The Weekly Print
1. Market Regime Snapshot
VIX Term Structure
| Tenor | Level | vs. Spot |
|---|---|---|
| VIX Spot | 18.8 | β |
| VIX 3M | 20.5 | +1.8 (Contango) |
Realized vs. Implied Vol: SPY 20D RV = 12.2%, VIX = 18.8. Spread = -6.5pp (realized vol below implied β calmer than feared).
Cross-Asset Momentum (1-Month)
| Asset | 1M Return | Signal |
|---|---|---|
| SPY (Equity) | +0.6% | π Positive |
| TLT (Bonds) | -1.7% | π Negative |
| GLD (Gold) | -5.2% | π Negative |
| UUP (US Dollar) | +0.5% | π Positive |
Regime: Mixed/transitional regime
Last updated: 2026-07-19 17:42 UTC
2. Factor Performance Dashboard
Source: ETF Proxies (MTUM, VLUE, QUAL, USMV, IWM vs SPY)
Note: ETF proxy returns include market beta and are not directly comparable to factor-neutral French library returns.
| Factor | Weekly | 1M | 3M | Mean (52W wkly) | Std (52W wkly) | Z |
|---|---|---|---|---|---|---|
| Momentum π¨ | -6.12% | -7.35% | +11.67% | +0.50% | 2.71% | -2.44 |
| Value | -3.08% | -4.69% | +21.59% | +1.04% | 2.56% | -1.61 |
| Quality | -0.83% | +0.05% | +6.82% | +0.36% | 1.55% | -0.76 |
| Low Volatility | -1.19% | +0.72% | +2.66% | +0.11% | 1.16% | -1.12 |
| Size | +0.89% | +1.31% | +2.86% | +0.23% | 1.57% | +0.42 |
π¨ Factor Stress: Momentum (z=-2.44) β weekly return β₯ 2Ο from trailing mean.
Last updated: 2026-07-19 17:42 UTC
3. Macro Signal Tracker
| Indicator | Current | 1W Change | Signal |
|---|---|---|---|
| HY Spread (OAS) | 271bps | +1bps | β οΈ Widening |
| IG Spread (OAS) | 78bps | +2bps | β οΈ Widening |
| 2s10s Yield Curve | 0.37% | +0.02% | Normal |
| 3M10Y Yield Curve | 0.70% | -0.01% | Normal |
| Fed Funds Rate | 3.63% | N/A | β |
| HY β IG Spread | 193bps | β | Risk sentiment proxy |
Macro Summary: Neutral macro backdrop
Last updated: 2026-07-19 17:42 UTC
4. Quant Research Digest
Three papers I found worth reading this week:
SciPhy Reinforcement Learning for Portfolio Optimization β Igor Halperin et al. arXiv
The authors formulate continuous-time portfolio optimization with explicit transaction costs using a physics-informed reinforcement learning framework. Itβs an interesting approach to embedding continuous state constraints directly into stochastic control policies for allocation problems.
Measuring Sentiment News with Transformer-Based Language Models β Maria Saveria Mavillonio et al. arXiv
This paper benchmarks transformer-based daily news mood indices against standard dictionary word-count approaches. It highlights how contextual language models handle edge cases like negation and semantic nuance much better than rigid word-counting heuristics.
How Much of a 10-K Matters? Aggregation-Dependent Value of Full-Text versus Risk-Factor Sentiment β Sanggyu Sean Choi arXiv
This study tests whether full 10-K text or specific Item 1A risk-factor disclosures provide a better predictive signal for volatility versus returns. It shows that targeting volatility directly with risk text yields cleaner results than using text simply to forecast equity direction.
Last updated: 2026-07-19 17:42 UTC
5. Stat of the Week
| Stat | Value | Context |
|---|---|---|
| CBOE Skew Index | 147 | Elevated tail risk (>130) |
Momentum was the main outlier on the dashboard this week, dropping over 6% (-2.44 z-score). Meanwhile, the Skew Index rose back to 147 even as 20-day realized volatility stayed low at 12.2%. Itβs a clean example of market divergence: headline index volatility looks calm, but factor-level drawdowns and option pricing point to lingering left-tail risk.
Last updated: 2026-07-19 17:42 UTC
Generated: 2026-07-19 17:42 UTC