Will ICON's final revenue restatement remain below 3% for all affected fiscal years?

The company disclosed <2% overstatement for FY2023 and FY2024 in a 6-K filing, which carries legal liability. This disclosure was made after ~4 months of investigation by forensic accountants, giving significant time to assess magnitude. ASC 606 cost-to-cost timing disputes in CROs are mechanically constrained — they involve percentage-of-completion estimates where small changes in total estimated costs shift revenue between periods. It is very difficult to generate a >3% overstatement through this mechanism alone without outright fabrication, which the investigation has not indicated. FY2025 could show a somewhat larger number if biases accelerated, but the 3% threshold is generous enough to absorb moderate upward revision from the preliminary <2% estimate.

Two separate considerations: (1) FY2023/2024 are likely locked at <2% given the specificity of the 6-K disclosure, and (2) FY2025 is the unknown. The question requires ALL affected years to be below 3%. If FY2025 shows a larger overstatement, it would need to exceed 3% to trigger NO resolution. The same ASC 606 mechanisms apply, and there's no indication the overstatement was accelerating year-over-year. The 'less than 2%' framing for both FY2023 and FY2024 suggests consistency, making a sudden jump to 3%+ in FY2025 unlikely. However, FY2025 includes Q4 2025 which may have had less oversight as the CEO transitioned and investigation began — creating a small window for larger errors. I assign ~85% for FY2023/2024 staying below 3% and ~90% for FY2025 staying below 3%, giving a combined ~78%.

The key question is whether 'less than 2%' in preliminary findings could become 3%+ in final findings. Historically, preliminary restatement estimates are generally accurate within ~50% of the final number — meaning <2% could become up to ~3%. However, this is the extreme case. The more common pattern is that preliminary estimates are slightly conservative (companies err on the side of overstating the problem in initial disclosures to manage expectations). The forensic accountants had 4 months to develop these figures before the February disclosure. The ASC 606 cost-to-cost method constrains the mechanics of overstatement. High confidence this stays below 3%.

The 6-K disclosure of <2% is the strongest evidence. However, 'less than 2%' is carefully worded — it could mean 1.9% (close to the 3% threshold with upward revision) or 0.5% (safely below). The multi-year pattern and expected material weaknesses suggest systemic issues that could produce larger adjustments in FY2025 when the investigation was ongoing and normal controls were disrupted. There is also the possibility that initial estimates grow as the forensic review deepens — this pattern has occurred in other restatements. Still, 3% is a generous threshold for a timing-related issue. Lean strongly toward YES but with some uncertainty about FY2025.

ASC 606 cost-to-cost method overstatements are mechanically limited. The method works by estimating total contract costs and recognizing revenue proportionally to costs incurred. To overstate by 3%+, you would need to either significantly underestimate remaining costs or dramatically overstate completion percentage. In CRO contracts spanning years, a 3%+ overstatement would require implausibly large estimation errors or intentional manipulation. The investigation has characterized this as a control failure (material weakness) rather than fraud. Control failures produce errors in the 1-3% range; fraud produces larger misstatements. Below-3% is the most likely outcome.

The risk scenario is scope expansion: what if the investigation discovers additional accounting areas beyond revenue recognition? The 6-K specifically scopes the investigation to 'revenue recognition in fiscal years 2023 through 2025,' but material weakness evaluation sometimes uncovers adjacent issues (e.g., contract cost estimation, goodwill impairment testing, expense classification). If additional areas are found, the total restatement magnitude could exceed 3% even if revenue recognition alone stays below 2%. This tail risk keeps my probability below 80%. The base case (revenue recognition only) strongly favors below 3%.

Company disclosed <2% for FY2023 and FY2024 in a legally binding 6-K filing. ASC 606 timing issues are mechanically constrained. The 3% threshold provides significant buffer. FY2025 is the unknown but same mechanisms apply. Strongly favors YES.

Two prior years at <2% strongly anchors expectations. FY2025 introduces uncertainty as controls were disrupted during CEO transition and investigation period. The 3% threshold is generous but not impossible to breach in a disrupted control environment. Main risk is FY2025 showing larger errors due to less oversight during Q4 2025. Still heavily favors YES given the established pattern.

The fundamental constraint is the ASC 606 cost-to-cost method: revenue recognition timing errors are bounded by the mechanics of percentage-of-completion accounting. A >3% overstatement would require either deliberate manipulation or implausibly large estimation errors in contract cost projections. The investigation has not indicated fraud. The 6-K disclosure is specific and legally liable. Strong YES.