise that they could not rule out that nothing had changed.
The rules
These are the rules of the Russell 1000 Value Index, which IWD tracks, in outline.
- Take the 1,000 largest US companies by market value.
- Score each on three measures: book value against price, analysts' forecast of earnings growth over the next two years, and growth in sales per share over the last five.
- Companies that are cheap on book value with low growth go into the value index. The opposite go into the growth index. Companies in between are split, part of their weight in each.
- Weight each holding by its market value in the value index.
- Rebuild the list at each scheduled reconstitution: once a year in June until 2025, and in June and December from 2026.
- Reinvest dividends.
A person holding the fund follows one rule: buy it and keep it. Steps 1 to 5 are done by the index.
Two consequences follow from the rules and explain most of the test. First, this is "cheap half of the market", not "the cheapest stocks": about half the market's value is in the index, so its results stay close to the market's. Second, cheapness on book value collects particular industries. Banks, insurers, energy companies and utilities own large balance sheets and trade at low multiples of them. Software companies own little that accounting counts. A value index is, in practice, heavy in the first group and light in the second.
What the backtest shows
The test, stated in full so it can be checked:
- Data. Month-end closes of IWD and SPY from June 2000, IWD's first full month, to October 2026: 317 months.
- Trading. Each fund is bought once and held.
- Costs. One purchase each. The funds' costs are inside their prices: IWD charges 0.18% a year and SPY 0.0945% (the issuers' fund pages, November 2026). The index's turnover at each rebuild is paid inside the fund as well.
- Left out. Dividends on both sides, and taxes. Prices are adjusted for splits only.
| Measure, price only | IWD | SPY |
|---|---|---|
| Total return | 371.9% | 420.2% |
| Return per year (CAGR) | 6.1% | 6.5% |
| Deepest fall from a peak | −57.4% | −52.2% |
| Longest time below a peak | 76 months | 80 months |
| Value of $1 at the end | $4.72 | $5.20 |
Over the whole period the two finished close together, with value slightly behind on price and with a deeper worst fall.
What the missing dividends change
Value stocks pay more of their profits out. IWD's 12-month trailing yield was 1.37% on October 31, 2026 (iShares) and SPY's 1.04% on November 17, 2026 (State Street), a gap of about a third of a point; at other times in the period it has been wider. The price gap in the table is 0.4 points a year. Adding dividends back would therefore remove much of it, possibly all. This page cannot compute the total return exactly, so the careful statement is that over these 26 years value and the market ended about level, with value possibly a little ahead, and that nothing here supports a claim of a premium.
The correction does not change the shape of the three stretches below. A yield gap of under a point cannot explain differences of four to six points a year.
Three stretches
| Period | IWD, per year | SPY, per year | Difference |
|---|---|---|---|
| June 2000 – May 2007 | 7.3% | 0.8% | +6.5 |
| May 2007 – August 2020 | 2.4% | 6.4% | −4.0 |
| August 2020 – October 2026 | 13.0% | 13.3% | −0.3 |
2000 to 2007. The test begins three months after the peak of the dot-com bubble. Technology stocks, which a value index barely held, fell by more than half. From August 2000 to September 2002 SPY's price fell 46.3% and IWD's 27.0%. By March 2007 a dollar in IWD was worth 1.55 times a dollar in SPY.
2007 to 2020. Then the industries value was heavy in went wrong in turn: banks in 2008, energy from 2014, and both again in 2020. Meanwhile the companies it was light in grew into the largest in the market. IWD's price divided by SPY's fell 48.8% from its March 2007 high to its low in October 2025.
Since 2020. Roughly level, with wide swings each year, and a recovery of value's relative price from 0.795 to 0.907 in the twelve months to October 2026.
Which of these is "the" result of value investing depends entirely on the starting date. A test that began in 2007 would show a failed strategy. One that ended in 2007 would show a brilliant one. Over every five-year window in the data, IWD's price was ahead of SPY's in 46 of 257, or 18%.
