Composite Market-Making Quotes from an Equity Signal and Perpetual Anchor
Code NautilusTrader
Summary
This plotting script illustrates a composite quoting framework for NVDA equity-linked perpetual trading. It uses an external equity mid-price as a directional signal and the perpetual market mid-price as the quote anchor. The quote center shifts with the equity signal residual and is adjusted against the trader's inventory; bid and ask prices are then placed around that center using a fixed half-spread. The panels visualize the reference prices, basis, signal and inventory adjustments, position limits, and quote bands. A separate panel demonstrates the operating-time constraint: the perpetual market remains active continuously while the external equity feed is available only during the cash session, so signal age increases outside those hours. The replay is deterministic synthetic data intended to explain the equations and session logic, not a record of live trading or a performance test. The script gives parameter examples but no evidence that the settings are profitable, robust, or suitable for live execution.
Key ideas
- The framework anchors quotes to the perpetual mid-price and shifts the center using an external equity price signal.
- Inventory is incorporated as an opposing adjustment to the signal-driven quote shift.
- A fixed half-spread determines bid and ask levels around the adjusted quote center.
- The external equity feed has a cash-session schedule while the perpetual market operates continuously.
- The plotted replay is synthetic and explains mechanics rather than demonstrating trading performance.
Tags
Full text
# render_panels.py
```py
"""
Render the Lighter NVDA composite market maker tutorial panels.
After building NautilusTrader from source, run these commands from the repository root:
make sync
uv run --project python --no-sync \
python docs/tutorials/assets/lighter_rwa_composite_mm/render_panels.py
The renderer uses deterministic replay data to show the quoting equations and
cash-session operating constraint without shipping a Databento capture or a
live Lighter account trace.
"""
from __future__ import annotations
from pathlib import Path
import numpy as np
import pandas as pd
import plotly.graph_objects as go
from plotly.subplots import make_subplots
from nautilus_trader.analysis.tearsheet import _write_figure
from nautilus_trader.analysis.themes import get_theme
OUT = Path(__file__).resolve().parent
SYMBOL = "NVDA"
HALF_SPREAD_BPS = 25
SIGNAL_SKEW_FACTOR = 55.0
INVENTORY_SKEW_FACTOR = 2.0
MAX_POSITION = 0.20
TRADE_SIZE = 0.05
THEME = get_theme("nautilus_dark")
TEMPLATE = THEME["template"]
COLORS = THEME["colors"]
PRIMARY = COLORS["primary"]
POSITIVE = COLORS["positive"]
NEGATIVE = COLORS["negative"]
NEUTRAL = COLORS["neutral"]
GRID = COLORS["grid"]
BACKGROUND = COLORS["background"]
def apply_layout(fig: go.Figure, title: str, height: int = 480) -> None:
fig.update_layout(
template=TEMPLATE,
title={"text": title, "x": 0.02, "xanchor": "left"},
paper_bgcolor=BACKGROUND,
plot_bgcolor=BACKGROUND,
font={"family": "Inter, system-ui, sans-serif", "size": 13},
margin={"l": 60, "r": 30, "t": 70, "b": 50},
height=height,
width=1200,
legend={"orientation": "h", "yanchor": "bottom", "y": 1.02, "xanchor": "right", "x": 1.0},
)
fig.update_xaxes(gridcolor=GRID, zeroline=False)
fig.update_yaxes(gridcolor=GRID, zeroline=False)
def build_replay() -> pd.DataFrame:
ts = pd.date_range("2026-06-17 13:30:00+00:00", periods=390, freq="1min")
x = np.linspace(0.0, 1.0, len(ts))
