+First time on the top traders desk?
The desk runs its own numbers and publishes its calls next to the account that trades them. The short route:
TradFi Daily Read beta
Candidate status, and the measurement is not flattering. Over the venue's whole session history the rule returns -1.8% on 7 trades while simply holding the eligible universe returns +5.7%. Shuffle the names and keep everything else (the same days, the same sizes) and 18 of 200 random draws do better. Split the rule in two and the blame lands on the gate: the gate alone with no selection returns -9.5%, so timing subtracted while name selection clawed back +7.7 pp on top of it. The sample is months, not years: 210 session bars on the oldest market, 116 on the median one. That is a statement about the sample as much as about the rule, and both belong here.
What is closest
Ranked by distance to its own high over the window each market actually has. A market has to be near that high, moving on more volume than usual, in a trend that is measurable, and stronger than the universe it sits in. The context column says which of those it failed first.
| #? | market? | context? | 30d? | % of high? | funding? |
|---|---|---|---|---|---|
| 1 | COPPER commodity | volume too thin | +6.2% | 97.1% | +5.5% |
| 2 | EUR FX | weaker than the universe | +2.4% | 96.7% | +3.4% |
| 3 | JPY FX | volume too thin | -1.8% | 96.7% | +0.0% |
| 4 | EWJ stocks | volume too thin | +0.4% | 96.4% | +5.5% |
| 5 | XYZ100 index/ETF | volume too thin | -1.6% | 95.3% | +5.5% |
| 6 | MSFT stocks | volume too thin | +25.5% | 92.8% | +5.5% |
| 7 | AMZN stocks | volume too thin | +5.7% | 90.4% | +5.5% |
| 8 | PLTR stocks | volume too thin | +42.0% | 90.4% | -9.7% |
| 9 | AAPL stocks | not near its high | -1.6% | 89.9% | +5.5% |
| 10 | NVDA stocks | not near its high | +2.2% | 89.4% | +5.5% |
| 11 | TSM stocks | not near its high | -3.8% | 87.6% | +15.3% |
| 12 | GOOGL stocks | not near its high | -3.4% | 84.4% | +5.5% |
| 13 | RIVN stocks | not near its high | -3.1% | 83.9% | +12.9% |
| 14 | GOLD commodity | not near its high | +12.0% | 82.4% | +23.6% |
| 15 | AMD stocks | not near its high | -15.7% | 80.8% | +5.5% |
| 16 | EWY stocks | not near its high | -2.8% | 80.8% | +5.5% |
| 17 | BRENTOIL commodity | not near its high | +21.4% | 78.4% | -11.7% |
| 18 | HOOD stocks | not near its high | -2.7% | 77.3% | +5.5% |
| 19 | PALLADIUM commodity | not near its high | +5.3% | 77.0% | +5.5% |
| 20 | MU stocks | not near its high | -2.0% | 76.8% | +6.4% |
| 21 | SMSN stocks | not near its high | +1.6% | 75.5% | -1026.8% |
| 22 | META stocks | not near its high | -17.3% | 74.3% | +5.5% |
| 23 | CL commodity | not near its high | +21.2% | 73.4% | +5.5% |
| 24 | TSLA stocks | not near its high | -10.8% | 72.8% | +5.5% |
| 25 | NFLX stocks | not near its high | +8.4% | 72.1% | +5.5% |
| 26 | SNDK stocks | not near its high | -17.5% | 67.1% | +27.5% |
| 27 | COIN stocks | not near its high | +18.5% | 66.6% | +27.8% |
| 28 | PLATINUM commodity | not near its high | +15.3% | 66.6% | +5.5% |
| 29 | BABA stocks | not near its high | +5.8% | 65.8% | +6.9% |
| 30 | URNM stocks | not near its high | +9.0% | 65.7% | +42.4% |
| 31 | CRCL stocks | not near its high | +33.0% | 63.3% | +5.5% |
| 32 | SKHX stocks | not near its high | -15.4% | 63.2% | -705.6% |
| 33 | CRWV stocks | not near its high | -1.4% | 63.1% | -21.1% |
| 34 | INTC stocks | not near its high | -18.1% | 62.1% | +5.5% |
| 35 | MSTR stocks | not near its high | +28.0% | 61.3% | +5.5% |
| 36 | NATGAS commodity | not near its high | -4.2% | 61.1% | +5.5% |
| 37 | USAR stocks | not near its high | +3.8% | 58.9% | +5.5% |
| 38 | ORCL stocks | not near its high | +4.1% | 58.6% | +5.5% |
| 39 | SILVER commodity | not near its high | +15.4% | 56.7% | +5.5% |
| 40 | HYUNDAI stocks | not near its high | -3.5% | 55.7% | -359.4% |
Read it as: the board is the shortlist, not the position. A market only becomes one when the gate is open and it clears every check in the rule.
How this was built
The method sits under the answer on purpose: the verdict first, then the machinery that produced it, for anyone checking the work.
The research path
Data and hypothesis → candidate → measurement → indicator. This one is at the measurement step and stays there: the sample is months, so a Lab verdict here is a reason to watch it live, never a result.
Timeframe and execution
Breadth inside this universe opens or shuts risk: is there any point looking for a long.
Picks the markets nearest their own high with volume and a measurable trend.
Checks the entry is not late, not crowded, and can be made at the price shown.
A position leaves on its own stop, its time limit or the gate, never on a trigger flip.
Regime
The gate is breadth measured inside this universe, not borrowed from crypto: how many of the 40 markets with enough history trade above their 50-day average. Below the line the book stands aside rather than shorting. There is no short side in this rule.
Breakout and the checks
Each one can only remove a candidate; none of them can add one.
- near its high — Within 10% of the high over the window this market actually has.
- volume confirms — Turnover at least 1.3x its own normal.
- trend measurable — ADX at or above 22.
- stronger than the universe — In the top third by 30-day return.
- not toxic — Not in the worst 15% by the toxicity reading.
- carry affordable — Funding paid over 14 days below 3.0% of the position.
- executable — Volume and open interest above $1M, premium under 1%, OI cap under 85% used.
What is not here
The measurement above is the whole of it: no multi-regime record, no out-of-sample window, no equity curve worth drawing from seven trades. Nothing here has been re-fitted to make it look better. The thresholds are the crypto radar's, and the one parameter that moved is how much history a market needs before it is admitted, recomputed in session bars rather than tuned against the returns table. The oracle behind these prices is checked against an independent source on 52 of 116 markets, 3 against on-chain twins and 49 equities against tokenised shares; the rest is unverified. Funding settles hourly here and is annualised x24x365, which is why this is a separate product from anything on the crypto side.
→ the tradfi tab of the screener · method and evidence · the crypto Daily Read
This is context, not advice. The rule is a candidate measured on months, not years, and the measurement is above. Sources: method and evidence · the tradfi tab.