Calendar Events

Integrate economic calendar events (FOMC, CPI) into trading pipelines. Time-decay sizing, event countdown, and pre-event risk reduction.

Ultra Feature. Requires an Ultra subscription. Get started at api.mathematicalcompany.com

Calendar Events

Major economic events (FOMC decisions, CPI releases) cause sharp moves in prediction markets. Horizon’s calendar module provides a live event feed, bundled FOMC/CPI dates through 2026, and pipeline functions for time-decay sizing and pre-event risk reduction. The feed maps event data to standard FeedSnapshot fields for seamless pipeline integration.

Overview

Event Feed

CalendarFeed tracks the nearest event with seconds-to-event countdown.

Time Decay

hz.time_decay() reduces position size as events approach.

Event Countdown

hz.event_countdown() exposes event data for downstream use.

Pre-Event Strategy

hz.pre_event_strategy() automatically reduces exposure before events.

---

Feed

CalendarFeed

Parses bundled or custom event data and tracks the nearest future event.

import horizon as hz

# Uses bundled FOMC + CPI dates automatically
feed = hz.CalendarFeed()

# Or with a custom API
feed = hz.CalendarFeed(api_url="https://api.example.com/events")

Parameter	Type	Default	Description
events_json	str	bundled dates	JSON array of event objects
api_url	str	""	Optional REST API for event data
interval	float	300.0	Poll interval in seconds

Bundled events include FOMC meeting dates and CPI release dates from 2024 through 2026.

FeedSnapshot mapping:

• price = urgency score (0.0 to 1.0, higher as event approaches)
• bid = seconds to next event
• ask = event type code (FOMC=1.0, CPI=2.0, NFP=3.0, GDP=4.0, EARNINGS=5.0)
• volume_24h = days to next event
• source = "calendar:EVENT_NAME"

---

Pipeline Functions

hz.time_decay

Reduce position size as events approach.

decay = hz.time_decay(
    calendar_feed="calendar",
    max_days=30.0,
    decay_type="linear",  # or "exponential", "sqrt"
)

Parameter	Type	Default	Description
calendar_feed	str	"calendar"	Feed name for calendar data
max_days	float	30.0	Days at which decay factor reaches 1.0
decay_type	str	"linear"	Decay curve: linear, exponential, or sqrt

Returns each cycle:

{
    "time_decay_factor": 0.67,   # multiply position size by this
    "days_to_event": 20.0,
    "event_urgency": 0.33,
    "event_type_code": 1.0,      # FOMC
}

hz.event_countdown

Expose raw event countdown data for downstream pipeline use.

countdown = hz.event_countdown(calendar_feed="calendar")

Parameter	Type	Default	Description
calendar_feed	str	"calendar"	Feed name

Returns each cycle:

{
    "seconds_to_event": 172800.0,
    "event_name": "FOMC 2026-03-18",
    "event_type_code": 1.0,
    "urgency": 0.33,
    "days_to_event": 2.0,
}

hz.pre_event_strategy

Automatically reduce exposure as events approach, with a linear ramp.

pre_event = hz.pre_event_strategy(
    calendar_feed="calendar",
    reduce_hours=24.0,      # start reducing 24h before
    reduce_factor=0.5,      # minimum 50% of normal size
)

Parameter	Type	Default	Description
calendar_feed	str	"calendar"	Feed name
reduce_hours	float	24.0	Hours before event to start reducing
reduce_factor	float	0.5	Minimum size multiplier at event time

Returns each cycle:

{
    "size_multiplier": 0.75,
    "pre_event": True,
    "seconds_to_event": 43200.0,
    "event_name": "FOMC 2026-03-18",
}

---

Examples

Event-Aware Trading

import horizon as hz

def my_strategy(ctx):
    if ctx.feed is None or ctx.feed.price <= 0:
        return []
    multiplier = ctx.params.get("size_multiplier", 1.0)
    size = int(10 * multiplier)
    if size < 1:
        return []
    return hz.quotes(fair=ctx.feed.price, spread=0.04, size=size)

hz.run(
    name="event_aware",
    exchange=hz.Polymarket(),
    feeds={
        "market": hz.PolymarketBook("will-fed-cut-rates"),
        "calendar": hz.CalendarFeed(),
    },
    pipeline=[
        hz.pre_event_strategy(reduce_hours=24, reduce_factor=0.3),
        hz.time_decay(max_days=7, decay_type="sqrt"),
        my_strategy,
    ],
)

Custom Event Data

import json
import horizon as hz

events = [
    {"name": "Earnings AAPL", "timestamp": 1711900800, "type": "earnings"},
    {"name": "Custom Event", "timestamp": 1712000000, "type": "other"},
]

feed = hz.CalendarFeed(events_json=json.dumps(events))

---

Event Type Codes

Code	Event Type
1.0	FOMC
2.0	CPI
3.0	NFP (Non-Farm Payrolls)
4.0	GDP
5.0	Earnings
6.0	PPI
7.0	PCE
99.0	Other