Research cutoff: Monday, August 24, 2026 at roughly 12:30 BST / 07:30 ET.
U.S. price cutoff: Friday, August 21 regular-session close.
Horizon: roughly the next 1–15 U.S. trading sessions.
Pre-open rule: extended-hours moves are context only. They do not count as confirmation for this screen.
Today's answer: Cadence is still the best transition, Riot is the most interesting new divergence, but neither is a clean opening-bell buy
Monday gives us a useful contrast between a laggard that has started repairing and a laggard whose gap just became enormous because the stock refused to follow its underlying driver.
Cadence Design Systems (CDNS) is still first. Friday was the first completed session in the recent sequence where Cadence rose in absolute terms and beat Synopsys across the 1-, 3- and 5-session windows while leaving a meaningful 20-session deficit. That is the shape this strategy is designed to find.
Riot Platforms (RIOT) is the new research watch. Bitcoin exposure ripped higher late last week — the iShares Bitcoin Trust (IBIT) gained more than 22% over five sessions — while Riot gained only about 4%. Friday was even more extreme: IBIT gained 6.0%, while RIOT fell 5.5% after having traded higher earlier in the day. The gap is real and Riot's latest operating update is not obviously broken, but that Friday reversal is negative confirmation, not a reason to buy more aggressively.
UPS remains the steadier secondary watch. Its 20-session deficit versus FedEx is still large, and the latest quarter included raised full-year revenue and EPS guidance, but the stock has not reclaimed the $106–$108 area that would show actual repair.
So there are three legitimate research candidates today, but no clean catch-up entry before the U.S. cash session opens.
Candidate summary
| Rank | Candidate | Genuine relationship | What the screen sees | Status | What would change the view |
|---|---|---|---|---|---|
| 1 | Cadence (CDNS) | Direct EDA peer Synopsys (SNPS) | Positive 1/3/5-session turn while a large 20-session deficit remains | Best transition watch | Another regular session holding roughly $316–$320; stronger confirmation through $325–$328 |
| 2 | Riot Platforms (RIOT) | Bitcoin mining + large BTC balance-sheet exposure versus IBIT / Bitcoin | Enormous multi-window lag, intact recent strategic news, but Friday sharply rejected the move | New divergence watch — not turned | Regular-session close back above roughly $21, then preferably $22.20 with positive relative performance versus Bitcoin |
| 3 | UPS | Direct parcel/logistics peer FedEx (FDX) | Small 3/5-session repair while 10/20-session deficits remain | Secondary watch | Reclaim roughly $106–$108, then hold into a second session while maintaining positive relative performance |
These are observation conditions, not market-open instructions.
Relative performance through Friday's close
The figures below are candidate return minus benchmark return, in percentage points. Negative means the candidate lagged.
| Candidate vs benchmark | 1 session | 3 sessions | 5 sessions | 10 sessions | 20 sessions |
|---|---|---|---|---|---|
| CDNS vs SNPS | +1.74 pp | +2.41 pp | +3.82 pp | -1.60 pp | -8.75 pp |
| RIOT vs IBIT | -11.50 pp | -14.48 pp | -18.28 pp | -22.06 pp | -32.15 pp |
| UPS vs FDX | -0.26 pp | +1.08 pp | +0.47 pp | -4.43 pp | -14.35 pp |
The table makes the important distinction obvious.
Cadence has a smaller old gap, but a real short-window turn. Riot has a spectacular gap, but no turn at all. In fact, Friday widened the gap dramatically.
That is why the largest negative number is not automatically the best trade.
Seven-session walk-forward: the filters still earn their keep
I re-ran the screen over the prior seven completed U.S. sessions — August 13, 14, 17, 18, 19, 20 and 21 — using only information that would have been available at each historical cutoff.
Because Monday's U.S. cash market has not opened yet, there is no new completed session to add since Sunday's report. The sample is therefore unchanged, and I am not going to invent a new rule simply to make Monday's article look different.
The useful result is that the current rules continue to separate relative arithmetic from actual price repair:
- CDNS showed several days where it improved relative to SNPS while Cadence itself was falling. Those days failed the absolute-stabilization rule. Friday was the first day that combined an absolute gain with positive 1/3/5-session relative performance.
