InsiderTrades cohort data for the PDG/DG · Mega bucket gives you a 90-day sample size of 15,098. That is a large enough pool to be useful, and large enough to keep you from cherry-picking one anecdote. The 90-day win rate is 50.6%, the average 90-day return is 1.91%, and the 365-day average return is 25.65%. Read that carefully. The short-horizon number is only modestly positive, and the win rate is barely above a coin flip. This is not a bucket that screams immediate edge on its own.
That is exactly why the caveat matters. The cohort data is historical, not predictive. It tells you what has tended to happen after similar filings by similar roles in similar size bands. It does not tell you what Bouygues will do now. If you are looking for a clean mechanical buy signal from a CEO sale, you will not find one here. If you are looking for a way to frame the trade inside a broader pattern, you have one. The distinction is the whole game.
The fundamentals are decent, which keeps the sell from becoming a thesis
BOUYGUES insider-trading story">
InsiderTrades fundamental data gives Bouygues a score of 67, with a rank of 4,251 out of 21,368. The value pillar is 81 and the quality pillar is 53, while growth is not provided in the dossier and should be left out. That is a respectable screen, and it is one reason this filing should not be treated as a red flag in isolation. A company can have a decent fundamental profile and still see insiders sell stock. Those are not contradictions. They are separate layers of the read.
The value score matters here because it tells you the name is not being treated as a stretched momentum vehicle. The quality score is middling rather than exceptional, which is consistent with a large industrial group that can produce cash but does not live in the same margin profile as a software compounder. Put differently, Bouygues is not the kind of stock where a sale by a senior executive automatically implies a broken story. It is the kind of stock where you need to ask whether the cluster is a timing issue, a portfolio issue, or a genuine change in internal confidence. The filing alone does not answer that.
The cluster is the part that deserves your attention
InsiderTrades data says this is a cluster, with five distinct insiders trading the name in the same direction over the past quarter and 12 recent declarations in the cluster set. The recent list includes Benoit Torloting on June 24, June 11, and June 10, Jean-Manuel Soussan on June 10, and Olivier Roussat on June 10. That is enough to move the read from routine to notable. A single sale can be shrugged off. A cluster, especially when it includes senior names, is harder to wave away.
The direction matters too. This is not a mixed-direction cluster with offsetting buys and sells. It is a selling cluster. That does not mean management is bearish in some dramatic, all-hands-on-deck sense. It does mean the internal flow is leaning one way, and the market usually pays more attention when several senior people lean the same way in close succession. If you are weighing Bouygues, that is the part to sit with. The cluster does not prove anything. It does change the burden of proof.
The market tape around the filing is thin, so do not invent one
There was no verified analyst commentary, no company statement, and no management quote in the recent sources tied to the June 24 transaction. That absence matters. It means the filing is doing the work on its own, without a convenient narrative from the company or from the sell side to explain it away. It also means you should resist the temptation to backfill a story from the stock chart or from sector chatter that is not actually in the record we have.
The grounded research also did not surface broader market data directly linked to the filing window. So the honest read is narrower than the market often likes. We know the trade size, the role, the date, the cluster context, and the historical cohort bucket. We do not know whether the executive was rebalancing, funding a personal obligation, or responding to some internal view on valuation. You can speculate, but you cannot file speculation as analysis. The right move is to keep the signal intact and the motive unresolved.
Risks, caveats, and where this read breaks down
The first risk is overfitting. A clustered sale at a mega-cap can look more ominous than it is if you ignore how often senior executives sell for reasons that have nothing to do with the next quarter. The second risk is treating the cohort data as a forecast. It is not. A 50.6% win rate and a 1.91% average 90-day return are useful only as a historical reference point. They do not tell you whether Bouygues will outperform from here. The third risk is forgetting that the company is large and diversified. That reduces the informational purity of any single subsidiary executive sale.
There is also a simpler risk. The filing summary does not include share count or execution price, so you cannot compute a more precise behavioural read from the public summary alone. That keeps the analysis honest, but it also limits the strength of the conclusion. If you want to push this further, you need to watch for subsequent declarations, any change in direction from the same names, and whether the cluster broadens or fades. A cluster that keeps going is one thing. A cluster that stops after a few prints is another.
What to watch next
The next useful question is whether the June 24 sale is followed by more filings from the same senior names. If Torloting keeps selling, or if other executives join on the same side, the cluster read gets stronger. If the flow stops, the signal weakens and the June sequence starts to look more like a short-lived de-risking window than a durable internal view. That is the practical test. Not the headline. The follow-through.
You should also watch whether Bouygues discloses anything that changes the context around the telecom unit or the broader group. The current record gives you no management explanation for the sale, so any new corporate disclosure would matter. For now, InsiderTrades data says the fundamentals are decent, the cluster is real, and the historical bucket is only mildly positive over 90 days. That combination does not hand you a trade. It gives you a frame. If you are already in the name, it is a reason to pay closer attention to the next filing, not to panic. If you are not, it is a reason to keep Bouygues on the watchlist rather than force a conclusion from one executive sale.
BOUYGUES story">