Do Stablecoin Inflows Push Down Treasury Bill Yields?
A BIS working paper finds a 3.5 billion dollar stablecoin inflow lowers 3-month T-bill yields by 0.71 basis points at once and up to 4 within 10 days.
We find the Quasimodo pattern in real time, draw the scenario on the chart with entry, stop and target, and keep honest statistics, wins and losses alike.
Scroll to watch the chart build the Quasimodo pattern. Every step is something the engine can actually see on closed candles.
A left shoulder forms, then price makes a higher low. Together they mark the structure the pattern is going to break.
Price pushes above the left shoulder and sweeps the stops resting there. The engine records the head and the time it became known.
Price closes below the higher low: a break of structure, detected by open-source engines. Until it happens the candidate stays out of the idea list.
Price returns to the level of the left shoulder. We mark the zone and show the idea once it is confirmed; an idea expires 48 bars after confirmation.
Entry at the edge of the zone, stop behind the head plus 0.1 ATR, target at exactly 2R. A quality score counts the reasons for and against.
A left shoulder forms, then price makes a higher low. Together they mark the structure the pattern is going to break.
Each page answers one question about the market and shows its raw numbers, so you can disagree with it.
Live Quasimodo scenarios with entry, stop and target on 30+ coins across five timeframes from 5m to 1d.
Open ideas →Where activity is unusual right now, with the raw numbers next to every alert: volume, order book imbalance, large trades, big resting orders.
Open the radar →A live stream of forced liquidations on Binance futures by coin and side. Binance sends at most one order per coin per second, and the page says so.
Open liquidations →A scanner for the break of a level and its retest, drawn on the chart. It is a marking tool, and its test on history is published on the page.
Open the scanner →The 4h trend state of each coin by a vote of three rules, plus a log of state changes written forward in time.
Open the board →Every forecast is frozen with its levels at the moment it is made, so results can be checked later without hindsight.
Open the journal →Mini pictures above are decorative illustrations; the numbers in the journal card are an example, not a real forecast.
Open an idea to see the chart with the scenario, the levels and the reasons for and against.
The same four rules for every hypothesis, so a good-looking chart cannot hide a weak test.
Every result is counted after a 0.08% fee per round trip. Without it, a small edge looks bigger than it is.
Rules are fixed first, then tested on a later period they have never seen. A rule that only works on the data it was picked on is dropped.
Each live forecast is saved with its levels at the moment it is made, so results can be checked without hindsight.
The methodology page shows the results of every test, the weak ones next to the strong ones. We do not promise returns.
Long reads built on peer-reviewed research and primary sources, every quotation checked against the original.
A BIS working paper finds a 3.5 billion dollar stablecoin inflow lowers 3-month T-bill yields by 0.71 basis points at once and up to 4 within 10 days.
S&P 500, 1990–2007: implied variance averaged over twice realized variance. The gap predicted quarterly returns, but it is not free money.
Under a random walk a stop-loss always lowers expected return; with momentum it can raise it. Kaminski and Lo derive the condition and test it.
The scanner is open and free. The methodology shows what works and what does not.
Open live ideas→OrderBlock.net is a market structure scanner for crypto futures. It reads public Binance USD-M futures data, finds swing highs and lows, breaks of structure (BOS), changes of character (CHoCH), order blocks (OB), fair value gaps (FVG), equal highs and lows and liquidity sweeps, and assembles them into one readable scenario. The main pattern we track is the Quasimodo (QM): a left shoulder, a structural low or high, a head that takes liquidity beyond the shoulder, and a break of the structural level. When the break is confirmed, the left shoulder becomes the QML level, the zone where price often returns.
Every idea comes with a plan: an entry at the edge of the QML zone, a stop beyond the head, a target at 2R, the confirmations that support it and the contradictions against it. You see why the setup exists, not just an arrow.
In Crypto → Ideas you pick a coin and see the current scenario on the chart. All setups lists every active Quasimodo across the top 30 coins and five timeframes (5m, 15m, 1h, 4h, 1d), updated live. The Journal records every published forecast with its levels frozen at the moment of publication, so the statistics cannot be edited after the fact. Radar highlights unusual market activity: bursts of trades, order book imbalances, new listings. The Methodology page explains the rules of every calculation.
Most pattern websites show winners and hide losers. We do the opposite. Over a year of history on 30 coins (about 7,650 trades) the Quasimodo pattern in our definition gives roughly +0.05R per trade before fees and roughly −0.11R after a 0.08% round-trip fee. Dozens of filters did not survive an out-of-sample test. That is why we treat QM as context: a map of levels where liquidity sits and where the market already reacted, not as a trading system that prints money. The decision and the position size are always yours.
Structure detection comes from open-source engines (MIT), charts from TradingView Lightweight Charts (Apache-2.0), and Radar from open-source detectors. Our own work is the Quasimodo recipe, the plan, the journal and the research.
No. It shows market structure and a scenario with entry, stop and target, plus the reasons for and against it. It does not place trades and does not promise profit.
From public Binance USD-M futures market data (candles, trades, order book). No exchange keys are used, and nothing is invented: if data is missing, the page says so.
The top 30 USDT perpetual contracts by volume plus BTC, ETH and SOL, on 5m, 15m, 1h, 4h and 1d.
Yes, the scanner, ideas, journal and research pages are free to read.
No. Everything is readable without registration; the language and a few display settings are stored in a cookie.