riskmodels-py 0.3.0

Python SDK for the RiskModels API (ERM3 factors, hedge ratios, explained risk)

riskmodels-py

Published on PyPI as riskmodels-py (import package riskmodels).

Python SDK for the RiskModels API (ERM3 factor model: hedge ratios, explained risk, batch portfolio analysis).

Install

pip install riskmodels-py
# Optional xarray panel interface:
pip install riskmodels-py[xarray]
# Optional Plotly/Matplotlib charts (L3 decomposition, portfolio cascades):
pip install riskmodels-py[viz]

From this monorepo:

cd sdk && pip install -e ".[dev]"

Requires Python 3.10+.

Local env files (monorepo / Next.js parity)

Put your API key in .env.local (recommended) or .env:

• Monorepo: RISKMODELS_API_KEY=rm_agent_... in the repository root .env.local (same file the Next.js app uses), and/or in sdk/.env.local if you only work under sdk/. Shell export values always win over files.

With pip install -e ".[dev]" or pip install "riskmodels-py[dotenv]", RiskModelsClient.from_env() loads .env then .env.local by walking up from the current working directory to the first folder that contains either file. Existing environment variables are never overwritten (shell exports and CI secrets win). Among files only, .env.local overrides .env for keys not already set.

Readme asset script: scripts/generate_readme_assets.py (run from repo root) also loads .env / .env.local from the repo root and from sdk/, so a key stored only in sdk/.env.local is picked up.

To call a local Next app (npm run dev), set RISKMODELS_BASE_URL=http://localhost:3000/api in .env.local (see repo root MAINTENANCE_GUIDE.md).

Quickstart

from riskmodels import RiskModelsClient

client = RiskModelsClient.from_env()  # RISKMODELS_API_KEY or OAuth client env vars; optional .env / .env.local (see below)
df = client.get_metrics("NVDA", as_dataframe=True)
pa = client.analyze({"NVDA": 0.5, "AAPL": 0.5})  # alias for analyze_portfolio
print(pa.portfolio_hedge_ratios["l3_market_hr"])
print(pa.to_llm_context())

Namespaces + charts (v0.3+)

With pip install riskmodels-py[viz]:

from riskmodels import RiskModelsClient

client = RiskModelsClient.from_env()

# Single- or multi-name L3 risk (σ-scaled horizontal bars)
fig = client.stock.current.plot(
    style="l3_decomposition",
    sigma_scaled=True,
    tickers=["AAPL", "MSFT", "NVDA", "AMZN", "GOOG", "META", "TSLA"],
)
fig.show()

MAG7 — L3 explained risk (variance shares to 100%) — same chart as RM_ORG/demos/article_visuals.py “risk DNA” variant (1): horizontal stacked bars, x-axis is fraction of variance, right-rail annotations use ER / systematic % (not σ-scaled). The SDK Plotly figure defaults to theme="light" to match the article / matplotlib reference (white plot area, vertical x-grid, left y-spine, airy bar spacing). Pass theme="terminal_dark" to plot_l3_horizontal / save_l3_decomposition_png for a dark canvas. Use GOOG as the canonical Alphabet symbol.

from riskmodels import RiskModelsClient, plot_mag7_l3_explained_risk, save_mag7_l3_explained_risk_png

client = RiskModelsClient.from_env()
fig = plot_mag7_l3_explained_risk(client)  # or pass tickers=[...]
fig.show()

save_mag7_l3_explained_risk_png(client, filename="mag7_l3_explained_risk.png")
# client.visuals.save_mag7_l3_explained_risk_png(filename="mag7_l3_explained_risk.png")

MAG7 portfolio risk cascade (weights ∝ market cap) — fetch the MAG7 list, pull market_cap from each get_metrics snapshot, normalize to weights, then plot a variable-width stacked L3 decomposition (bars touch; width ∝ weight):

import pandas as pd
from riskmodels import RiskModelsClient

client = RiskModelsClient.from_env()

mag7 = client.search_tickers(mag7=True)
caps: list[tuple[str, float]] = []
for sym in mag7["ticker"].astype(str).str.upper():
    snap = client.get_metrics(sym, as_dataframe=True)
    row = snap.iloc[0]
    cap = row.get("market_cap")
    if cap is None or (isinstance(cap, float) and pd.isna(cap)):
        continue
    caps.append((sym, float(cap)))

weights = pd.DataFrame(caps, columns=["ticker", "market_cap"])
weights["weight"] = weights["market_cap"] / weights["market_cap"].sum()
positions = weights[["ticker", "weight"]].to_dict("records")

fig = client.portfolio.current.plot(
    positions=positions,
    style="risk_cascade",
    sort_by="weight",
    include_systematic_labels=True,
)
fig.show()

# Optional: bundled PDF snapshot (premium endpoint; same weights)
# pdf_bytes = client.portfolio.current.pdf(positions=positions)

Portfolio risk snapshot PDF (POST /portfolio/risk-snapshot) — raw bytes without the visuals namespace:

from pathlib import Path
from riskmodels import RiskModelsClient

client = RiskModelsClient.from_env()
pdf_bytes, _lineage = client.post_portfolio_risk_snapshot_pdf(
    [("NVDA", 0.3), ("AAPL", 0.25), ("MSFT", 0.25), ("GOOGL", 0.2)],
    title="My sleeve",
)
Path("snapshot.pdf").write_bytes(pdf_bytes)

Stdlib-only example from repo root: python examples/python/portfolio_risk_snapshot_pdf.py (see docs/portfolio-risk-snapshot-runbook.md).

