EARNINGS-WEEK UPSIDE SCREEN — S&P 500, 18–21 Aug 2026 Analysis run 17 Aug 2026, 08:15 ET. Prices = 14 Aug 2026 closes unless noted.

S&P 500 constituents reporting in the window (12 names): HD and KEYS and JKHY (18 Aug), LOW, TGT, EL, TJX, ADI, NDSN (19 Aug), WMT, DE, ROST (20 Aug). All 12 were researched in full.

FULL SCREEN — the two criteria you weighted most

Immediate honest observation: nothing this week cleanly combines BOTH strong upward estimate revisions AND a beaten-down price. The two deepest price discounts (LOW −25.5%, EL −29.2%) are exactly the two names with the most negative revision trends. That is the trap your criteria are designed to avoid, and it is worth stating plainly rather than forcing a name to fit.

===================================================== TOP 3

#1 — TGT (Target), reports Wed 19 Aug premarket, consensus EPS $2.24 on $25.7B

Why it ranks first: the revision cycle is the strongest of any name in the index this week, and it is fresh, not a May artifact. Eleven price-target hikes in 90 days, and critically five of them landed in the last two weeks directly into the print: Citi 133→148 (3 Aug), Evercore 135→150 (4 Aug), UBS 144→166 (6 Aug), JPM 129→157 (7 Aug), Oppenheimer 140→170 and BMO 130→150 (11 Aug). Wolfe upgraded to Buy on 23 Jun. Only one downgrade in the window (Freedom Broker, 20 May).

The revisions are backed by an actual fundamental turn. Q1'2026 (20 May) printed EPS $1.71 vs $1.35 estimate = +26.67% surprise, with revenue $25.4B vs $24.3B = +4.54% surprise, EPS +31.54% YoY and revenue +6.70% YoY. That followed five straight quarters of negative YoY revenue. The prior two prints also beat (+2.89%, +12.96%).

The discount is in the valuation, not the chart. PE (TTM) 17.03 vs department/specialty-retail median 45.08; EV/EBITDA 9.30 vs 19.20; P/S 0.55 vs 1.49. Marked to today's price:

Options and flow support it. Front-week ATM IV is 87% (155-strike call) vs 78% puts — calls bid over puts, which is unusual and constructive. Volume on the 21 Aug expiry is heavily skewed to upside strikes: 170 calls 1.0K contracts, 165 calls 859, 160 calls 267, versus 214 and 213 contracts at the 150 and 155 puts. Large flow on 13–14 Aug was net bullish: a $460.0K at-ask trade in 21 Aug $157.5 calls (538% of prior OI), a $105.0K at-ask sweep in 21 Aug $130 calls, plus three at-bid put sweeps (put selling) — 20 Nov $130 put $190.1K, 20 Nov $145 put $207.4K, 18 Dec $150 put $145.6K. Options price ±6.9%, i.e. roughly $143 to $165.

The bias you must confront: TGT is NOT trading at a price discount. It closed at $154.48, RSI(14) = 71.03 (overbought), +11.8% in one month, and far above SMA50 137.71 and SMA200 117.23. It has run from $112.69 (12 Feb) to $155.51 (13 Aug), +38%. And the run has taken it past the Street: So the consensus target no longer implies upside — the stock is ahead of the revisions rather than behind them. The bull case here is that the revision cycle continues (analysts have been chasing, not leading) plus a still-cheap multiple; it is not a "buy the dip" case. Biggest risk: an in-line quarter with the multiple already re-rated from 11.3x (Q3'25) to 19.4x and RSI at 71 — a 6.9% implied move cuts both ways, and there is no oversold cushion. Score: 7.5/10.

#2 — ADI (Analog Devices), reports Wed 19 Aug premarket, consensus EPS $3.33 on $3.9B

This is the cleanest literal fit to your criteria: rising estimates plus an actual price pullback. All 20 analyst actions in the last 90 days raised the price target — zero cuts, zero downgrades. Cantor 400→510→550, KeyBanc 430→500→525, Wells 410→470→515, Stifel 405→450→498, Jefferies 410→475, Evercore 387→474. PT trend: Latest target per firm across 15 firms averages $470.80 (median $460) versus a $389.39 close = +20.9% average upside / +18.1% median. The stock trades below every single target on the sheet except one hold at 405.

Fundamentals are the strongest of the cohort: 8 of 8 EPS beats (median surprise +5.75%), Q2'26 revenue +37.2% YoY, GAAP EPS +293% YoY, gross margin 67.3% (up 1,266bp in two years) and operating margin 38.1% vs 28.4% a year earlier.

The pullback is real: −12.66% from the 22 Jun high of $445.83, having bottomed at $353.37 on 29 Jul, and price is still 0.56% below SMA50 (391.57) while holding 15.1% above a rising SMA200 (338.22). RSI(14) 54.1 — neutral, no stretch. Large options flow was 4 of 5 prints bullish (Jun'27 $360 calls $163.8K at ask, Dec'26 $490 calls $148.7K at ask, Sep $350 puts sold at bid $152.8K), though total premium was only ~$691K across ten sessions — thin.

