CRAI · CRA International, Inc. · Stock research

Current Thesis

CRA International sells economists and expert testimony by the billable hour into two end-markets that both ran hot in the June quarter: merger review and litigation. The 2026-08-06 Q2 FY2026 print delivered record revenue of $210.815M (+12.8% YoY) against a $198.903M consensus, and management lifted FY2026 constant-currency revenue guidance to $805–820M from $785–805M — the new floor sits exactly at the old ceiling, and above the $801.21M consensus. The narrative leg an investor buys here is a toll on deal and litigation volume, with an operating-leverage kicker from utilization.

What the print did not deliver is margin. Non-GAAP diluted EPS came in at $2.16 against $2.15 consensus — an in-line bottom line on a roughly $12M revenue beat — and the non-GAAP EBITDA margin band was reaffirmed at 12.0–13.0% rather than raised. The stock closed $169.31 on 2026-08-07, 23.6% below its $221.74 52-week high, with RSI(14) at 48.0. Records are printing at the revenue line; the multiple has not followed.

Life-cycle: MATURING. The antitrust engine has now produced six consecutive record quarters per management's 2026-08-06 commentary, so this is a known, working story rather than a fresh one. But the equity is a quarter below its high after a beat-and-raise, RSI is mid-range, and there is no new marginal bid visible in the price. Well known, still working, moderating flow.

Bullish and bearish views on CRA International, Inc.

The model's bull view on CRA International, Inc. (CRAI), in brief: Record quarter, and it beat wide. Q2 FY2026 revenue $210.815M vs $198.903M consensus; +12.8% YoY, the highest quarterly revenue in company history (2026-08-06 release). Guidance raised, not merely reaffirmed. FY2026 constant-currency revenue guidance moved to $805–820M from… The bear view: The beat did not reach the bottom line. Roughly $12M of revenue upside converted to a $0.01 non-GAAP EPS beat ($2.16 vs $2.15). The EBITDA margin guide stayed at 12.0–13.0%. Growth is being bought with headcount and comp, and there is no evidence yet of scale economics.… Both cases follow in full.

Bull Case

• Record quarter, and it beat wide. Q2 FY2026 revenue $210.815M vs $198.903M consensus; +12.8% YoY, the highest quarterly revenue in company history (2026-08-06 release).
• Guidance raised, not merely reaffirmed. FY2026 constant-currency revenue guidance moved to $805–820M from $785–805M on 2026-08-06, against $801.21M consensus.
• Utilization at 77% with net headcount +3.3% (Q2 FY2026 call, 2026-08-06). Consulting revenue is headcount × utilization × rate; both inputs moved the right way in the same quarter.
• International +32.9% YoY, described by management as organic expansion in European Life Sciences and Antitrust rather than acquired (2026-08-06 call).
• The demand backdrop is quantified, not asserted. Management cited $1.6 trillion of global M&A in the June quarter, case filings +11% and court judgments +5% (2026-08-06 call). Antitrust & Competition Economics logged its sixth consecutive record quarter.
• A second demand line is forming. Management flagged Energy-practice growth off utility data-center load forecasts and "demand amplification from AI complexity" in Forensic Services and incident response (2026-08-06 call). Neither is yet a disclosed dollar figure.
• a $0.57/share quarterly dividend was declared 2026-08-06, record 2026-08-25, payable 2026-09-14. The buyback ran while the shares sat well below their 52-week high.
• Balance-sheet capacity added. The credit facility was expanded on 2026-08-06 to a five-year $400M structure — a $75M term loan plus a $325M revolver — replacing a $300M revolver.

Bear Case

• The beat did not reach the bottom line. Roughly $12M of revenue upside converted to a $0.01 non-GAAP EPS beat ($2.16 vs $2.15). The EBITDA margin guide stayed at 12.0–13.0%. Growth is being bought with headcount and comp, and there is no evidence yet of scale economics.
• Receivables lengthened. Days sales outstanding moved to 113 days from 100 days in the prior quarter (2026-08-06 call). In a labor firm, cash collection lagging revenue growth is the first place quality-of-earnings problems show up.
• The M&A input is cyclical and near a high. The antitrust streak is levered to $1.6T of quarterly global deal volume. Expert-economics work is a lagging derivative of announcements; when the announcement pipeline thins, the backlog rolls off with a quarter or two of delay.
• Antitrust intensity is policy-dependent. Second Request and merger-challenge workloads shift with the enforcement stance of the reviewing agencies. That variable is outside management's control and is not disclosed as a backlog number.
• Price is not confirming the fundamentals. A record quarter and a raise left the stock at $169.31 on 2026-08-07, 23.6% under $221.74. The market is compressing the multiple against improving revenue — a pattern that continues until it doesn't, and nothing in the 2026-08-06 disclosure forces it to stop.
• FX sits between the guide and the print. The $805–820M range is constant-currency; with international revenue growing 32.9%, translation is a live wedge between guided and reported figures.