Publication PNGs (Kaleido)

With pip install riskmodels-py[viz] (Plotly + Kaleido), you can save finished charts in one or two lines. Defaults target ~1600×1000 logical pixels with scale=3 for crisp PNGs; optional dpi maps to Kaleido scale as dpi / 96 (96 dpi baseline). Optional figsize=(width_px, height_px) overrides width / height.

from riskmodels import RiskModelsClient, save_l3_decomposition_png, save_mag7_l3_explained_risk_png, save_portfolio_risk_cascade_png
from riskmodels.visuals import mag7_cap_weighted_positions

client = RiskModelsClient.from_env()

save_mag7_l3_explained_risk_png(client, filename="mag7_l3_explained_risk.png")

save_l3_decomposition_png(
    client,
    ticker="NVDA",
    filename="nvda_l3.png",
    title="NVDA — L3 risk decomposition",
    width=1600,
    height=1000,
    scale=3,
)

positions, weight_source = mag7_cap_weighted_positions(client)
save_portfolio_risk_cascade_png(
    client,
    positions=positions,
    filename="mag7_risk_cascade.png",
    subtitle=f"weights: {weight_source}",
)

# Equivalent facades on the client:
# client.visuals.save_l3_decomposition_png(filename="nvda_l3.png", ticker="NVDA")
# client.visuals.save_portfolio_risk_cascade_png(positions=positions, filename="mag7_risk_cascade.png")

# Optional: write NVDA + MAG7 L3 ER + MAG7 risk + MAG7 attribution PNGs to a folder (gallery helpers)
# from riskmodels.visuals import run_gallery_all
# run_gallery_all(client, output_dir="figures")

CLI (monorepo): from the repo root, with a key in .env.local or the environment:

python scripts/run_visuals_gallery.py -o figures
# Regenerate the readme figure into sdk/images/:
# python scripts/run_visuals_gallery.py -o sdk/images --charts mag7-l3-er

Use --charts nvda / mag7-l3-er / mag7-risk / mag7-attribution for a single chart.

mag7_cap_weighted_positions uses market_cap from each get_metrics snapshot when enough values exist; otherwise it falls back to documented illustrative early 2026 cap-share weights (see MAG7_SNAPSHOT_DATE_DOC in riskmodels.visuals.gallery). save_portfolio_attribution_cascade_png needs the batch returns panel (the helper requests it automatically).

Weights are holdings-style fractions (sum to 1); analyze_portfolio renormalizes if a ticker fails batch resolution. Ticker aliases (e.g. GOOGL→GOOG) follow the usual ValidationWarning path.

Readable metrics snapshot (CLI): after pip install -e ".[dev]" from sdk/, run:

export RISKMODELS_API_KEY=...
python examples/quickstart.py

Optional: RISKMODELS_QUICKSTART_TICKER=AAPL. The script prints L3 hedge ratios, explained risk, optional market fields, and the ERM3 legend. Use format_metrics_snapshot(row) in your own code for the same text layout from a get_metrics dict row.

Metrics + macro factor correlation (one row)

ERM3 snapshot plus macro_corr_* columns (Pearson/Spearman vs bitcoin, VIX, etc.). macro_corr_* values are return correlations, not dollar hedges (l3_market_hr) or variance shares (l3_residual_er). Use return_type="gross" for total-equity co-movement with macro; use "l3_residual" for the idiosyncratic sleeve vs macro.

from riskmodels import RiskModelsClient, to_llm_context

client = RiskModelsClient.from_env()
snap = client.get_metrics_with_macro_correlation(
    "NVDA",
    factors=["bitcoin", "vix"],
    return_type="l3_residual",
    window_days=252,
)
print(snap["macro_corr_bitcoin"].iloc[0], snap["l3_market_hr"].iloc[0])
print(to_llm_context(snap))

Raw macro factor series (no ticker)

get_macro_factor_series() calls GET /macro-factors — long table of factor_key, teo, return_gross for charts or offline checks. Requires the macro-factor-series scope on your API key (included in the SDK default scope string).

README and docs site PNGs (maintainers)

From the repository root (not sdk/), with a free-tier RISKMODELS_API_KEY:

export RISKMODELS_API_KEY='paste-your-key-here'
python scripts/generate_readme_assets.py

Use single quotes around the key. If you add an end-of-line comment, it must start with # (ASCII). Otherwise put the comment on its own line above.

