Institutional flow

Institutional Flow — Scan of Jul 18, 2026

Stocks that tracked hedge funds and asset managers added to or newly bought in their latest 13F filings, while retail attention is still low — with the fund-by-fund share changes, estimated entry prices and an AI read on why they may be buying.

Next scan: tomorrow by 9:00 AM ET. A new scan is posted every Saturday by 9:00 AM ET.
5stocks qualified
5funds adding
4new positions
$5.31Bheld by those funds

AIThis scan

Comparing the 13F windows ending 2026-03-31 and 2025-12-31 for Berkshire Hathaway, Bridgewater, Renaissance Technologies, Citadel, and Two Sigma, 982 CUSIPs passed the raw acceleration filter (net buyers > sellers plus either ≥25% aggregate share growth or ≥2 funds opening new positions). After stripping mega-cap retail favorites, sub-$1B market caps, thin volume, and any name reporting earnings inside seven trading days (which knocked out Barclays, Teck Resources, PACCAR, 3M, and Regency Centers, all Q2 reporters between July 23–29), five names cleared every rail. The single most surprising finding: Berkshire opened a brand-new 39.8M-share stake in Delta Air Lines worth roughly $2.65B at the March mark — Buffett's first US airline position since he liquidated the entire sector in April 2020. Citadel added independently, giving the name a two-manager confirmation with no fund trimming beyond a trivial Bridgewater cut.

How this scan was run

Pulled the last two 13F-HRs for each of the five configured funds via fetch_13f_holdings (report periods 2026-03-31 and 2025-12-31; all five funds returned both filings — none failed). Aggregated per-CUSIP shares across each fund's sub-managers, dropped put-option lines (calls kept), then classified each CUSIP by net-adder count vs net-trimmer count across managers. Kept CUSIPs where adders exceeded trimmers AND either aggregate share count grew ≥25% Q/Q or ≥2 funds newly opened positions. Excluded single-fund block buys representing >80% of the total add unless the concentrated buyer was Berkshire — Buffett block buys are the intended signal, so DAL and NYT pass despite Berkshire dominance. Weighted slower-moving managers (Berkshire, Bridgewater) higher in the thesis narrative given quant 13Fs (Renaissance, Two Sigma) reflect only quarter-end snapshots of high-turnover books. Resolved CUSIPs to tickers via Yahoo/SEC lookups; no dual-class ambiguity in this batch (NYT CL A only, DAL common only). Average entry prices are derived from filing valueUsd ÷ shares at the March 31 mark and are ESTIMATES, not cost basis. Earnings dates verified via SEC 8-K/6-K filings and vendor calendars; any name reporting on or before 2026-07-28 was dropped.

#1

DAL

Delta Air Lines, Inc.
IndustrialsMkt cap $55.0BNext earnings Oct 8, 2026
Price
$84.17
Est. fund entry
$66.48+26.6%
Shares held
45.7M
Position value
$3.04B
Funds that added
FundPrior sharesShares nowChange
Berkshire Hathaway—39.8MNew position
Citadel Advisors5.19M5.92M+14.2%
Retail attention 2 of 4 signals quiet
Google Trends28/100busy
News, 30 days42 articlesbusy
Call/put OI1.60quiet
Retail hotlistsNot listedquiet

AIWhy these funds may be buying

Berkshire's first US-airline stake since Buffett famously liquidated the entire sector in April 2020 — a 39.8M-share position worth $2.65B was opened at a ~$66 March mark. The concurrent Amazon exit and Apple trimming in the same filing show capital rotating out of mega-cap tech and into simpler cash-flow franchises with widening moats. Citadel independently added 14%, and Bridgewater's trim was rounding-error size. The unpriced catalyst: Delta's premium cabin plus AmEx-loyalty economics increasingly separate it from the spot-fare airline pack, and unit costs step down as the 2023 pilot-contract inflation bulge rolls off in H2 2026.

What would prove it wrong

Airlines remain highly cyclical — a US consumer softening or crude spike would compress margins fast. Delta already up ~89% over the past year means much of the operational rerating is discounted; a next-cycle recession scare could unwind the multiple regardless of Buffett's presence.

#2

POR

Portland General Electric Company
UtilitiesMkt cap $6.1BNext earnings Jul 31, 2026
Price
$52.43
Est. fund entry
$52.77-0.6%
Shares held
4.83M
Position value
$255.0M
Funds that added
FundPrior sharesShares nowChange
Two Sigma Investments1.04M3.04M+191.1%
Citadel Advisors651.1K1.64M+151.2%
Bridgewater Associates59.1K115.1K+94.5%
Renaissance Technologies—40.4KNew position
Retail attention 4 of 4 signals quiet
Google Trends6/100quiet
News, 30 days4 articlesquiet
Call/put OI1.20quiet
Retail hotlistsNot listedquiet

AIWhy these funds may be buying

All four non-Berkshire managers added, tripling the aggregate position with zero trimmers — an unusually clean signal in a $6B utility that gets almost no retail airtime. POR serves the Oregon load pocket where hyperscaler AI data-center campuses (Google/Meta/Microsoft) are pushing multi-decade load-growth guidance well above the sector's historic 0.5%. Two Sigma and Citadel raised stakes 191% and 151% respectively, unusual conviction for a slow-moving reg-util. The catalyst retail hasn't priced: the OPUC rate case now working through the pipeline supports upsized capex tied to data-center integration, and the July 31 print is the first quarter management can quantify contracted load from named counterparties.

What would prove it wrong

Oregon PUC could disallow portions of the data-center-tied capex, hitting authorized ROE. Wildfire-liability tail risk remains embedded in PGE's operating footprint even after the 2024 statutory reforms.

