The point of the cohort data is not to talk you out of every buy. It is to stop you from confusing insider activity with easy alpha. A 32.4% win rate at 90 days means the bucket has historically been right less often than wrong, and the negative average returns tell you the losers have outweighed the winners on average. That is the kind of record that should make you more selective, not less alert. When the role is CEO and the company is micro-cap, the signal deserves attention. When the bucket history is this weak, the burden of proof shifts back onto the trade itself.
The dispersion implied by those numbers also matters. A negative average return can coexist with occasional sharp winners, especially in small-cap and micro-cap names where outcomes are lumpy. But the dossier does not give us a clean dispersion table, so we do not pretend otherwise. What we can say is simpler. Historically, this is not a bucket that has rewarded blind following. If you are using Predilife as a test case, the right question is not whether the CEO bought. He did. The right question is whether this particular buy has enough size, repetition, and company context to overcome a weak historical base rate. That is a harder question, and the data says you should treat it as such.
The cluster picture is narrow but persistent
The cluster is the strongest part of the story, and it is also the easiest part to overstate. The dossier flags the filing as part of an insider cluster, and the recent declarations list six June buys by Stéphane Ragusa on June 25, June 18, June 12, June 9, June 9, and June 8. The cluster is therefore not a one-day blip. It is a run of repeated buys over most of the month. That is meaningful because repetition tends to matter more than a single isolated transaction, especially when the same insider keeps showing up.
But the cluster is also narrow. The dossier says there was one distinct insider. So this is not a broad insider accumulation across several executives. It is the CEO, repeatedly. That can be a strong read if you believe the chief executive is the best informed actor in the room, which is often true in a small company. It can also be a sign of conviction that is limited by personal capacity, liquidity, or simple habit. The filing record alone cannot tell you which. It can only tell you that the same name kept buying.
That is why the cluster matters most when paired with the score. Our scoring leaned on the CEO role and the clustering, but it did not ignore the microscopic filing size. That balance is exactly right here. If the buys were larger, the cluster would carry more weight. If the cluster were broader, the same size could be easier to excuse. Instead, you have repeated buys from one executive in a micro-cap, and a score that reflects interest without excess enthusiasm. That is a disciplined read, and it is the one the data supports.
Fundamental health is not available, so do not invent it
The internal dossier does not provide fundamental pillars for Predilife, so there is no transparent screen to walk through here. That absence is itself a fact. It means we cannot dress up the insider buy with a neat list of revenue momentum, margin expansion, balance-sheet strength, or cash burn metrics that were not supplied. In a lot of stories, this is where the copy gets lazy and starts borrowing confidence from a press release. We are not doing that.
What we can say is that the lack of fundamental detail makes the insider read more, not less, important as a piece of evidence. In a micro-cap Healthcare & Pharma name, where the market cap is EUR 9.46 million and the public record around the filing is thin, the trade is one of the few observable actions from inside the company. That does not make it predictive. It makes it relevant. The absence of a fundamental dossier also means the trade should not be mistaken for a valuation argument. A CEO can buy for many reasons, and without the company’s operating metrics in front of us, the filing cannot be converted into a clean fundamental thesis.
So the right discipline is to keep the insider signal in its lane. It tells you that management is still buying into the name in June. It does not tell you whether the business is improving, whether the balance sheet is stable, or whether the next quarter will be better than the last. If you are going to own a micro-cap on insider activity, you need to know which parts of the story you actually have and which parts you are filling in with hope. Here, the dossier gives you the former and not the latter.
Strategy context and what the backtest can, and cannot, claim
Our strategy context is a 90-day holding period with a maximum position size of 0.08. The out-of-sample Sharpe is 0.56 and the out-of-sample CAGR is 17% on a restricted EU venue universe. Those numbers survived a narrow test, but they do not survive search-aware deflation, and the window is short and single-regime. That caveat belongs in the sentence with the numbers, not in a footnote after the fact. The strategy context is useful because it tells you how our system has behaved in a constrained setting. It is not a promise that Predilife will fit the same pattern.
The universe win rate is 51.5%, which is only modestly above a coin flip. That is another reason to stay sober. The strategy is not a magic machine that turns every CEO buy into a trade. It is a filtered framework that has shown some edge in a restricted setting, with clear limits. In a case like Predilife, the framework says the filing is worth attention because the role is high, the cluster is active, and the company is small. It also says the trade size is tiny and the historical bucket is weak. Those two truths sit together.
If you are using this as a live screen, the practical takeaway is simple. The strategy context supports looking, not leaning. It gives you a reason to keep Predilife on the page for another filing or another disclosure. It does not give you a reason to treat a EUR 25 buy as a decisive vote of confidence. That would be too much drama for too little capital at risk.
Risks, caveats, and what to watch next
The first risk is obvious. The filing value is about EUR 25, and that is a very small number in any market. A tiny buy can be a signal, but it can also be a token gesture. Without exact share quantity, execution price, or ownership change, you do not know how much economic weight the transaction carried. That is a real limitation, and it is why the filing should be read as part of a sequence rather than in isolation.
The second risk is historical. The cohort record for PDG/DG micro-cap names is weak, with a 32.4% 90-day win rate, a -5.61% average 90-day return, and a -14.1% average 365-day return. That does not condemn the trade, but it does warn you against overconfidence. The third risk is structural. Predilife is a micro-cap, and micro-caps can be thin, noisy, and vulnerable to narrative drift. A CEO can keep buying while the market does very little with the information.
What to watch is straightforward. Watch for whether Ragusa keeps filing additional buys beyond June 25. Watch for whether another insider joins the pattern, because a broader cluster would change the read. Watch for any company communication that gives the buying a business context, because right now the public record is silent. And watch the share price around 2.290 EUR, not because the price itself proves anything, but because micro-cap insider activity often matters most when it starts to line up with a change in the tape. Until then, this is a CEO who kept buying in June, a weak historical bucket, and a signal that deserves respect without surrendering judgment.