This calls MAG7 POST /correlation and get_rankings, then writes assets/*.png and mirrors the same files to public/docs/readme/ for the Next.js docs hub. Commit both trees so GitHub and the portal stay in sync.

Recursive Visual Refinement (MatPlotAgent)

Generate professional financial visualizations through automated Vision-LLM feedback:

from openai import OpenAI
from riskmodels import RiskModelsClient

client = RiskModelsClient.from_env()
llm = OpenAI(api_key="...")

result = client.generate_refined_plot(
    plot_description="L3 risk decomposition stacked area chart for NVDA over 2 years",
    output_path="nvda_risk.png",
    llm_client=llm,
    max_iterations=5,
)

print(f"Generated in {result.iterations} iterations")
print(f"Saved to: {result.output_path}")

The generate_refined_plot method implements the MatPlotAgent Pattern:

1. Execute: Runs generated matplotlib code in a subprocess
2. Capture: Collects execution errors or output PNG
3. See: Sends the image to a Vision-LLM (GPT-4o or Claude 3.5 Sonnet)
4. Evaluate: LLM audits for overlapping text, legibility, legend accuracy, styling
5. Refine: Iterates until "COMPLETE" or max iterations reached

Requirements: pip install openai matplotlib (or anthropic)

Financial Color Standards (enforced automatically):

• Market Risk (SPY): Indigo (#4B0082)
• Sector Risk: Green (#228B22)
• Residual/Idiosyncratic: Gray (#808080)

Environment variables:

• RISKMODELS_API_KEY — static Bearer token, or
• RISKMODELS_CLIENT_ID + RISKMODELS_CLIENT_SECRET — OAuth2 client credentials (JWT ~15m),
• RISKMODELS_BASE_URL (default https://riskmodels.app/api),
• RISKMODELS_OAUTH_SCOPE (optional).

Agent-native helpers (vibe coding)

Use these so agents and humans never guess wire names or ERM3 semantics:

Tool	Purpose
client.discover()	Markdown or JSON digest (format="json", to_stdout=False): each method includes description, parameters (name, type, required, defaults, enums), returns, plus tool_definition_hints for Claude Desktop / MCP-style tool synthesis.
Ticker alias	Curated remap (e.g. GOOGL→GOOG) logs info and emits ValidationWarning (Warning: … Fix:) so agents refresh symbols.
to_llm_context(obj)	One call → Markdown tables + lineage + semantic cheatsheet + ERM3 legend (obj = DataFrame, PortfolioAnalysis, xarray.Dataset, or dict).
df.attrs["legend"]	Short ERM3 text on every tabular result from the client (same as SHORT_ERM3_LEGEND).
df.attrs["riskmodels_semantic_cheatsheet"]	Wire→semantic map + column hints + units (JSON + bullet list). Ground truth for field names.
df.attrs["riskmodels_lineage"]	JSON string: model version, as-of, factor set, universe size when the API sent them.
df.attrs["riskmodels_kind"]	What produced the frame (ticker_returns, portfolio_per_ticker, tickers_universe, …).
validate="warn" | "error" | "off"	ER sum + HR sign checks; Error: / Warning: … Fix: strings for self-correction.
attach_sdk_metadata / ensure_dataframe_legend	If you build a DataFrame manually, attach the same attrs so to_llm_context stays consistent.
build_semantic_cheatsheet_md()	Standalone cheatsheet string for custom prompts.

Semantic names (always use in code and LLM explanations): l3_market_hr, l3_sector_hr, l3_subsector_hr, l3_market_er, … — not raw V3 keys like l3_mkt_hr. Batch Parquet/CSV wire columns l1/l2/l3 are renamed to those three L3 component HR series (not “L1/L2/L3 model levels”). Full reference (repo root):

• SEMANTIC_ALIASES.md (same file as ../../SEMANTIC_ALIASES.md when this README is viewed inside the monorepo)
• docs/ERM3_ZARR_API_PARITY.md

Tip for agents: Prefer get_metrics(..., as_dataframe=True) so you get attrs; the plain dict return has no attrs.

Cursor: .cursorrules (math, naming, batch semantics).

PyPI distribution name vs import

• Install from PyPI: pip install riskmodels-py (and optionally pip install riskmodels-py[xarray]).
• Import in Python: from riskmodels import … — the distribution on PyPI is riskmodels-py; the package directory is riskmodels.

Core runtime dependencies are pandas, pyarrow, and httpx (HTTP). xarray is optional ([xarray] extra). The SDK does not depend on requests.

PyPI releases (maintainers)

Upload steps (version bump, build, twine, PyPI token format) are not in this public README. They are maintained in the private BWMACRO monorepo at docs/RISKMODELS_PY_PYPI_PUBLISHING.md — open that file from your internal BWMACRO clone.

License

Proprietary — same terms as RiskModels API access.