Setup & Price Structure

Measured, as of the 2026-08-07 close: $169.31; 52-week high $221.74; distance from high −23.6%; three-month return +21.8%; RSI(14) 48.0.

Inferred from that set: the +21.8% three-month advance is a partial retracement of a larger drawdown, and a mid-range RSI on the session after a beat-and-raise says the advance stalled going into the print rather than extending through it. There is no momentum extension to fade and no evidence of a stretched position — RSI 48 is the middle of the range, and a name 23.6% below its high is not where crowding usually shows up.

Positioning observables, stated without a verdict: the earnings binary is behind rather than ahead, so there is no imminent print compressing the window; new debt capacity ($400M facility, 2026-08-06) was added rather than equity issued, so there is no issuance-into-strength signature; and no insider transactions appear in the filing record reviewed for this note. Retail-sentiment coverage on this name is sparse — the most recent third-party headline before the print was a 2026-07-07 client engagement item — which is the opposite of the clustering that marks a saturated narrative.

The structure to watch is whether the recovery leg extends toward the $221.74 high or gives back the three-month advance. The $155 area marks where most of that advance would be surrendered.

Catalyst Calendar (next 30 days)

• 2026-08-25 — Dividend record date, $0.57/share (payable 2026-09-14, declared 2026-08-06). Confirms capital-return cadence; it resolves nothing about demand.
• ~2026-08-14 (est.) — Q2 FY2026 Form 10-Q. The receivables detail behind the 113-day DSO and practice-level revenue splits land here; the 2026-08-06 release and call gave the headline but not the composition.
• ~2026-10-29 (est.) — Q3 FY2026 results. Outside the 30-day window and the genuine binary: it is the first test of whether the $805–820M FY guide holds and whether 12.0–13.0% EBITDA margin becomes a floor or a ceiling.

No scheduled regulatory, product or index event is identified inside 30 days.

What Would Change Our Mind

The structure that has to hold is the three-month recovery leg. A weekly close below $155 says the market has read the 2026-08-06 beat-and-raise as a cycle high and is re-pricing the antitrust streak as peak-cycle earnings rather than a run-rate — at that point the fundamental improvement has stopped mattering to the tape and the case is broken on its own terms.

Three fundamental conditions would do the same work independently:

1. A trimmed or unchanged-to-lower FY2026 guide at the Q3 print (~2026-10-29, est.). The raise to $805–820M is the whole reason the second half is de-risked; walking it back removes the leg.
2. A third consecutive sequential DSO extension past 113 days, or any allowance/write-off commentary in the Q3 10-Q. Cash conversion deteriorating alongside record revenue changes what the revenue is worth.
3. The antitrust streak breaking — management describing Antitrust & Competition Economics as flat or declining after six consecutive record quarters, or a global quarterly M&A figure materially below the $1.6T cited for Q2 2026.

Conversely, the case strengthens if the Q3 print pairs revenue inside or above the raised band with an EBITDA margin at or above the 13.0% top of the guided range — that would be the first evidence that headcount growth is producing operating leverage instead of consuming it.

Correlation Notes

• Primary driver: announced global M&A volume and litigation filing counts. Management anchored the quarter to $1.6T of Q2 deal activity and +11% case filings. CRAI trades with the deal cycle on a lag, not with the tape on the day.
• Peer complex: expert-services and litigation-consulting names — FTI Consulting, Exponent, Huron, ICF. Read their prints for the same two inputs; divergence between CRAI and that group on a common quarter is more informative than CRAI's absolute move.
• The AI/data-center linkage is second-order and small. Management cited Energy-practice growth from utility data-center load forecasting and AI-driven complexity in Forensic Services (2026-08-06), but neither is a disclosed revenue line. Treating CRAI as a data-center infrastructure proxy overstates the exposure; it is a professional-services firm whose customers happen to include utilities doing load studies.
• Rate and small-cap beta. With a market capitalization in small-cap territory and a newly enlarged $400M facility, the name carries the discount-rate sensitivity of the Russell complex independent of anything in the operating business.