⭐️ Week 18, 2025: Why the S&P’s 9-Day Streak Doesn’t Mean Smooth Sailing
No fluff—just the must-know Fed, data releases, and sector calls
To Smart Investors,
The Most Important Stuff:
Jack’s notes for this week:
The Stock Arbitrage Scanner is coming out of alpha this week and will be available to all paying and lifetime users.
It will be suitable for both short-, mid- and long-term traders. There will be options too including LEAPs. Everything under 1 roof.
The wait will be worth it.
I will explain more today, including the new Roshi Ratio that I have been working on for months, in today’s Weekly Report.
We will have beautiful, interactive charts:
And simple answers for people who don’t want to delve too deep into numbers, candles, and charts:
Lifetime subscriptions are back and, hopefully, here to stay.
A discovery:
Roshi Ratio “Secret Sauce”
The table below compares the performance of several Roshi Ratio vs. popular SOTA indicators:
Table: Backtest performance of different Roshi Ratio formulations (Top 10% stocks portfolio, 2005–2025). Risk-free rate averaged ~1.5% over the period. The market (benchmark) had ~10% CAGR and 0.60 Sharpe in the same period for reference.
As shown, the regression-based weighting delivered the highest return and Sharpe ratio, making it the top performer. It achieved about a 14.2% CAGR, significantly above the market, and a Sharpe of ~1.05, indicating strong risk-adjusted returns. The PCA-derived composite was next best, also outperforming Black-Scholes. The Black-Scholes version, while better than the broad market, had a lower Sharpe (~0.85) – it seems the lack of differentiation among metrics cost some performance. The volatility-adjusted scheme gave a minor improvement over equal weight, likely because it effectively reduced the redundancy of Sharpe/Sortino a touch.
Statistical significance: The differences in Sharpe (0.85 vs 1.05) are material over 2 decades – the regression-based composite’s excess Sharpe indicates a considerably more efficient strategy. A long-short spread (top decile minus bottom decile by Roshi score) showed an information ratio above 0.8 for the regression version, versus ~0.5 for equal weight, again underscoring improvement in stock selection efficacy.
By sectors: I found the Roshi Ratio worked across most sectors, but with some nuance:
Sectors with inherently lower ROIIC (e.g. Utilities, Materials) still saw the composite picking winners, but the ROIIC component tended to favor companies that were above-average within their sector. In extremely capital-intensive sectors, ROIIC differences were smaller, so the composite leaned more on the risk-return metrics in those cases.
Technology and Healthcare, where ROIIC variance is high, benefited strongly – many of the top composite picks in Tech were those with both high ROIIC and high Sharpe (e.g. profitable growth companies), which delivered excellent returns.
No single sector dominated the top picks – the composite naturally found a mix of sectors (a positive sign that it’s not implicitly just a sector bet). I did observe that in certain years, the top Roshi stocks were concentrated in a booming sector (like Tech in 2020), but the composite helped avoid some frothy cases by requiring strong risk-adjusted past performance as well (which bubble stocks often lack once volatility picks up).
By market cap: The strategy showed robustness across large and mid-caps. Among small-caps, the efficacy was present but slightly reduced – small companies with extremely high ROIIC or past Sharpe often faced liquidity or volatility issues not fully captured by these metrics. If small-cap coverage is desired, one might include an additional penalty for very high volatility or low liquidity. But within the scope of more liquid stocks, the Roshi Ratio held up well.
In summary, the backtest confirmed that combining fundamental quality with risk-adjusted return metrics yields a powerful predictive indicator. The regression-optimized Roshi Ratio in particular consistently identified stocks that delivered superior subsequent returns for the risk taken. Notably, this composite outperformed any single-factor strategy (e.g., just picking highest Sharpe or highest ROIIC stocks), which underscores the benefit of a multifaceted approach. This aligns with the principle that “risk matters as much as return” and combining them into one score allows easy comparison of investments.
My Stock Arbitrage Scanner predicted the market crash 2 weeks before it happened:
That is exactly what I was doing at Knight Capital Group.
As always, no discounts. The price only goes up as I release new tools and eBooks.
All prior memberships include new SaaS tools (an options scanner, an undervalued stocks detector based on fundamentals, and another secret tool), 12 eBooks, and more. These will be released over the next 12 months. The books are already written and are being edited by a NY Times Best Seller editor.