- UPS improved over short windows but never reclaimed the $106–$108 repair zone. Friday then slipped relative to FedEx again.
- FTNT and FFIV previously produced attractive relative numbers mainly because cybersecurity peers were falling harder. The peer-shock and post-invalidation quarantine rules stopped those from becoming false signals.
- RIOT now validates the same rule from the other direction. A huge benchmark gap does not deserve a buy label when the candidate itself closes sharply lower. Friday's gap expansion is a warning until the stock starts reclaiming price.
The core method therefore stays intact:
- Require a genuine economic, operating or underlying-asset relationship.
- Require multi-window divergence; one dramatic day is not enough.
- Reject company-specific bad news, guidance cuts, dilution stress, legal/product problems, balance-sheet trouble or structural competitive breaks that explain the lag.
- Require a credible catalyst or supportive estimate backdrop.
- Require a fresh 5- or 10-session gap; a large 20-session number alone can be stale.
- Require absolute stabilization or a reclaim by the candidate itself.
- Require a second regular session of confirmation after a reclaim.
- Keep failed setups in post-invalidation quarantine until broken support is rebuilt.
- Do not count relative strength created by a peer-specific collapse.
- If a direct peer has earnings within two sessions, require post-event confirmation or agreement from a second benchmark / the candidate's own price action.
I am still not publishing a raw hit rate from a seven-session sample. The audit is being used to identify failure modes, not manufacture a probability badge.
1. Cadence: still the cleanest shape, but Wednesday can rewrite the pair
Why the relationship is genuine
Cadence and Synopsys are the two dominant U.S.-listed electronic-design-automation companies. They sell into overlapping semiconductor and systems customers and benefit from the same rising design complexity, advanced-node adoption, multi-die architectures and AI infrastructure investment.
That makes CDNS/SNPS a real operating relationship rather than two charts that happened to move together.
Fundamentals are not the obvious reason for the lag
Cadence's Q2 was strong: revenue growth was roughly 24%, backlog reached a record level, and management raised its full-year revenue-growth outlook to about 19%. Management continued to point to broad AI-driven demand across both design-for-AI and AI-for-design workloads.
The sell side remains constructive. StockAnalysis currently shows a Strong Buy consensus, with the average 12-month target around $404. That is not a short-term signal, but it supports the idea that the medium-window lag is not being universally interpreted as a broken franchise.
Public investor discussion around the Q2 beat-and-raise has also been broadly positive on backlog and AI demand. I treat that only as crowd temperature — not trade validation.
Friday finally gave the right kind of turn
CDNS closed Friday at $319.02, up 1.73%, while SNPS was almost flat. That ended four consecutive losing sessions for Cadence and left the 1-, 3- and 5-session relative windows positive.
Early Monday premarket indications have CDNS around the high $317s / low $318s, slightly below Friday's close. That does not invalidate the setup, and it does not confirm it either. Premarket liquidity is too thin for this strategy's confirmation rule.
| Item | CDNS plan |
|---|---|
| Observation area | Roughly $316–$320 |
| First useful confirmation | Another regular-session hold above that area while CDNS remains positive in absolute terms and continues to beat SNPS |
| Better trigger | Close through roughly $325–$328 with healthy technology/EDA breadth |
| Invalidation | Fresh close below roughly $311–$312 |
| Likely catalysts | AI/EDA demand, estimate support, Cadence's Wednesday conference appearance, Nvidia validating AI-capex demand |
| Main risks | Rich valuation, rising long yields, broad growth-stock de-rating, Synopsys earnings shock, AI-capex disappointment |
The problem: Wednesday is stacked
Synopsys reports fiscal Q3 after the close on Wednesday, August 26. Cadence is also scheduled to present at the Deutsche Bank Technology Conference that day, and Nvidia reports Wednesday evening.
That is a lot of fresh information for one relationship.
If SNPS drops sharply on a company-specific issue, CDNS can suddenly look brilliant in the relative table without doing anything. Under the existing peer-event rule, that would not count as clean confirmation.
Verdict: CDNS remains the best transition watch. I would rather see Monday hold Friday's improvement and eventually earn $325–$328 than buy simply because the 20-session gap is attractive.
