STRL · Sterling Infrastructure, Inc. · Stock research

Current Thesis

The 2026-08-15 note set the thesis break at a weekly close below $550. That condition landed inside four sessions: the last completed daily close was $516.81 on 2026-08-21, 48.0% under the $993.74 closing high of 2026-06-04, with RSI(14) at 32.3 and a three-month price change of -29.5%. Nothing in the filings deteriorated in the interval. Q2, reported 2026-08-03, put revenue at $1,168.2M (+90% YoY) against a ~$989.8M consensus, adjusted diluted EPS at $5.80 versus $4.62 expected, combined backlog at $5.62B (+150% YoY), and lifted the FY26 adjusted EPS guide to $19.70–20.30. What is repricing is the multiple, and published research is following the tape rather than leading it — D.A. Davidson opened coverage at Buy with a $700 target on 2026-08-21, roughly a quarter below the $922–$956 marks that stood in early June. The leg on offer to a buyer here is a valuation reset that has not yet produced a higher low, and no company-scheduled event falls before the ~2026-11-02 Q3 report.

Bullish and bearish views on Sterling Infrastructure, Inc.

The model's bull view on Sterling Infrastructure, Inc. (STRL), in brief: Q2 2026 (released 2026-08-03): revenue $1,168.2M, +90% YoY, against a ~$989.8M Street bar; adjusted diluted EPS $5.80 vs $4.62 expected; GAAP diluted EPS $5.00; adjusted EBITDA $256.7M, +104% YoY. The bear view: The de-rate survived the best print the company has produced. Both cases follow in full.

Bull Case

• Q2 2026 (released 2026-08-03): revenue $1,168.2M, +90% YoY, against a ~$989.8M Street bar; adjusted diluted EPS $5.80 vs $4.62 expected; GAAP diluted EPS $5.00; adjusted EBITDA $256.7M, +104% YoY. Roughly half the growth was organic per the release, with CEC and Stone Ridge supplying the balance.
• Combined backlog $5.62B at 2026-06-30, +150% YoY; signed backlog $4.33B, +116% YoY, including $1.28B of unsigned awards. Mission-critical work — data center, manufacturing, semiconductor — is 92% of E-Infrastructure backlog, so forward coverage is contracted rather than pipeline commentary.
• FY26 guidance raised on 2026-08-03 to revenue $4.00–4.15B (from $3.70–3.80B), GAAP diluted EPS $17.25–17.85, adjusted EPS $19.70–20.30, adjusted EBITDA $891–916M. The adjusted guide midpoint of $20.00 stands against a 2026-08-21 close of $516.81.
• E-Infrastructure Solutions revenue $905.0M in Q2, +192% YoY, segment operating income +148%. The operating leverage sits where hyperscaler site work is being awarded.
• Cash and equivalents $464.5M at 2026-06-30; H1 2026 operating cash flow $328.0M; $35.3M of stock repurchased in H1. The growth has been funded internally through the half.
• Every published mark found remains above the market: the S&P Global poll compiled by stockanalysis.com and retrieved 2026-08-23 shows a Strong Buy consensus and an average target of $876 across 8 analysts, with Stifel at $804 (2026-08-11), KeyBanc at $754 (2026-08-05), Thompson Davis at $1,000 (2026-08-04), and the two most recent marks at $700 (2026-08-20/21). No rating cut has been published this cycle.

Bear Case

• The de-rate survived the best print the company has produced. Investing.com covered the Q2 deck on 2026-08-03 under the headline "90% revenue jump, shares fall"; Simply Wall St's account of the same session records the stock at $529.61 after hours against a $611.47 prior close. Eleven weeks after 2026-06-04, no higher low has held.
• The $550 area, named as the break level on 2026-08-15, gave way — the 2026-08-21 close of $516.81 is beneath both it and the prior note's reference close of $576.48 on 2026-08-14.
• The target cluster keeps migrating down: $956 (Cantor) and $922 (KeyBanc) in May–June, cut to $742 and $754 on 2026-08-05, then two fresh $700 marks on 2026-08-20/21. The polled average has moved from $918.67 across 7 analysts (CNN/S&P Global, retrieved 2026-08-15) to $876 across 8 (retrieved 2026-08-23) — still far above the market, which leaves the revision cycle unfinished.
• Momentum made a new low with price: RSI(14) at 32.3 on 2026-08-21 versus 42.9 on 2026-08-14. There is no positive divergence in that pair.
• The mix narrowed in Q2. Transportation Solutions revenue fell 20% YoY to $156.7M as crews were redeployed; Building Solutions was -1% at $106.5M. Growth and margin now ride on one end market whose customers set capex annually.
• A mixed shelf was filed 2026-05-12 with size undisclosed. Issuance capacity exists and a takedown would arrive without notice.