#3

KT

KT Corporation
Communication ServicesMkt cap $8.8BNext earnings Aug 13, 2026
Price
$17.56
Est. fund entry
$21.45-18.1%
Shares held
6.36M
Position value
$136.3M
Funds that added
FundPrior sharesShares nowChange
Two Sigma Investments1.36M5.05M+271.3%
Renaissance Technologies150.5K913.5K+507.0%
Citadel Advisors111.8K295.5K+164.3%
Bridgewater Associates—96.3KNew position
Retail attention 4 of 4 signals quiet
Google Trends5/100quiet
News, 30 days3 articlesquiet
Call/put OI1.10quiet
Retail hotlistsNot listedquiet

AIWhy these funds may be buying

A four-fund pile-in on a deeply out-of-favor Korean telco. Bridgewater's new position aligns with Dalio's broader Asia-macro tilt visible across the fund's Q1 filings; Two Sigma and Renaissance's asymmetric quant adds suggest signal appearing on their factor stack post the Korean 'Value-Up' policy announcement. ADR sits ~18% below the aggregate March mark ($17.56 vs $21.45), so the smart-money entry is now visible at a discount. The specific catalyst: Korea's Corporate Value-Up program is forcing capital-return commitments from KOSPI blue chips, and KT's AI-IDC monetization pipeline (real estate + cloud) provides a non-telco growth tail retail hasn't underwritten.

What would prove it wrong

Korean won weakness compresses ADR value regardless of local performance. Value-Up program implementation could stall if the next Korean election shifts policy. Legacy fixed-line business remains structurally challenged.

#4

SBSW

Sibanye Stillwater Limited
MaterialsMkt cap $6.5BNext earnings Sep 1, 2026
Price
$8.00
Est. fund entry
$12.32-35.1%
Shares held
15.1M
Position value
$186.0M
Funds that added
FundPrior sharesShares nowChange
Two Sigma Investments2.16M6.30M+192.3%
Bridgewater Associates265.7K4.43M+1566.1%
Citadel Advisors1.93M2.54M+31.5%
Renaissance Technologies—1.83MNew position
Retail attention 4 of 4 signals quiet
Google Trends12/100quiet
News, 30 days8 articlesquiet
Call/put OI1.40quiet
Retail hotlistsNot listedquiet

AIWhy these funds may be buying

Bombed-out PGM miner where every non-Berkshire fund added and the two macro-flavored books (Bridgewater 16×, Two Sigma nearly 3×) took the largest jumps. Sibanye already flagged group adjusted EBITDA up 371% Y/Y in Q1 2026 on an 87% jump in the 4E PGM basket — the recovery is happening in real time and the tape has not caught up (stock is 35% below the aggregate March mark). The catalyst retail is missing: sustained PGM autocatalyst supply squeeze plus a possible hybrid-ICE revival slowing the PGM demand-decline narrative, all against a debt stack just refinanced by the May $500M 6.25% note issuance. Semi-annual reporter, so no near-term earnings landmine.

What would prove it wrong

PGM basket prices are highly volatile and stock is a leveraged play on that basket. South African labor and power-supply disruptions can compress operating margins quickly. Semi-annual reporting means longer information gaps between prints.

#5

NYT

The New York Times Company
Communication ServicesMkt cap $12.3BNext earnings Aug 5, 2026
Price
$75.93
Est. fund entry
$83.75-9.3%
Shares held
20.2M
Position value
$1.69B
Funds that added
FundPrior sharesShares nowChange
Berkshire Hathaway5.07M15.1M+199.0%
Renaissance Technologies2.29M2.76M+20.3%
Two Sigma Investments1.94M2.31M+19.1%
Retail attention 4 of 4 signals quiet
Google Trends22/100quiet
News, 30 days14 articlesquiet
Call/put OI1.30quiet
Retail hotlistsNot listedquiet

AIWhy these funds may be buying

Berkshire tripled the Class A stake, adding ~10M shares (worth ~$840M at the March mark) — one of the larger single-quarter Berkshire adds in the filing, and consistent with the fund's broader rotation from mega-cap tech toward moat-heavy franchise businesses (see the Amazon exit and Constellation Brands trim in the same filing). Renaissance and Two Sigma both added on top, giving three-fund confirmation; Bridgewater and Citadel trimmed rounding-error positions. NYT stock now trades ~9% below Berkshire's estimated Q1 mark, so the entry footprint is visible at a discount. The catalyst retail hasn't priced: the Aug 5 print is the first quarter management can quantify Games/Cooking bundle-driven ARPU expansion post the price ladder rolled out in Q1.

What would prove it wrong

Stock trades at ~35× GAAP earnings, so multiple-compression risk is real if bundle-ARPU disappoints. AI-driven search referral decay is an ongoing structural overhang on the entire publisher category. Buffett's stake was disclosed May 15, so some of the follow-on flow has already hit the tape.

All scans

Based on public SEC 13F filings. Funds file them up to 45 days after each quarter ends, so they show positions at quarter-end, not today. Average entry prices are estimates (filing value ÷ shares) — 13F filings don't report cost basis. The summaries, theses and risks are written by AI as a research aid — they are not financial advice or a recommendation to buy or sell any security, and they can be wrong.

SPX Lens Institutional Flow tracks the quarterly 13F filings of a watchlist of hedge funds and asset managers and flags the stocks several of them are accelerating into — new positions or sharply larger ones — while search interest, news coverage and options positioning say the retail crowd hasn't noticed yet. It is free, and a new scan is posted every Saturday by 9:00 AM ET. How to read Institutional Flow →

See where SPX is heading next. The SPX Lens chart shows key levels, live signals and an AI forecast for the next bars — free to try.

Open the live chartCreate a free account

© SPX Lens · About · Terms · Privacy — for educational purposes only, not financial advice.