Here is a question that will make most ETF investors uncomfortable: do you actually know what factor bets your portfolio is making? You might think you own "small-cap value" because you hold an ETF with those words in the name. But ETF labels are marketing, not science. The only way to know your true factor exposure is to run a regression.
That is exactly what FactorLens does. It is a free tool that runs a Fama-French 5-factor regression on any portfolio of US ETFs, and shows you the factor loadings, statistical significance, return attribution, and per-fund decomposition. No account required. No paywall. Just enter your tickers and weights.
What Is Factor Analysis (and Why Should You Care)?
In 1993, Eugene Fama and Kenneth French published research showing that stock returns are largely explained by a small number of systematic "factors." The original model had three: market risk (beta), company size (small vs. large), and valuation (cheap vs. expensive). In 2015, they expanded it to five factors by adding profitability (high vs. low profit firms) and investment (conservative vs. aggressive capital spenders).
Why does this matter for you? Because factor exposure is the primary driver of portfolio returns. If your portfolio has a large loading on the value factor, it will behave very differently from one tilted toward growth, even if both are labeled "diversified." Factor analysis strips away the marketing and tells you what is actually happening under the hood.
Think of it like a nutrition label for your portfolio. You would not eat a protein bar without checking the macros. You should not invest in a portfolio without checking the factor exposure.
The ETF Label Problem
ETF issuers have every incentive to use appealing names. "Small-Cap Value" sounds precise, but the actual methodology behind the index varies wildly between providers. Some use price-to-book as their value metric. Others use price-to-earnings, price-to-cash-flow, or a composite. Some apply aggressive screens, others barely tilt.
The result is that two ETFs with nearly identical names can have very different factor profiles. One "small-cap value" fund might deliver a genuine, statistically significant value loading of 0.6. Another might come in at 0.15, which is barely distinguishable from a broad market fund.
You cannot tell the difference by reading the fund name, the prospectus summary, or even the top holdings list. You need to run the regression.
How FactorLens Works
Using the tool is straightforward:
- Enter your tickers and weights. Add up to 25 US-listed ETFs with their portfolio weights (in percent). Or pick one of the pre-built example portfolios to get started quickly.
- Hit Analyze. The tool fetches historical monthly returns for each ETF and runs an OLS regression against the Fama-French 5-factor model.
- Read the results. You get a factor scorecard showing each loading with its t-statistic and significance, a return attribution chart breaking down where your returns came from, and a per-fund decomposition so you can see which ETFs are contributing which factor exposures.
The regression equation is the standard one from academic finance:
Rp - Rf = α + β1(MKT-RF) + β2(SMB) + β3(HML) + β4(RMW) + β5(CMA) + ε
Factor data comes from the Ken French Data Library, updated monthly. This is the same data source used in thousands of academic papers. It is the gold standard.
Example: Small-Cap Value Showdown (AVUV vs. VBR vs. VIOV)
Let us put this to work. Three popular small-cap value ETFs:
- AVUV (Avantis US Small Cap Value) - the Bogleheads favorite
- VBR (Vanguard Small-Cap Value) - the low-cost classic
- VIOV (Vanguard S&P Small-Cap 600 Value) - the index purist choice
All three say "small-cap value" on the tin. But run them through FactorLens and the differences are striking. AVUV typically shows the highest SMB (size) and HML (value) loadings of the three, often significantly so. This is because Avantis uses a multi-dimensional screening approach that deliberately targets small, cheap, profitable stocks.
VBR, despite its name, often has a more moderate value loading because the CRSP index it tracks uses a relatively broad definition of "value." VIOV tends to land somewhere in between, with a tighter universe from the S&P 600 index.
The practical takeaway: if you want maximum small-cap value exposure per dollar invested, the regression tells you which fund actually delivers it. Names and expense ratios alone will not give you that answer.
Example: Is SCHD Really a Value Fund?
SCHD (Schwab US Dividend Equity) is one of the most popular ETFs among income-focused investors. Many holders think of it as a value fund because it selects high-dividend-yield stocks, and high yield is often associated with value.
Run SCHD through FactorLens and the picture is more nuanced. You will typically see a moderate HML (value) loading, but the standout factor is often RMW (profitability). That makes sense: SCHD screens for quality metrics like return on equity and debt-to-equity alongside dividend yield. The fund is arguably more of a "profitability fund" than a "value fund."
You will also often see a negative SMB loading, meaning SCHD tilts toward larger companies. This is not surprising for a dividend-focused fund (large companies pay more dividends), but it is worth knowing if you thought you were getting broad market-cap exposure.
None of this makes SCHD a bad fund. It just means you should understand what you are actually buying instead of relying on the label.
What the Numbers Mean
FactorLens shows several key metrics. Here is how to interpret them:
- Factor loadings are the regression coefficients. A market beta of 1.0 means your portfolio moves in lockstep with the market. An SMB loading of 0.3 means you have meaningful small-cap exposure. Loadings near zero mean the factor is not driving your returns.
- T-statistics tell you whether a loading is statistically significant. A |t-stat| above 2 generally means the loading is real, not noise. FactorLens highlights significant factors so you can focus on what matters.
- R-squared tells you how much of your portfolio's return variation is explained by the five factors. Most equity portfolios will show R-squared above 0.90, meaning the Fama-French model captures over 90% of the return variation. If R-squared is unusually low, your portfolio contains exposures the model does not capture (alternatives, options strategies, or very niche sectors).
- Alpha (α) is the annualized return not explained by any of the five factors. Positive alpha means your portfolio outperformed what the factor exposures alone would predict. But be careful: alpha is often statistically insignificant over short periods, so do not read too much into small positive or negative values.
- Return attribution breaks down your portfolio's annualized return into the contribution from each factor. This answers the question: "How much of my return came from market beta vs. my value tilt vs. my size bet?" It is one of the most useful views because it connects the abstract factor loadings to actual dollars.
Limitations and What Is Coming Next
FactorLens is a powerful tool, but it has boundaries:
- US tickers only. The tool currently supports US-listed ETFs and uses US Fama-French factors. International factor models (developed ex-US, emerging markets) are on the roadmap.
- Momentum factor coming soon. The current default is the 5-factor model. A 6-factor version including the momentum (UMD) factor is in development and will be available shortly.
- Historical, not forward-looking. Factor regressions tell you what your portfolio's exposures were over the sample period. Factor loadings can drift over time as fund managers change their methodology or as holdings turn over.
- The model is not perfect. The Fama-French model is the most widely used framework in academic finance, but it does not capture every source of return. Sector bets, currency exposure, and alternative risk premia fall outside the model.
Try FactorLens for Free
If you have ever wondered whether your portfolio's factor exposures match your intentions, there is now a quick way to find out. Head over to bestfolio.app/tools/factor-lens, enter your tickers and weights, and see exactly what you own in factor terms.
It takes about 30 seconds. No signup. No credit card. Just the regression.
And if the results surprise you (they often do), that is exactly the point. Better to know your true factor exposure now than to discover it during the next drawdown.