Other funds on the same idea
| Fund | Index | From | Fund, per year | SPY, same months | Deepest fall |
|---|---|---|---|---|---|
| IWD | Russell 1000 Value | June 2000 | 6.1% | 6.5% | −57.4% |
| IVE | S&P 500 Value | June 2000 | 5.5% | 6.5% | −58.8% |
| VTV | CRSP US Large Value | February 2004 | 6.9% | 8.7% | −57.1% |
| IWF | Russell 1000 Growth | June 2000 | 6.6% | 6.5% | −64.3% |
Three index providers define value with different measures and reach the same place. The growth index, shown for contrast, ended level with the market after falling 64.3% on the way: it lost 5.1% a year to May 2007 and then gained 10.5% a year to August 2020.
Where it fails: the losing periods
Thirteen years, May 2007 to August 2020. IWD's price gained 36.5% in total while SPY's gained 127.8%. This is the central fact about value in this century, and it lasted longer than most investors' patience and many fund managers' careers.
The financial crisis. Cheap did not mean safe. From October 2007 to February 2009 IWD fell 55.8% and SPY 52.2%. The index was heavy in banks, which looked cheapest on book value just before their book values were written down.
2020. From December 2019 to March 2020 IWD fell 27.3% against 19.9%, and over the calendar year it gained 0.2% while SPY gained 16.2%.
2023 and 2024. From December 2022 to December 2024 SPY's price rose 53.3% and IWD's 22.1%, as a handful of very large technology companies led the market.
| Year | IWD | SPY |
|---|---|---|
| 2007 | −3.0% | 3.2% |
| 2015 | −6.3% | −0.8% |
| 2017 | 11.0% | 19.4% |
| 2020 | 0.2% | 16.2% |
| 2023 | 9.0% | 24.3% |
| 2024 | 12.0% | 23.3% |
Other ways it fails or misleads:
- Value traps. Some stocks are cheap because the business is shrinking. A rule based on book value buys them along with the bargains, and cannot tell which is which.
- Book value has aged. Research, software and brands are mostly expensed, not recorded as assets, so companies built on them look expensive on book value however profitable they are. Whether that has broken the measure or only delayed it is still argued.
- An index is not a deep-value portfolio. The academic results came from the cheapest tenth or third of stocks, often small ones, held against the most expensive. A fund holding the cheaper half of large companies captures a diluted version.
- The start date. Beginning in June 2000 gives value its best seven years first. This test is kinder to it than a start in 2007 or 2010 would be.
Compared with its neighbours
| Strategy | Selects on | Result against SPY's price here | Worst stretch |
|---|---|---|---|
| Value (IWD) | Low price to book, low growth | −0.4 points a year, 2000 to 2026 | 13 years behind, 2007 to 2020 |
| Growth (IWF) | The opposite | +0.1 points a year | −64.3% fall; 7 years behind to 2007 |
| Dividend growth | Years of dividend increases | −0.9 points a year, 2006 to 2026 | Behind since February 2009 |
| Market (SPY) | Nothing | n/a | −52.2% in 2007 to 2009 |
How Finabase does it
Finabase does not label stocks as value or growth and no Finabase tool says what to buy.
- Screener. The screener filters on P/E, price to sales, dividend yield, earnings per share and market capitalization, within scopes such as the S&P 500. It has no price-to-book field, so it cannot reproduce rule 2. The "Low P/E large caps" screen lists profitable companies above $10 billion trading under 15 times earnings: a different, earnings-based definition of cheap, and a starting list, not a portfolio.
- Ratios. Each stock's page shows its price to book and P/E, for checking one company against the idea.
- Comparison. The comparison chart linked above draws IWD against SPY as percentage change, on prices, like this page.
- Backtest and Strategy. These tools replay indicator signals and if/then rules on one symbol, buying a fixed $10,000 per signal with no trading cost; Strategy covers two years. They cannot pick stocks by valuation, and two years is far too short for a strategy whose weak stretch lasted thirteen.
Frequently asked questions
Does value investing still work?
On this test it neither worked nor failed: IWD's price rose 6.1% a year from 2000 to 2026 against 6.5% for SPY, and its higher dividends close much of that gap. But it led by 6.5 points a year until 2007 and trailed by 4.0 points a year for the next 13 years.
Why did value stocks underperform for so long?
Three things coincided. Banks, a large part of value indexes, were hit in 2008. Energy fell from 2014. And the companies value indexes held least, large technology firms, became the most profitable in the market. Whether that was bad luck or a lasting change is unsettled.