Setup & Price Structure

• The narrative is dead — the price structure, not the order book. What dates it: the 2026-08-03 triple beat-and-raise producing a lower low rather than a higher one; the loss of the $550 shelf at the 2026-08-21 close of $516.81; and a 48.0% drawdown from the 2026-06-04 high of $993.74 with no reclaimed level since. The picks-and-shovels story is intact in the filings and has stopped functioning as a price driver.
• The 2026-07-18 intraday low of $593.62 was the last visible structural reference above; it broke in early August and has not been retested from below. Below $516.81 there is no recent shelf in this series — the drawdown is now deep enough that the nearest reference is round-number, near $500.
• Positioning observables rather than a verdict on them: consensus rating remains Strong Buy at an $876 average with the market at $516.81 (retrieved 2026-08-23), so the downgrade lever is unspent; two initiations landed at $700 within 48 hours of the low, adding coverage into weakness; no insider purchase or sale has surfaced in the window; the company bought back $35.3M of stock in H1; the 2026-05-12 shelf remains undrawn as far as public filings show. The next earnings print is roughly ten weeks out, so none of the crowding is print-driven.
• RSI(14) of 32.3 is close to, but not through, the conventional 30 line. That is a description of momentum, not evidence a low is in.

Catalyst Calendar (next 30 days)

• Through 2026-09-22: no company-scheduled event is on the calendar. Sterling fixed the Q2 date via a scheduling press release ahead of the 2026-08-03 release; the equivalent Q3 scheduling notice has historically landed around mid-October and would be the first dated marker.
• ~2026-10-27 (est.): large-cloud-platform Q3 prints with 2027 capex framing. Outside the 30-day window, but it is the first external input that touches award flow.
• ~2026-11-02 (est.): Q3 2026 results and the FY26 guidance update. First scheduled test of the raised $4.00–4.15B revenue and $19.70–20.30 adjusted EPS guide.

Elapsed catalysts

• Any 424B or 8-K takedown against the 2026-05-12 shelf is undated and could land inside the window. (passed 106d ago)

What Would Change Our Mind

The structure that failed is the $550 shelf, and the argument that Q2's record set a floor under the equity failed with it on 2026-08-21. The next level that would extend the same sequence is a weekly close below $500; that would put the drawdown past half the 2026-06-04 high and remove the last round-number reference in this series. Two secondary conditions would carry independent information: an FY26 adjusted EPS guide cut beneath $19.70 at the ~2026-11-02 Q3 print, or combined backlog printing lower than the $5.62B recorded at 2026-06-30. On the other side, the read that this remains a de-rate rather than a reset would be challenged by a weekly close back above $593.62 — the 2026-07-18 intraday low, the first level lost on the way down — held for a second week, or by a target being raised by a firm that cut on 2026-08-05. A downgrade from Overweight or Buy by any of KeyBanc, Cantor, Stifel or D.A. Davidson would be the first rating action of the cycle and would confirm that the revision phase has moved past arithmetic on the multiple.

Correlation Notes

• The engineering-and-construction/AI-datacenter cohort — Comfort Systems, EMCOR, Quanta, Vertiv, Powell — has repriced as a basket since early July; the prior note flagged the 2026-07-02 session where the group fell together on no single-name news. Another such session would say the de-rate is cohort-level.
• E-Infrastructure awards trace back to hyperscaler capex budgets, so the name carries the beta of large-cloud 2027 capex commentary at the late-October prints more than it carries its own backlog print.
• Building Solutions ties the residential-slab work to the homebuilder complex and to rates; at $106.5M of Q2 revenue and -1% YoY it is now a small share of the story.
• Benzinga's 2026-05-27 piece framed AI-linked small caps as the driver of the Russell 2000's record run. That framing was the flow behind the June high, and its reversal is the flow behind the current tape.