As an introduction to the VerifiedBeta framework, this comparison looks at IJS and DFSVX, with DFSV treated as the ETF wrapper for the same underlying DFA small-value implementation. The key question is which fund captures the small-value sleeve more efficiently once returns are normalized for risk, implementation drag, and valuation.
The practical answer is that both funds capture the intended style, but over the study period shown here the DFA implementation comes through as the cleaner factor harvester. The gap is not driven by a single factor. It comes from a slightly stronger aggregate mix of market, size, value, profitability, and residual alpha.
Comparison Table
| Metric | IJS | DFSVX / DFSV proxy |
|---|---|---|
| Fund Score | 85 | 91 |
| Backtested Relative Sharpe | 1.17 | 1.26 |
| Small Factor Score | 94.7 | 97.2 |
| Value Factor Score | 85.6 | 100.0 |
| Profitability Factor Score | 56.5 | 29.5 |
| Investment Factor Score | 0.0 | 35.9 |
| Key takeaway | Strong small and value capture, simpler implementation | Broader factor mix and better overall expected risk-adjusted outcome |
Interpretation
IJS still looks like a legitimate small-value fund. The issue is not that it fails to capture the style. The issue is that the DFA implementation captures slightly more of the return-enhancing mix once the full factor decomposition is laid out. That shows up in the higher fund score and the stronger backtested relative Sharpe.
This is also a good example of why VerifiedBeta separates factor purity from fund score. A fund can have strong exposure to one or two desired factors without necessarily ranking as highly on the total expected package after fees, implementation drag, and volatility are incorporated.
Factor Contribution Waterfalls
The figures below show the factor decomposition used in this comparison. They reflect the study periods noted in each chart.
Conclusion
The central point is straightforward: implementation quality matters. Two funds can sit in the same style box and still deliver meaningfully different expected outcomes once factor strength, volatility, and implementation drag are normalized into a single framework.