Is a value fund less risky than the market?
Not in this test. IWD fell 57.4% in 2007 to 2009 against 52.2% for SPY and fell more in early 2020. It fell less in 2000 to 2002 and in 2022. A low price does not limit how far a price can fall.
What is the difference between value investing and a value index fund?
Value investing in Graham's sense is judging one business at a time and buying well below what it is worth. A value index fund applies a formula to a thousand companies and holds the cheaper half. The fund is a test of the formula, not of the judgment.
Summary
As a rule set, value investing holds the cheaper, slower-growing half of the large-company market and rebuilds the list on a schedule. Tested through IWD from June 2000 to October 2026, on prices without dividends, it returned 6.1% a year against 6.5% for SPY, with a deeper worst fall. That average hides three stretches: far ahead to 2007, far behind for thirteen years to 2020, and level since. Value's higher dividends would close much of the overall gap, which this page cannot measure exactly, so the fair summary is a tie reached by a very different road. Two other value indexes show the same pattern. The test starts at the most favorable moment for value in a generation and still finds no premium, and it says nothing about the cheapest small stocks that the original research studied.
Check your understanding
4 questions on this lesson. Each answer comes with the reason, and nothing is scored.
IWD’s price returned 6.1% a year from 2000 to 2026 and SPY’s 6.5%. Why is that average a poor description of value investing over the period?
- Because the funds have different fees
- Because it is made of three very different stretches: far ahead to 2007, far behind to 2020, level since
- Because 26 years is too short to say anything
Show the answer
B. Because it is made of three very different stretches: far ahead to 2007, far behind to 2020, level since Value led by 6.5 points a year for seven years and trailed by 4.0 points a year for thirteen. Nobody lived through the average. A result that changes sign with the start date should be reported by period.
A value index weights toward stocks that are cheap against book value. Which industries does that rule collect?
- Software and other companies with few recorded assets
- Banks, insurers, energy companies and utilities, which own large balance sheets
- An even spread of all industries
Show the answer
B. Banks, insurers, energy companies and utilities, which own large balance sheets Book value counts physical and financial assets, not research or brands. So the cheap-on-book list is heavy in asset-rich industries, and the index’s results follow theirs: banks in 2008, energy after 2014.
From October 2007 to February 2009 IWD fell 55.8% and SPY 52.2%. What does that show about buying cheap stocks?
- A low price limits how far a stock can fall
- Cheapness is not protection: the cheapest-looking stocks were banks whose book values were about to be written down
- The index was calculated wrongly
Show the answer
B. Cheapness is not protection: the cheapest-looking stocks were banks whose book values were about to be written down A stock that looks cheap against book value is only cheap if the book value holds. In the financial crisis the measure itself was wrong for the industry the index held most.
The test leaves out dividends, and value funds yield more than the market. How does that change the reading of the 0.4-point price gap?
- It does not: both funds pay dividends
- The gap is about the size of the yield difference, so on total return the two were roughly level
- It proves value beat the market by a wide margin
Show the answer
B. The gap is about the size of the yield difference, so on total return the two were roughly level A yield gap of a third of a point or more cancels much or all of a 0.4-point price gap. That supports "roughly level", not a premium, and it cannot explain the four-to-six-point differences inside the three stretches.
Sources and further reading
- Eugene F. Fama and Kenneth R. French, “The Cross-Section of Expected Stock Returns,” Journal of Finance 47(2), 1992
- Josef Lakonishok, Andrei Shleifer and Robert W. Vishny, “Contrarian Investment, Extrapolation, and Risk,” Journal of Finance 49(5), 1994
- Eugene F. Fama and Kenneth R. French, “The Value Premium,” Review of Asset Pricing Studies 11(1), 2021
- FTSE Russell — Russell U.S. Equity Indexes: construction and methodology (style indexes)
- iShares — Russell 1000 Value ETF (IWD) fund page: expense ratio 0.18%, inception May 22, 2000, 12-month trailing yield 1.37% as of October 31, 2026
- State Street — SPDR S&P 500 ETF Trust (SPY), fund page: gross expense ratio 0.0945%, fund distribution yield 1.04% as of November 17, 2026
- Finabase chart data — IWD, IWF, VTV, IVE and SPY monthly bars to October 2026; results computed by Finabase