2. Riot Platforms: maybe the most interesting gap on the screen, but Friday was ugly
Why Bitcoin is a genuine driver — with one important caveat
Riot is no longer a pure bitcoin miner. It is increasingly building an AI/data-center business, which means the relationship to Bitcoin is becoming less one-dimensional.
But Bitcoin is still economically important. Riot's Q2 included $113.7 million of Bitcoin Mining revenue out of $174.2 million total revenue, production of 1,587 bitcoin, and 11,380 bitcoin held at quarter-end. That makes spot Bitcoin / IBIT a legitimate underlying driver for the mining economics and a meaningful part of Riot's balance-sheet exposure.
The divergence is enormous
Friday was the striking part.
IBIT rose 6.02%, while RIOT fell 5.48% to $19.83. Over five sessions, IBIT gained roughly 22.6% while RIOT gained only about 4.3%. Over 20 sessions the relative deficit exceeds 32 percentage points.
An Investing.com intraday report captured why Friday matters: crypto equities, including Riot, were trading higher earlier in the session as Bitcoin approached $80,000. Riot still finished the day sharply lower.
That is not a normal “it hasn't noticed yet” setup. The stock noticed and rejected the move.
The business update is strong enough to keep it on the research list
Riot's August 10 Q2 release included several important positives:
- total revenue of $174.2 million, up 14% year over year;
- a 20-year, 191 MW data-center lease with a frontier AI lab, expected to generate about $9.1 billion of initial-term contract revenue;
- continued AMD data-center deployment;
- more than $1.2 billion of liquid assets at quarter-end.
Recent analyst revisions have also been notably positive. StockAnalysis shows a Strong Buy consensus and an average target around $32.40; recent firms including Clear Street, JPMorgan, Morgan Stanley, Bernstein, Citi and Needham have reiterated or raised targets after the data-center announcements.
Community discussion is predictably much more excited, especially around the AI-data-center contracts and possible additional Corsicana leasing. That excitement is exactly why I want price confirmation rather than using social enthusiasm as evidence.
| Item | RIOT plan |
|---|---|
| Observation area | Roughly $19–$20 |
| First useful repair | Regular-session close back above roughly $21 while no longer losing ground to Bitcoin / IBIT |
| Better confirmation | Reclaim Friday's $22.20 high, then hold into another session |
| Invalidation | Close below roughly $18.75 would weaken the setup; a break of the recent $17.85–$17.90 low would invalidate the current catch-up structure |
| Likely catalysts | Continued Bitcoin strength, crypto-regulatory clarity, AI/data-center lease execution, further Corsicana commercialization |
| Main risks | Bitcoin reversal, mining difficulty/hash-rate economics, capex/execution, high volatility, and the fact that Riot's growing data-center business makes IBIT a less pure benchmark over time |
Verdict: this is a fascinating divergence, not a buy signal. If Bitcoin stays firm and RIOT reclaims $21–$22.20 on its own strength, the setup becomes much more interesting. If it keeps falling while Bitcoin holds up, the market is telling us the relationship has changed or company-specific risk is being priced.
3. UPS: still a valid laggard, still missing the price confirmation
Why the relationship is genuine
UPS and FedEx share the same parcel/logistics cycle: U.S. business shipping, e-commerce, international air freight, fuel, labor costs, network utilization and global trade.
The pair therefore remains one of the cleanest operating comparisons in the screen.
Fundamentals are improving, but investors remain skeptical
UPS reported roughly $22.8 billion of Q2 revenue, adjusted EPS of $1.76, and raised its full-year expectations to approximately $91.2 billion of revenue and $7.22 of adjusted EPS. Management also said the planned Amazon volume glide-down and related network reconfiguration had been completed.
That keeps the 20-session gap worth investigating rather than dismissing UPS as fundamentally broken.
But the market's skepticism has a basis. Investors still need proof that cost cuts and the post-Amazon network reset can translate into durable domestic margin improvement. Recent sell-side views show the disagreement clearly: BMO reiterated a Hold with a target around $115 on August 21, while Citi, Goldman Sachs and Argus have recently carried Buy ratings with targets around the low $130s.
Employee/community discussion remains mixed on network volume, staffing and Amazon-related volume. Useful operating color, yes; evidence of a rerating, no.
