WASDE Ending Stocks: Tracking Revision Momentum

The Setup

The USDA's World Agricultural Supply and Demand Estimates (WASDE) report is released on or around the 12th of every month. It covers supply, demand, and ending stocks estimates for major crops — corn, soybeans, wheat, cotton, rice, and sorghum — for both the U.S. and global markets.

Ending stocks (also called carryout) is the estimated inventory remaining at the end of the marketing year. It's the single most important fundamental number in grain markets because it captures the net balance of supply and demand in one figure. Lower ending stocks mean a tighter market; higher ending stocks mean a looser market.

But here's what most analysis misses: the USDA doesn't get it right on the first estimate. The ending stocks number for a given crop year gets revised every month from May (first projection) through January of the following year (final). That's nine revisions, and the direction of those revisions — consecutive tightenings or loosenings — carries more information than any single month's number.

The Data

The Commodity Fundamentals API provides WASDE data as a time series of monthly estimates. Here's how to pull the ending stocks history for corn:

import requests

API_KEY = "your_api_key_here"
BASE = "https://commodityfundamentals.com/api/v1"

# Fetch WASDE corn ending stocks estimates
response = requests.get(
    f"{BASE}/commodities/corn/series",
    headers={"Authorization": f"Bearer {API_KEY}"},
    params={
        "source": "usda_wasde",
        "start": "2020-01-01",
        "end": "2026-01-01"
    }
)

data = response.json()
print(f"Retrieved {data['meta']['total']} monthly estimates")

Each observation is a monthly WASDE estimate with the ending stocks value in millions of bushels:

{
  "data": [
    {
      "date": "2025-10-10",
      "value": 1738.0,
      "commodity": "corn",
      "source": "usda_wasde",
      "unit": "million_bushels",
      "marketing_year": "2025/26"
    },
    {
      "date": "2025-11-11",
      "value": 1702.0,
      "commodity": "corn",
      "source": "usda_wasde",
      "unit": "million_bushels",
      "marketing_year": "2025/26"
    }
  ],
  "meta": {
    "total": 2,
    "request_id": "req_ghi789"
  }
}

The marketing_year field is critical — it tells you which crop year this estimate applies to. The October and November estimates above are both for the 2025/26 marketing year, and the drop from 1,738 to 1,702 million bushels is a tightening revision of 36 million bushels.

The Analysis

The revision momentum indicator tracks the direction of month-over-month changes in ending stocks for a given marketing year. The logic is simple: count consecutive revisions in the same direction.

import pandas as pd

df = pd.DataFrame(data["data"])
df["date"] = pd.to_datetime(df["date"])

def revision_momentum(group):
    """Calculate revision momentum for a single marketing year."""
    group = group.sort_values("date")
    group["revision"] = group["value"].diff()
    group["direction"] = group["revision"].apply(
        lambda x: "tighter" if x < 0 else ("looser" if x > 0 else "unchanged")
    )

    # Count consecutive same-direction revisions
    momentum = 0
    results = []
    prev_direction = None

    for _, row in group.iterrows():
        if pd.isna(row["revision"]):
            results.append(0)
            continue

        if row["direction"] == prev_direction:
            momentum += (1 if row["direction"] == "tighter" else -1)
        else:
            momentum = 1 if row["direction"] == "tighter" else -1
        prev_direction = row["direction"]
        results.append(momentum)

    group["momentum"] = results
    return group

# Apply per marketing year
df = df.groupby("marketing_year", group_keys=False).apply(revision_momentum)

# Display the 2024/25 corn marketing year
my = df[df["marketing_year"] == "2024/25"].sort_values("date")
print(f"\nCorn Ending Stocks — 2024/25 Marketing Year")
print(f"{'Date':<12} {'Stocks':>8} {'Revision':>10} {'Momentum':>10}")
print("-" * 44)
for _, row in my.iterrows():
    rev = f"{row['revision']:>+10.0f}" if pd.notna(row["revision"]) else "       ---"
    print(f"{row['date'].strftime('%Y-%m'):>12} {row['value']:>8.0f} {rev} {row['momentum']:>+10d}")

A momentum reading of +3 or higher (three consecutive tightening revisions) is a historically reliable signal that the USDA is chasing a fundamental shift — typically a yield disappointment, an export surge, or an ethanol demand surprise. Once the USDA starts revising in one direction, the bureaucratic inertia of their estimation process means the next revision is more likely to continue the trend than to reverse it.

Conversely, a momentum reading of -3 or lower (three consecutive loosenings) signals that supply is outrunning demand, and prices tend to grind lower as the market digests the growing surplus.

