“I’m Bud. Let’s look at the numbers.”
An original WarRoom Labs research character — sharp on the tape, honest about the record. He never trades for you and never sugar-coats a loss.
A ranked basket you can inspect, step by step.
Bud is a deterministic research engine — not a stock tipster and not an adviser. Here’s exactly how the Bud Basket is ranked, published, held on its review cycle, and kept on the record, with nothing hidden.
- Step 1
Market data enters through controlled providers
Supported stock, ETF, and official U.S. Treasury data flow in through licensed providers. Bud never trades and never touches your brokerage.
- Step 2
Bud freezes a dated input snapshot
Before evaluating anything, Bud captures a timestamped snapshot of the inputs and hashes it — so every basket is tied to exactly what was known at that moment.
- Step 3
A deterministic model ranks the universe cross-sectionally
The Bud Basket v2.6 model scores the securities relative to each other and ranks them. The same inputs always produce the same ranking — no black box, no improvisation.
- Step 4
Bud publishes a ranked basket with defined weights
Top ranks receive defined, long-only weights that sum to 100%. It is disciplined ranking for research — not a buy/sell recommendation or a trade signal.
- Step 5
The basket is held through a defined review cycle
Each basket is published, held, and reviewed on a fixed schedule. Between reviews it stays frozen — no quiet mid-cycle edits.
- Step 6
Every snapshot becomes part of the record
Ranks, weights, sessions, and each change are preserved and timestamped. What was ranked, when it changed, and the full history stay on the record.
- Step 7
Why not up/down predictions?
Earlier binary prediction approaches were not adopted for launch because they did not demonstrate reliable edge versus baseline. Bud therefore centers on disciplined ranking, a defined review/holding cycle, transparency, history, and accountability.
Imaginary instrument “XMPL” on 2026-01-15: the engine freezes the snapshot, computes a score of +42 (bullish-leaning), model/data confidence 61, and a 1-standard-deviation expected range of −4.8% to +5.1% over a 20-day horizon. It writes an immutable receipt. Twenty trading days later it’s graded against those exact frozen terms. These numbers are invented to explain the flow — they are not a real call or a forecast.
A model volatility band (about one standard deviation of realized volatility over the horizon). It describes how much the instrument has typically moved — not a guaranteed range, floor, ceiling, or promise.
Model/data confidence — how complete and internally consistent the inputs are. It is not the probability of profit and not a measure of how likely the call is to “win.”
Every outcome stays visible
Wins, losses, flats, voids, corrections, and abstentions all remain on the record. Nothing favourable is kept and nothing unfavourable is buried.