The chart still has not earned the trade
UPS closed Friday at about $102.01. It remains modestly positive versus FedEx over the 3- and 5-session windows, but Friday itself slipped relative again and the stock is still below the upper-$100s repair zone.
| Item | UPS plan |
|---|---|
| Observation area | Roughly $101–$103 |
| First useful repair | Reclaim roughly $106–$108 |
| Confirmation | Hold the repair into a second regular session while maintaining positive relative performance versus FDX |
| Invalidation | Decisive close below roughly $101 weakens the setup further |
| Likely catalyst | Evidence that the network reset is improving volume quality and domestic margin; upward estimate revisions |
| Main risks | Weak parcel volumes, restructuring charges, competition from FedEx/Amazon, fuel/labor costs, second-half execution |
Verdict: still a legitimate secondary watch. The gap is there; the price turn is not.
Near-misses and rejected setups
A good catch-up screen needs a rejection table, otherwise it quietly becomes “buy whatever is down.”
| Setup | Why it is not promoted today |
|---|---|
| TER vs COHU | Genuine semiconductor-test relationship, and Teradyne's latest quarter was excellent. But the relative pattern is inconsistent: roughly -2.41 / +1.13 / -1.97 / -3.76 / +0.14 pp across 1/3/5/10/20 sessions. There is no clean persistent lag to mean-revert. |
| BURL vs ROST | Ross's Friday jump followed a strong earnings beat and raised guidance — a peer-specific event. Burlington itself reports August 27, so buying a “Ross catch-up” before Burlington's own earnings would violate the event-risk rule. |
| TJX vs ROST | TJX management explicitly described the Marmaxx slowdown as self-inflicted. That is a company-specific explanation for relative weakness, so the gap is rejected rather than treated as mispricing. |
| FTNT / FFIV | Both remain in post-invalidation quarantine. Recent relative strength has been contaminated by violent moves in cybersecurity peers, and neither has rebuilt its broken price structure. |
| NOC vs ITA | Friday underperformed the defense ETF, but the medium-window relationship has not produced a clean persistent lag. One weak session is not enough. |
Monday's backdrop makes confirmation more valuable, not less
U.S. futures are softer before the open. Reuters reports that investors are juggling potential new Iran sanctions, elevated long-term Treasury yields, Nvidia's August 26 earnings, and the Jackson Hole symposium later this week. The S&P 500 technology sector fell more than 3% last week, while the Philadelphia Semiconductor Index lost roughly 5%.
That backdrop matters most for Cadence and Riot: both can move hard when long-duration growth assets or risk appetite reprice.
The practical rule is simple: premarket movement does not count. A Monday regular-session close is more valuable to this strategy than an exciting 8:15 a.m. chart.
Bottom line
There are three worthwhile catch-up relationships on today's screen, but no clean U.S. buy before the opening bell.
Cadence is still first because it has something the others do not: the old gap remains while the newest completed session actually turned positive in both absolute and relative terms. Another hold in the $316–$320 area would strengthen the case; $325–$328 would be cleaner.
Riot is the new high-dispersion watch. The gap versus Bitcoin is enormous and recent business/analyst news is supportive, but Friday's 5.5% decline while IBIT gained 6% is the opposite of confirmation. $21 first, then $22.20, is a much better test than simply buying the biggest negative number in the table.
UPS remains the patient secondary setup. It still needs $106–$108 and a second-session hold.
The lesson from the seven-session audit remains the same: catch-up trades become interesting when a genuine relationship is still intact and the laggard begins taking performance back — not merely when the spreadsheet says the gap is large.
Educational disclaimer: This report is for informational and educational purposes only. It is not personalized financial advice, investment research, or a recommendation to buy or sell any security. Relative relationships can break permanently instead of mean-reverting.