The bias you must confront: ADI is expensive on any trailing measure — PE 58.5 vs semiconductor median 14.3, EV/EBITDA 32.8 (highest of its last nine quarters), P/S 15.2 (also a nine-quarter high). And the last print, 20 May, was a double beat that still closed −3.92%. Beats are priced in. Options price ±6.3% versus a 4.52% average absolute reaction over the last four prints — you are paying above realized. Biggest risk: guidance that merely matches the raised bar, into a 58x trailing multiple with put IV 12–20 vol points over calls (the market is already paying for downside). Score: 7/10.

#3 — HD (Home Depot), reports Tue 18 Aug premarket, consensus EPS $4.73 on $47.3B

Ranked third on the "discount" leg of your criteria, not the revision leg. The discount is genuine: −20.60% from the 52-week high of $426.75 (17 Sep 2025), −1.52% YTD, and unlike every other cheap name here HD trades roughly in line with its peer group (PE 23.0 vs retail-building-materials median 22.3, EV/EBIT 18.1 vs 18.6, P/FCF 22.6 vs 27.2) while sitting at its own cheapest multiple in two years (PE was 26.3 last quarter, 27.6 at peak). Average target $373.2 = +10.4% upside; median $377 = +11.5%.

Revisions are recovering off a mid-May reset, not deteriorating: Wells Fargo 360→400 (11 Aug), RBC 340→343 (12 Aug), Stifel 320→340 (17 Aug, today). Supporting flow: the single largest options print in the last two weeks was a bullish $679.7K at-ask sweep in 21 Aug $330 calls (14 Aug), three of five put prints executed at the bid (put selling), short volume fell from ~22% of tape in late July to 11–15% in mid-August (−19% in absolute terms), and 13F filers were net buyers with Sixth Street establishing a new 20.3M-share ($7.1B) position. Implied move is the smallest of the top three: ±4.2%, i.e. a $324–$352 band.

The bias you must confront, and it is the reason HD is third not first: HD has had ZERO upgrades and ONE downgrade (Wolfe, 23 Jun) in 90 days, and the fundamentals are the weakest in the group — operating margin has compressed from 15.13% (Q2'24) to 11.93% (Q1'26), net margin 10.56% → 7.88%, and EPS declined YoY in five of the last six quarters. Only 5 of 8 EPS beats, and all surprises are tiny (median absolute 2.6%), so this stock trades on guidance, not on the number. Target dispersion is $310–$430, meaning the Street does not agree on the trajectory. Biggest risk: another guidance cut. The 200-day SMA at $347.11 is falling and sits directly overhead as resistance, and price has already lost 4.71% in the last week. Score: 6/10.

===================================================== CLOSE MISSES AND WHY THEY DIDN'T MAKE IT

ROST (#4, 20 Aug postmarket, EPS $1.81): the most bullish large options flow of the entire cohort — $668.0K 18 Sep $220 put sold at bid (1145% of OI), $885.0K Jan'28 $240 put sold at bid, $232.5K Jun'27 $220 put sold below bid, versus only $816.0K of bearish call selling. Fundamentals are accelerating hard: 8/8 EPS beats, Q1'26 +20.24% EPS surprise, revenue +20.56% YoY with a +7.95% revenue surprise. But it fails both of your gates — targets average $259.0 (median $260) versus $245.36, only +5.6% upside; it carries a Wells Fargo downgrade (23 Jun); PE 32.0 vs industry median 10.7 and P/S 3.09 vs 0.44; and at RSI 53.7 sitting 3.9% under its 7 Aug high it is not a discount.

JKHY (18 Aug postmarket, EPS $1.44): genuinely beaten down — $153.31 versus SMA200 of $159.38 — and Stephens carries a $200 target (11 Aug) implying +30%. But coverage in the window is only three records, one of which was RBC cutting 180→173. Too thin a revision base to rank.

LOW and EL: the biggest price discounts of the week (−25.5% and −29.2% from highs) and both must be avoided on your own stated logic. LOW had 19 target cuts out of 24 records, four of them in the final three weeks (JPM −27, WF −10, Citi −18, Piper −2), PT trend −4.2%. EL had zero upgrades in 90 days and a PT trend of −7.4%. The headline "+17.9% analyst upside" on LOW is analyst inertia, not conviction.

DE, NDSN, TJX, KEYS, WMT: all fail at least one gate hard — DE has falling targets and a majority-hold rating mix at a two-year-high multiple; NDSN is 2.2% off its high with no analyst note in 11 weeks; TJX has had zero analyst actions since 21 May and faces a potential 50/200 SMA bearish cross; KEYS is 4.6% off its high at 57.6x earnings; WMT's revisions split 4 up / 4 down by firm with an Oppenheimer downgrade on 4 Aug.

DATA LIMITATIONS TO NOTE Dark pool (ATS) data terminates the week of 06 Jul 2026 for every ticker checked — roughly six weeks stale — so it could not inform pre-earnings positioning for any name. ATS share was flat in a 34–48% band across all names, i.e. no signal even where available. 13F data reflects 30 Jun positions. Forward PE is not available from the fundamentals feed; all forward multiples cited are approximations built from single-quarter consensus. Unusual options premium is thin for ADI (~$691K) and absent entirely for NDSN.

This is research, not trading advice, and I can't make the decision for you. Every one of these three has a live, quantified bear case above, and options are pricing 4.2%–6.9% single-event moves — position sizing matters more than name selection into prints like these.