IS/OOS 2016-01-01, sqrt(365). Independence = corr to EqMom and to market. Real new edge = positive Sharpe AND low corr.
A. Factor/style ETFs as standalone time-series momentum
| ETF | Factor | OOS Sharpe | p | Corr EqMom | Corr Mkt |
|---|---|---|---|---|---|
| MTUM | Momentum | 0.60 | 0.055 | 0.80 | 0.62 |
| QUAL | Quality | 0.74 | 0.022 | 0.71 | 0.70 |
| VLUE | Value | 0.75 | 0.021 | 0.69 | 0.69 |
| USMV | MinVol | 0.48 | 0.097 | 0.61 | 0.75 |
| SLYV | SmValue | 0.62 | 0.048 | 0.55 | 0.52 |
| SLYG | SmGrowth | 0.63 | 0.045 | 0.64 | 0.58 |
| IWD | R1000Val | 0.76 | 0.021 | 0.60 | 0.58 |
| IWF | R1000Gro | 1.01 | 0.003 | 0.72 | 0.63 |
| VTV | LValue | 0.85 | 0.011 | 0.60 | 0.60 |
| VUG | LGrowth | 1.06 | 0.002 | 0.74 | 0.65 |
| VBR | VSmlValue | 0.64 | 0.040 | 0.61 | 0.59 |
B. Global equity cross-sectional momentum (top-2 of 7)
| Factor | OOS Sharpe | p | Corr EqMom | Corr Mkt |
|---|---|---|---|---|
| GlobalXS(2) | 0.72 | 0.028 | 0.79 | 0.85 |
C. Global equity per-region TSMOM (standalone)
| Region | OOS Sharpe | p | Corr Mkt |
|---|---|---|---|
| EFA (Dev ex-US) | 0.59 | 0.060 | 0.48 |
| EEM (EM) | 0.55 | 0.068 | 0.44 |
| VGK (Europe) | 0.51 | 0.086 | 0.52 |
| EWJ (Japan) | 0.66 | 0.040 | 0.35 |
| VWO (EM(V)) | 0.41 | 0.134 | 0.49 |
| VPL (Pac exJP) | 0.51 | 0.085 | 0.47 |
D. Volatility risk premium (VRP) timing on SVXY
VRP = VIX - 21d realized vol of market; long SVXY (short vol) when VRP>0 (contango), else flat.
| Factor | OOS Sharpe | p | Corr EqMom | Corr Mkt | maxDD |
|---|---|---|---|---|---|
| VRP-SVXY | 0.50 | 0.094 | 0.58 | 0.63 | -95.2% |
Reading
- Factor ETFs (quality/value/momentum/minvol) are equity factors -> all show HIGH corr to market (0.52-0.75) and to EqMom (0.55-0.80). They are beta, confirming the single-name rejections: fundamental factors as standalone = beta, not independent edges.
- Global equity cross-sectional momentum: Sharpe 0.72 but corr 0.79/0.85 -> global beta (all equity markets co-move). Equal-weight of 6 regional TSMOM gives Sharpe 0.61 but corr to US market still 0.54 -> only a PARTIAL diversifier, still equity beta, and inferior to Treasuries (negative corr).
- VRP/SVXY: Sharpe 0.50 (p=0.094, marginal), corr 0.58-0.63 (risk-on beta), and maxDD -95% (blew up in 2018/2020 vol spikes) -> REJECTED.
Verdict — ROUND 5: NO new independent edge
Across factor ETFs, global equity, and the vol complex, everything equity-related is beta (high correlation) and the one vol trade has a catastrophic tail. The free yfinance universe (equities/ETFs/bonds/commodities/ FX/vol) is now EXHAUSTED for independent edges. The ONLY genuine non-beta diversifier we have found remains Treasuries (Round 4, negative equity corr). Global equity is at best a weak partial diversifier (corr ~0.5).
To find truly new, independent edges we would need data NOT obtainable free:
- point-in-time fundamentals (earnings, book value, quality time series)
- options-implied surfaces (real variance-risk-premium via VIX futures, not decaying ETNs)
- rates-futures term structure, FX carry (interest-rate differentials)
- sentiment / flows / short-interest, on-chain crypto metrics. Until then, the diversified book = equity momentum + crypto gate60 + Treasury diversifier is the best achievable with available data.