Evidence-first desk
DragonPulse publishes every setup with its entry zone, its invalidation level, and the evidence behind it — before it triggers. Then it records what actually happened, whether that flatters us or not.
Research and paper trading only
Three made money, two were stopped out, one never triggered, one is in trade, one is still forming. None of them were removed from the list after the fact.
The method
This is the state machine behind the portal, not a diagram drawn for a landing page. A stage is only finished once it has left a durable artefact the next one can read.
Publications, filings, social velocity, and prediction markets land in one store. Nothing is dropped for being inconvenient.
A normalized claim, with its sourceA claim only counts once independent sources agree. Conflicting evidence stays marked as conflicting.
An evidence score, and its inputsEntry zone, invalidation, and targets are timestamped before the setup can trigger.
A timestamped ticketForming, active, invalidated, expired, closed. States change in the open, never by deletion.
A full state historyExpectancy, hit rate, and drawdown are computed from the complete record, losers included.
A settled row in the journalThe hard part was never finding setups. It was deciding what the system is not allowed to claim — and then enforcing that in code rather than in a style guide.
Issued before entry · amendments appended, never overwritten
What lands
You never receive a symbol and a direction on their own. Each ticket carries the zone you are allowed to enter, the level that ends the argument, the size that keeps a loss survivable, and the evidence that justified publishing it at all.
If the thesis breaks before the entry fills, the ticket is marked invalidated and stays in the record. That is the part most services delete.
Read the full methodThe record
412 setups over twelve months on a paper account with modelled fills. More than half do not reach target, and the expectancy is still positive at +0.38R. That gap is the whole method — it only works if you take the losers too.
The curve is the same computation the journal runs inside the portal. It goes down as well as up: the deepest peak-to-trough leg is −7.9R, and it is drawn here because a curve without its drawdown is a sales chart, not a record.
Inside the portal
Every subscription opens the same portal. The panels below are drawn from the real workspace — the queue that keeps its failures, and the depth ladder that admits when its feed is stale.
Invalidated and expired setups keep their row permanently
Feed health is printed on every panel, including when it is stale
The queue, with every setup's state visible: forming, active, invalidated, expired, closed.
Normalized claims with their sources, corroboration level, and prediction-market context.
Charting and live depth, with an honest feed-health state printed on every panel.
Expectancy, hit rate, and drawdown from the complete record — sample size always shown.
Subscriptions
Not by how many features can be withheld. A quarterly investor and an intraday trader need genuinely different work; everyone opens the same portal, and the plan decides which surfaces unlock and how fast new work reaches you.
Position research for multi-month conviction
Multi-day setups with defined risk
Intraday structure, depth, and session planning
Pricing to be announced — the plan structure is settled, the commercial terms are not. Nothing can be subscribed to until they are published.
Direct answers
Answered here in full rather than summarised. If the one you need is missing, it gets added to the FAQ page rather than sent to you privately.
No. We publish research: a setup, the evidence behind it, the level that invalidates it, and the size that keeps a loss survivable. The decision to take it is yours, and we are not licensed to make it for you.
Investing runs 2–5 researched theses a month, Swing Trading 5–12 setups a week, and Day Trading publishes continuously through market hours. Quiet periods stay quiet — we do not manufacture setups to hit a quota.
Yes, and they are the reason the record is worth reading. Invalidated and expired setups keep their place in the queue and their row in the journal permanently.
No. Outcomes are recorded on a paper account with modelled fills. That is stated on every page showing results, and it is why the numbers are labelled as research performance rather than returns.
The desk
Questions about the method or a specific setup get a real answer from the person who published it. Requests for a personal recommendation get a polite refusal, because we are not licensed to give one.
The reliable route for anything about subscriptions, billing, or access.
desk@dragonpulse.techSubscribers raise a question on a specific setup, so the answer sits next to the evidence.
See subscriptionsWithin one trading session. Questions about a live setup are answered first.
Mon–Fri · 08:00–18:00 CETOpen the portal, work through the queue, and read the journal. If the evidence does not convince you, that is the correct outcome.
Evidence-first desk