Sources
Market / upcoming catalysts
- Reuters — S&P, Nasdaq futures slip as markets await Iran sanctions and Nvidia results, Aug. 24: https://www.reuters.com/business/sp-nasdaq-futures-slip-markets-await-iran-sanctions-nvidia-results-2026-08-24/
- Reuters — Nvidia earnings and Jackson Hole to test the stock rally, Aug. 21: https://www.reuters.com/business/wall-st-week-ahead-nvidia-earnings-jackson-hole-test-pillars-stock-rally-2026-08-21/
- Synopsys Investor Relations — Q3 fiscal 2026 earnings scheduled Aug. 26: https://investor.synopsys.com/news/news-details/2026/Synopsys-Announces-Earnings-Release-Date-for-Third-Quarter-Fiscal-Year-2026/default.aspx
- Cadence — Deutsche Bank Technology Conference appearance, Aug. 26: https://www.cadence.com/en_US/home/company/newsroom/press-releases/pr-ir/2026/cadence-to-present-at-deutsche-bank-technology-conference.html
Cadence / Synopsys
- Cadence — Q2 2026 results and raised outlook: https://www.cadence.com/en_US/home/company/newsroom/press-releases/pr-ir/2026/cadence-reports-second-quarter-2026-financial-results.html
- StockAnalysis — CDNS historical prices: https://stockanalysis.com/stocks/cdns/history/
- StockAnalysis — SNPS historical prices: https://stockanalysis.com/stocks/snps/history/
- StockAnalysis — CDNS analyst forecasts: https://stockanalysis.com/stocks/cdns/forecast/
- Public — CDNS premarket reference: https://public.com/stocks/cdns/pre-market
- Reddit — recent CDNS earnings discussion: https://www.reddit.com/r/TradeVerseNetwork/comments/1v8cr60/cadence_design_systems_cdns_q2_2026_earnings/
Riot / Bitcoin
- Riot Platforms — Q2 2026 results and data-center lease: https://www.riotplatforms.com/riot-platforms-reports-second-quarter-2026-financial-results-and-strategic-highlights/
- Reuters — crypto shares climb after Treasury buybacks and crypto-policy push, Aug. 20: https://www.reuters.com/legal/government/bitcoin-crypto-shares-climb-after-trump-pushes-clarity-act-2026-08-20/
- Investing.com — crypto stocks climb as Bitcoin approaches $80,000, Aug. 21: https://www.investing.com/news/stock-market-news/crypto-stocks-climb-as-bitcoin-nears-80000-mark-93CH-4871811
- StockAnalysis — RIOT historical prices: https://stockanalysis.com/stocks/riot/history/
- StockAnalysis — IBIT historical prices: https://stockanalysis.com/etf/ibit/history/
- StockAnalysis — RIOT analyst forecasts: https://stockanalysis.com/stocks/riot/forecast/
- Reddit — Riot data-center deal discussion: https://www.reddit.com/r/wallstreetbets/comments/1vl38lj/anthropic_strikes_9_billion_cloud_deal_with_riot/
UPS / FedEx
- UPS Investor Relations — Q2 2026 results: https://investors.ups.com/news-events/press-releases/detail/2164/ups-releases-2q-2026-earnings
- Reuters — UPS raises full-year forecasts: https://www.reuters.com/business/ups-raises-full-year-revenue-forecast-2026-07-28/
- StockAnalysis — UPS historical prices: https://stockanalysis.com/stocks/ups/history/
- StockAnalysis — FDX historical prices: https://stockanalysis.com/stocks/fdx/history/
- StockAnalysis — UPS analyst forecasts: https://stockanalysis.com/stocks/ups/forecast/
- Reddit / UPSers — operating-volume discussion: https://www.reddit.com/r/UPSers/comments/1vi6lqt/amazon_coming/
Rejected / near-miss references
- Teradyne — Q2 2026 results: https://investor.teradyne.com/news-events/press-releases/detail/351/teradyne-reports-second-quarter-2026-results
- StockAnalysis — TER historical prices: https://stockanalysis.com/stocks/ter/history/
- StockAnalysis — COHU historical prices: https://stockanalysis.com/stocks/cohu/history/
- Reuters — Ross Stores raises annual profit forecast, Aug. 20: https://www.reuters.com/business/retail-consumer/ross-stores-raises-annual-profit-forecast-again-discounted-apparel-demand-2026-08-20/
- Reuters — TJX Marmaxx slowdown described as self-inflicted, Aug. 19: https://www.reuters.com/business/retail-consumer/off-price-retailer-tjx-raises-annual-profit-forecasts-2026-08-19/
- Burlington Investor Relations — Q2 earnings scheduled Aug. 27: https://www.burlingtoninvestors.com/news-and-events/events-and-presentations/
Community links are anecdotal and included only for sentiment / operating color, not as trade validation.