LTQS
LTQS
THE GROVE
Sapling
Stage 1 / 9
0% grown · each analysis waters it

FEATHERSCALE

Quantitative Stock Ranking & Screening Model
--
Composite Score
Alpha--
Macro--
Forecast--

Built for the long term.

Scroll down

What we do

LTQS is an independent research project. Two rule-based engines trade US stocks with real money, in public, losses included.

Gen 01 · Live since Feb 2026

DRAUPNIR

Our first engine. A concentrated portfolio of up to 20 S&P 500 stocks, listed trade by trade from its first day; since late July 2026, every trade is confirmed with the broker.

Gen 02 · Live since Jun 2026

ARGO

Our second engine, built from a different starting point. Live since June 2026 on earlier builds; the current build has run since 27 July 2026, and its record is the one we publish.

Each engine trades its own brokerage account. Returns come from the balance the broker reports each day, trades are listed rebalance by rebalance, and the drawdowns stay in.

Portfolio PerformanceLive record

DRAUPNIR’s live account since funding in January 2026, against the S&P 500 price index. ARGO’s live record is on the performance page. Past performance does not guarantee future results.

Every trading day we collect prices, SEC filings, index changes, corporate actions and the exchange calendar, cross-checked, with renames and mergers tracked.

Model Release NotesHow the engines evolved

FeatherScale is the stock scoring model LTQS began with. It weighs six families of evidence into one score. Try a ticker.

How it runs

The system stays on watch through every session, and the engines act only when their rules say so. Most days, nothing trades.

The daily cycle · New York time
  1. ARGO09:30 – 16:01In session, through the closing print
  2. DRAUPNIR09:30 – 16:00In session
  3. Data16:00 – 20:00Company filings, prices and index changes
  4. Records16:01 – 17:00Reconciled against the broker
Windows shift with the close on early-close days.

Concentrated portfolios swing hard. What the rules control is the process.

  • Exposure and volatility limits
  • Regime guards that cut risk when markets turn
  • Data checks that stop a run on stale or incomplete prices
  • A daily reconciliation with the broker
Portfolio PerformanceDrawdowns, on the record
Live record
FEATHERSCALE · LTQS

A Quantitative Framework,
Forged From Discipline.

Every term orbiting above is a pillar of the FeatherScale engine. Six families of signals, spanning valuation, risk-adjusted return, market microstructure, distribution statistics, derivatives and sentiment, converge into a single conviction score.

01VALUATION
02RISK-ADJUSTED RETURN
03EXECUTION
04DISTRIBUTION
05DERIVATIVES
06SENTIMENT
THE ENGINES

Two Generations.
One Discipline.

GEN 01 · LIVE
DRAUPNIR

Our first portfolio engine has traded live capital daily since February 2026. Part of its stock screening runs on FeatherScale, the scoring model demonstrated above. Since late July 2026 every order it fills is published on this site, rebalance by rebalance.

GEN 02 · LIVE
ARGO

Our second generation engine reads the market from a different starting point than DRAUPNIR. Its current build, v11r, has traded live capital on its own account since 27 July 2026, and both engines are plotted side by side on the performance page.

READ THE RELEASE NOTES →
LIVE, NOT BACKTESTED

Numbers From a Real
Brokerage Account.

--
STRATEGY RETURN
--
S&P 500
--
DAYS TRACKED
--
MAX DRAWDOWN

DRAUPNIR, synchronised from the live brokerage ledger after each trading day. ARGO's live record sits beside it on the performance page, where the simulated part of its track is drawn as a dashed line and labelled as such. Past performance does not guarantee future results.

VIEW LIVE PERFORMANCE →
🌱
UNDER CONSTRUCTION
The Grove is being replanted —
此主题正在搭建中,部分视觉与功能即将更新。