# Multi-commodity momentum comparison
commodities = ["corn", "soybeans", "wheat"]

for commodity in commodities:
    resp = requests.get(
        f"{BASE}/commodities/{commodity}/series",
        headers={"Authorization": f"Bearer {API_KEY}"},
        params={"source": "usda_wasde", "start": "2024-01-01"}
    )
    c_df = pd.DataFrame(resp.json()["data"])
    c_df["date"] = pd.to_datetime(c_df["date"])
    c_df = c_df.groupby("marketing_year", group_keys=False).apply(
        revision_momentum
    )

    latest = c_df.sort_values("date").iloc[-1]
    print(f"{commodity:<10} momentum: {latest['momentum']:>+3d}  "
          f"({latest['direction']}, {latest['date'].strftime('%b %Y')})")

When multiple grain markets show tightening momentum simultaneously, the signal is stronger. The 2020/21 marketing year was a textbook example: corn, soybeans, and wheat all showed 3+ consecutive tightenings starting in the September 2020 WASDE — foreshadowing the commodity rally that ran through May 2021.

Try It Yourself

Here's a complete script that pulls WASDE ending stocks data, calculates revision momentum across all major crops, and outputs a dashboard of the current signal state:

"""
WASDE Revision Momentum Dashboard
Tracks the direction of consecutive USDA ending stocks revisions
to identify sustained tightening or loosening trends.

Requires: pip install requests pandas
API key:  https://commodityfundamentals.com/registration/new
"""
import requests
import pandas as pd

API_KEY = "your_api_key_here"
BASE = "https://commodityfundamentals.com/api/v1"
COMMODITIES = ["corn", "soybeans", "wheat", "cotton"]
SIGNAL_THRESHOLD = 3  # consecutive same-direction revisions

def fetch_wasde(commodity, start="2020-01-01"):
    response = requests.get(
        f"{BASE}/commodities/{commodity}/series",
        headers={"Authorization": f"Bearer {API_KEY}"},
        params={"source": "usda_wasde", "start": start}
    )
    response.raise_for_status()
    df = pd.DataFrame(response.json()["data"])
    df["commodity"] = commodity
    return df

def calculate_momentum(df):
    df["date"] = pd.to_datetime(df["date"])
    df = df.sort_values("date")
    df["revision"] = df.groupby("marketing_year")["value"].diff()

    momentum = 0
    prev_direction = None
    results = []

    for _, row in df.iterrows():
        if pd.isna(row["revision"]):
            results.append(0)
            prev_direction = None
            continue

        direction = "tighter" if row["revision"] < 0 else "looser"
        if direction == prev_direction:
            momentum += (1 if direction == "tighter" else -1)
        else:
            momentum = 1 if direction == "tighter" else -1
        prev_direction = direction
        results.append(momentum)

    df["momentum"] = results
    return df

def main():
    print("WASDE Revision Momentum Dashboard")
    print("=" * 55)

    all_data = []
    for commodity in COMMODITIES:
        df = fetch_wasde(commodity)
        df = calculate_momentum(df)
        all_data.append(df)

        # Latest marketing year summary
        latest_my = df["marketing_year"].iloc[-1]
        my_data = df[df["marketing_year"] == latest_my].sort_values("date")
        latest = my_data.iloc[-1]

        signal = ""
        if latest["momentum"] >= SIGNAL_THRESHOLD:
            signal = "TIGHTENING"
        elif latest["momentum"] <= -SIGNAL_THRESHOLD:
            signal = "LOOSENING"

        print(f"\n{commodity.upper()} ({latest_my})")
        print(f"  Current stocks: {latest['value']:,.0f} {latest.get('unit', '')}")
        print(f"  Last revision:  {latest['revision']:>+,.0f}")
        print(f"  Momentum:       {latest['momentum']:>+d}  {signal}")

        # Show revision history for current marketing year
        print(f"  {'Month':<8} {'Stocks':>10} {'Change':>10}")
        for _, row in my_data.iterrows():
            rev = f"{row['revision']:>+10,.0f}" if pd.notna(row["revision"]) else "       ---"
            print(f"  {row['date'].strftime('%b %y'):<8} {row['value']:>10,.0f} {rev}")

    # Cross-commodity signal
    print("\n" + "=" * 55)
    tightening = sum(
        1 for df in all_data
        if df.iloc[-1]["momentum"] >= SIGNAL_THRESHOLD
    )
    print(f"Crops with tightening momentum: {tightening}/{len(COMMODITIES)}")
    if tightening >= 3:
        print(">> BROAD TIGHTENING — historically bullish for grain complex")

if __name__ == "__main__":
    main()

Get your free API key to run this analysis — sign up here. The free tier includes 1,000 API calls per day — plenty for monthly WASDE tracking across all crops.

For detailed field documentation, see the USDA WASDE API reference.