Risk · research · record

Financial Risk and Research Terminal Built for Investment Desks

Risk that updates as your book moves, and research you can check.

6 asset classes20 yearspoint-in-time2T+ rows
Book a walkthrough →or start free with your own portfolio →
See what it does
I · the platform

Aegis for risk. Vega for research.

Both run on the same data, backtester and Strategy Builder.

Portfolio risk

Aegis

Risk that recalculates every time the portfolio changes.

Measures
VaR and CVaR, ex-ante volatility, beta, tracking error
Drivers
Factor exposure and each position's share of risk
Stress tests
Past crises, custom factor shocks, the current regime
Limits
Breach alerts and a constrained optimizer
More on Aegis →
Research

Vega

Research answers with a chart and a source for each claim.

Sources
Filings, earnings call transcripts, our data lake and the web
Output
The chart and the query behind it
Strategies
A backtestable strategy from your idea
Gaps
Says so when it can't find a source
More on Vega →
01Buildwrite the rules once
02Backtestgraded on data it never saw
03Researcha source for each claim
04Manage riskrecalculated as the book moves
II · your book

Start with a book you actually hold.

Paste your tickers. It runs in your browser, no account needed.

Separate tickers with spaces or commas, add optional weights like AAPL 20, or paste from Excel or Bloomberg (AAPL US Equity works).

AEGIS · 7 names
● local
VaR 95% · 1-day−2.46%
CVaR
−3.09%
Beta
1.27
Ann. vol
23.8%
Pos / eff
7 / 7.0
Factor exposure
market
+1.27
momentum
+0.39
quality
+0.48
value
−0.38
size
−0.39
Stress · scenario P&L
2020 Covid crash−33.2%
2022 rate shock−44.0%
Momentum unwind−25.1%
Rates +100bp−7.4%
Ask Vegacited answer, chart and query →

Illustrative betas, not live market data. Computed in your browser, so your holdings never leave this page.

III · Vega

Ask the question you’d give an analyst.

Vega runs the query, draws the chart and cites every claim.

Snapshot from Jul 1, 2026.

vega · researchchart + sources
Ask

Momentum has outperformed value by 11.4pp over twelve months, but the spread sits 2.1σ above its twenty-year mean. The last three episodes this wide mean-reverted inside two quarters.

momentumvaluetotal return · rebased
1 Fama–French factor library2 TradePolaris lake · 2006–20263 query.sql
Suggested nextBacktest the reversalShow it beta-neutralWhat breaks this?
IV · the gates

Walk-forward gates that reject curve-fits.

Graded only on data the strategy never saw, with an overfitting check that gets stricter on every rerun.

How the checks work →
WALK-FORWARD · 6 foldsexample: 20-day breakout, S&P 500
FOLD 1+0.71
FOLD 2+0.44
FOLD 3−0.12
FOLD 4+0.58
FOLD 5+0.39
FOLD 6+0.62
trained ongraded on, never seen in training
Deflated Sharpe 0.93 · passUniverse integrity · passTrade-block stability 74% · fail

Not eligible for paper trading. Only 74% of rolling 50-trade windows had a positive Sharpe, and the bar is 80%.

V · the lake

Use the data that was
available at the time.

Where a dataset supports it, set an as-of date and see only what was public on that day.

2026-07-01

Only data known by this date is shown, including inactive and delisted companies where the source has them.

Coverage
equities30k+
optionschains
futurescontinuous
fxmajors + crosses
cryptospot + perps
macro100+
filings13F · F4
VI · Strategy Builder

Write the rules once
and use them everywhere.

The backtest and the risk views read the same rules.

quality-momentum.spec v12 · data pinned
  1. UniverseS&P 1500ADV > $5mex-financials
  2. Signal12−1 momentum× gross profitabilityz-scored in sector
  3. Constructsector, beta, size neutralrisk parity3% name cap
  4. Executemonthly, 5-day tranchesspread + 12bp impact8% trailing exit
  5. Validatewalk-forward, 6 foldsDeflated Sharpe500 trials
Example runwalk-forward · 6 folds
strategy, out of sample benchmark
OOS Sharpe
1.08
Max drawdown
−11.4%
Turnover
18%/yr
  • Deflated Sharpe 0.96
  • Universe integrity
  • Trade-block stability 86%
Eligible for paper trading
Or start from a template

Quality–momentum, beta-neutral

Long high-quality winners, short low-quality names, with market beta neutralized.

monthlylong/short

Value + momentum composite

Holds cheap stocks and recent winners together in one portfolio.

monthlymulti-factor

Vol-targeted trend

Sizes momentum positions up or down to hold volatility steady.

dailyvol target

Low-vol quality, regime-aware

Low-volatility, high-quality names, with risk cut when the market regime turns.

monthlydefensive
VII · the rest of the platform

Everything else runs on the same data.

Monitor

Markets & monitoring

Quotes, breadth, movers, and alerts that fire on your rules.

Brief

Daily market brief

Movers and a sourced market summary, ready before the open.

Data

The data lake

Six asset classes, with each dataset's dates and as-of support shown.

Build apps

Apps

Screeners, 13F views and what-if charts, or describe the app you need.

Build & prove

Strategy Builder

Write the rules once and test them before they touch your portfolio.

Validate

Walk-forward backtesting

Graded only on data it never saw, after costs, with an overfitting check that tightens each rerun.

AI workspace

TradePolaris MCP

TradePolaris data and your portfolio inside your AI workspace, read-only.

Data API

Kappa Data API

Market data by API, with dataset versions you can pin.

VIII · the access layer

Bring the data into
your code and AI workspace.

Pin a dataset version so the same request returns the same result later.

Kappa Data API

Results you can rerun exactly

Corrections become new versions, so a pinned version never changes.

# as_of goes on the request itself
GET /v1/curated/fundamentals
    ?universe=sp1500
    &as_of=2022-06-30          # what was knowable that day
    &version=v41               # pinned; v41 never changes
    &cursor=eyJvIjoyMDAwfQ

# -> 200  x-dataset-version: v41   x-rows: 1500   x-complete: true
Kappa Data API →
TradePolaris MCP

TradePolaris inside your AI workspace

Look up assets and your portfolio mid-conversation, read-only.

search_assets
finds the right ticker before answering
get_portfolio
adds your watchlists, strategies or positions
TradePolaris MCP →
IX · who it’s for

For investment desks and anyone with a portfolio.

Investment desks

Live risk and a decision record for your book.

  • Aegis. VaR, stress tests and factor exposure, recalculated as the book moves.
  • The record. Every figure keeps the method and data version behind it.
How desks use it →
Individual investors

See which holdings carry your risk.

  • Ask Vega. Get a chart and a source for each claim.
  • Test a template. Backtest it, then paper-trade it before real money.
How people use it →
X · pricing

From a free plan to multi-book desks.

Charts and market data stay free, and every paid run shows its credit cost first.

Investment desks$1,000/mo+

Multi-book live risk, the decision record, onboarding and priority support

Aegis$249/mo

Live portfolio risk, stress tests, optimization and alerts

Pro$49/mo

1,000 credits · paper trading · 10 saved apps

Free$0

Vega previews · preview backtests · free credits monthly · no card

See full pricing →

Backtested and walk-forward results are hypothetical, do not represent actual trading, and are not indicative of future results. TradePolaris does not provide investment advice.