databento_mid = (
207.20
+ 2.10 * x
+ 0.85 * np.sin(2 * np.pi * x * 2.2)
+ 0.45 * np.exp(-(((x - 0.62) / 0.08) ** 2))
- 0.35 * np.exp(-(((x - 0.22) / 0.05) ** 2))
)
basis_bps = 5.5 * np.sin(2 * np.pi * x * 3.0 + 0.4) + 10.0 * np.exp(
-(((x - 0.66) / 0.055) ** 2),
)
lighter_mid = databento_mid * (1 + basis_bps / 10_000.0)
position = np.zeros(len(ts))
position[55:125] = TRADE_SIZE
position[125:205] = TRADE_SIZE * 2
position[205:260] = TRADE_SIZE
position[260:335] = -TRADE_SIZE
baseline = databento_mid[0]
signal_residual = databento_mid / baseline - 1.0
signal_shift = SIGNAL_SKEW_FACTOR * signal_residual
inventory_shift = INVENTORY_SKEW_FACTOR * position
total_shift = signal_shift - inventory_shift
quote_center = lighter_mid + total_shift
half_spread = lighter_mid * (HALF_SPREAD_BPS / 10_000.0)
return pd.DataFrame(
{
"ts": ts,
"databento_mid": databento_mid,
"lighter_mid": lighter_mid,
"basis_bps": basis_bps,
"position": position,
"signal_residual_bps": signal_residual * 10_000.0,
"signal_shift": signal_shift,
"inventory_shift": inventory_shift,
"total_shift": total_shift,
"quote_center": quote_center,
"bid": quote_center - half_spread,
"ask": quote_center + half_spread,
"quote_shift_bps": total_shift / lighter_mid * 10_000.0,
},
)
def build_session_frame() -> pd.DataFrame:
ts = pd.date_range("2026-06-17 12:00:00+00:00", periods=337, freq="5min")
cash_open = pd.Timestamp("2026-06-17 13:30:00+00:00")
cash_close = pd.Timestamp("2026-06-17 20:00:00+00:00")
is_cash_session = (ts >= cash_open) & (ts <= cash_close)
stale_minutes = np.zeros(len(ts))
last_databento_ts = cash_open
for i, current in enumerate(ts):
if is_cash_session[i]:
last_databento_ts = current
stale_minutes[i] = max((current - last_databento_ts).total_seconds() / 60.0, 0.0)
return pd.DataFrame(
{
"ts": ts,
"lighter_active": 1.0,
"databento_active": is_cash_session.astype(float),
"stale_minutes": stale_minutes,
},
)
def panel_a_reference_overlay(df: pd.DataFrame) -> go.Figure:
fig = go.Figure()
fig.add_trace(
go.Scatter(
x=df["ts"],
y=df["databento_mid"],
mode="lines",
name=f"Databento {SYMBOL}.EQUS mid",
line={"color": PRIMARY, "width": 1.7},
),
)
fig.add_trace(
go.Scatter(
x=df["ts"],
y=df["lighter_mid"],
mode="lines",
name=f"Lighter {SYMBOL}-PERP mid",
line={"color": NEUTRAL, "width": 1.2},
),
)
fig.add_trace(
go.Scatter(
x=df["ts"],
y=df["ask"],
mode="lines",
name="Composite ask",
line={"color": NEGATIVE, "width": 0.9, "dash": "dot"},
),
)
fig.add_trace(
go.Scatter(
x=df["ts"],
y=df["bid"],
mode="lines",
name="Composite bid",
line={"color": POSITIVE, "width": 0.9, "dash": "dot"},
),
)
fig.add_trace(
go.Scatter(
x=df["ts"],
y=df["quote_center"],
mode="lines",
name="Quote center",
line={"color": "#eeeeee", "width": 1.2},
),
)
apply_layout(
fig,
f"{SYMBOL} composite quote center against Databento signal and Lighter anchor",
height=540,
)
fig.update_yaxes(title_text="USD")
return fig
def panel_b_signal_basis(df: pd.DataFrame) -> go.Figure:
fig = make_subplots(
rows=2,
cols=1,
shared_xaxes=True,
vertical_spacing=0.08,
row_heights=[0.52, 0.48],
subplot_titles=("Signal residual and Lighter basis", "Resulting quote-center shift"),
)
fig.add_trace(
go.Scatter(
x=df["ts"],
y=df["signal_residual_bps"],
mode="lines",
name="Databento residual",
line={"color": PRIMARY, "width": 1.4},
),
row=1,
col=1,
)
fig.add_trace(
go.Scatter(
x=df["ts"],
y=df["basis_bps"],
mode="lines",
name="Lighter basis",
line={"color": NEUTRAL, "width": 1.0},
),
row=1,
col=1,
)
fig.add_trace(
go.Scatter(
x=df["ts"],
y=df["quote_shift_bps"],
mode="lines",