Listen to today’s free weekly Podcast here:
Stuff We Published Last Week:
As Always In The Sunday Report:
I have written a detailed recap of last week’s market, my predictions for next week, and an ELI5 (Explain Me Like I’m 5).
You can also find my typical quant data and the stock insiders’ significant buys/sells with my interpretation.
Every day, I post summaries of news relevant to Investors. I try to post about 30 minutes before the markets open and cover the last 24 hours of news. On the weekends, I post in the afternoon.
This daily newsletter is read by decision-makers at companies ranging from Berkshire Hathaway, Citadel, and BlackRock, Inc. employees to the oil traders in the Emirates and Saudi Arabia and back to Alphabet in Mountain View, California, Apple, and NVIDIA.
These news are instrumental to my success.
$5/month — less than the cost of a Starbucks latte.
We recently crossed 1,000 paying subscribers 🙏🙏🙏
S&P500 Heatmap over the last week
This graph was generated by our friends at TradingView. Did you know you can get 60% off the Premium plan forever, even though it's not Black Friday?
SPY Analysis
Our friends at TradingView generated these graphs. Did you know you can get 60% off the Premium plan forever, even though it's not Black Friday?
*Also, the chart above shows precisely why you need the Premium TradingView plan in your life. Plus, being able to monitor 8 graphs simultaneously while running complex indicators on them in parallel. Look into this:
SPY Daily Structure · 2 May 2025
Price Action
4 Apr capitulation low ≈ 518 reversed on 104 M shares; price has rallied ≈ 9 % to 566.76 with eight higher closes while volume decayed to ≈ 61 M.
Rally now retesting March supply shelf at 570-580. Failure to print higher high in today’s session noted.
Key Levels
Resistance: 570-580 (supply shelf), 600-610 (March gap zone), 620+ (all-time extension).
Support: 548-550 (20-DMA / breakout pivot), 515-525 (early-April capitulation demand), 480 (secondary demand pocket).
Macro Paradox Read
Green = 71.6, Red = 64.5. Spread positive but compressing; green slope negative, red slope positive. Historical sample: spread compression with this configuration precedes stall or pullback inside 2-7 sessions 70-80 % of the time.
Forward Probability Map · 2-3 Week Horizon
Mean-reversion to 545-550 – 65 %. Macro signal compression + waning volume favors pullback to first support.
Low-energy grind through 580 toward 600-610 – 20 %. Requires close > 580 on volume > 80 M.
Sharp retest of 515-525 – 15 %. Triggered by daily close < 548 with red line crossing above green.
Tactical Bias
Short-term risk skewed lower/sideways; durable upside only on decisive close > 580 confirmed by volume expansion.
Disclaimer: This analysis is purely educational and not financial advice. Always consider multiple factors and your own risk tolerance before making trading decisions.
*Macro Paradox is available for free here
Please send feedback and ideas using comments, PMs, or email. I answer all emails and PMs personally. There is no personal assistant BS here.
And, as always—stay informed—and do your own due diligence.
With every good wish, I remain
Yours sincerely in Christ,
Jack Roshi
Applied Mathematics Department, MIT
Lead Quant and Board Member, Alpha LLPOpinions are my own
Executive Summary of Last Week:
Market Grind Continues: S&P 500 notches a nine-day win streak through May 2, its best since 2004, despite GDP contracting 0.3% in Q1 (April 30).
Earnings Fireworks: Mega-caps (Microsoft, Meta) beat—and sometimes boost—futures, but Apple and Amazon delivered muted reactions or cautious guidance (May 1).
Tariffs & Trade Tension: U.S.–China tariff jitters ebb and flow, yet front-running imports distorted Q1 GDP and leave consumer sentiment near five-year lows (April 28–May 2).
Labor Market Resilience: April nonfarm payrolls added 177K jobs, unemployment steady at 4.2%, suggesting Fed rate cuts might not come soon (May 2).
Sector Divergence: Tech and industrials led (+4.0% and +4.3% weekly), energy alone in red (-0.6%). Oil prices plunged 18% in April, gold up 22% YTD, bitcoin flirting with $100K.
Please subscribe to my Substack to access the premium report, which includes a comprehensive day‐by‐day breakdown, in‐depth weekly analysis, a list of the most shorted stocks, Excel spreadsheets of options flow, the most active stocks, projections for the next week, and insider trades and their interpretation.
Members-Only Deep Dive:
Keep reading with a 7-day free trial
Subscribe to The Stock Insider to keep reading this post and get 7 days of free access to the full post archives.