name="Quote-center shift",
line={"color": POSITIVE, "width": 1.4},
),
row=2,
col=1,
)
fig.add_hline(y=0, line={"color": GRID, "width": 1}, row=1, col=1)
fig.add_hline(y=0, line={"color": GRID, "width": 1}, row=2, col=1)
apply_layout(
fig,
"Databento residual, Lighter basis, and composite quote-center shift",
height=620,
)
fig.update_yaxes(title_text="bps", row=1, col=1)
fig.update_yaxes(title_text="bps", row=2, col=1)
return fig
def panel_c_inventory_skew(df: pd.DataFrame) -> go.Figure:
fig = make_subplots(
rows=2,
cols=1,
shared_xaxes=True,
vertical_spacing=0.08,
row_heights=[0.45, 0.55],
subplot_titles=("Net position", "Price-unit skew terms"),
)
fig.add_trace(
go.Scatter(
x=df["ts"],
y=df["position"],
mode="lines",
name="Net position",
line={"color": PRIMARY, "width": 1.7, "shape": "hv"},
),
row=1,
col=1,
)
fig.add_trace(
go.Scatter(
x=df["ts"],
y=df["signal_shift"],
mode="lines",
name="Signal shift",
line={"color": POSITIVE, "width": 1.3},
),
row=2,
col=1,
)
fig.add_trace(
go.Scatter(
x=df["ts"],
y=-df["inventory_shift"],
mode="lines",
name="Inventory adjustment",
line={"color": NEGATIVE, "width": 1.3},
),
row=2,
col=1,
)
fig.add_trace(
go.Scatter(
x=df["ts"],
y=df["total_shift"],
mode="lines",
name="Total shift",
line={"color": "#eeeeee", "width": 1.4},
),
row=2,
col=1,
)
fig.add_hline(
y=MAX_POSITION,
line={"color": NEUTRAL, "dash": "dash", "width": 1},
row=1,
col=1,
)
fig.add_hline(
y=-MAX_POSITION,
line={"color": NEUTRAL, "dash": "dash", "width": 1},
row=1,
col=1,
)
fig.add_hline(y=0, line={"color": GRID, "width": 1}, row=2, col=1)
apply_layout(
fig,
f"Inventory skew with {TRADE_SIZE:.2f} {SYMBOL} trade size and {MAX_POSITION:.2f} cap",
height=620,
)
fig.update_yaxes(title_text=f"{SYMBOL}", row=1, col=1)
fig.update_yaxes(title_text="USD", row=2, col=1)
return fig
def panel_d_session_clock(df: pd.DataFrame) -> go.Figure:
fig = make_subplots(specs=[[{"secondary_y": True}]])
fig.add_trace(
go.Scatter(
x=df["ts"],
y=df["lighter_active"],
mode="lines",
name="Lighter RWA market",
line={"color": PRIMARY, "width": 1.5, "shape": "hv"},
),
secondary_y=False,
)
fig.add_trace(
go.Scatter(
x=df["ts"],
y=df["databento_active"],
mode="lines",
name="Databento EQUS.MINI feed",
line={"color": POSITIVE, "width": 1.5, "shape": "hv"},
),
secondary_y=False,
)
fig.add_trace(
go.Scatter(
x=df["ts"],
y=df["stale_minutes"],
mode="lines",
name="Signal age",
line={"color": NEGATIVE, "width": 1.3},
),
secondary_y=True,
)
fig.add_vrect(
x0="2026-06-17T13:30:00+00:00",
x1="2026-06-17T20:00:00+00:00",
fillcolor=PRIMARY,
opacity=0.10,
line_width=0,
annotation_text="regular session",
annotation_position="top left",
)
apply_layout(
fig,
"Lighter trades continuously while the Databento equity signal has a cash-session clock",
height=500,
)
fig.update_yaxes(title_text="active flag", range=[-0.05, 1.15], secondary_y=False)
fig.update_yaxes(title_text="minutes", secondary_y=True)
return fig
def main() -> None:
replay = build_replay()
session = build_session_frame()
panels = {
"panel_a_reference_overlay.png": panel_a_reference_overlay(replay),
"panel_b_signal_basis.png": panel_b_signal_basis(replay),
"panel_c_inventory_skew.png": panel_c_inventory_skew(replay),
"panel_d_session_clock.png": panel_d_session_clock(session),
}
for name, fig in panels.items():
path = OUT / name
_write_figure(fig, str(path))
print(f"wrote {path} ({path.stat().st_size / 1024:.1f} KB)")
if __name__ == "__main__":
main()
```Shown in full with attribution under the source's licence. Licence: LGPL-3.0
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.