I pull a global equity snapshot most mornings before I read any headline. The reason is simple. Headlines explain the move after it happens. A clean table of closes tells you what already moved, and you can ask why later.
This morning the table was more interesting than usual because the regions did not agree.
what the tape actually showed
I pulled twelve major indices in one call. The split was clear.
Asia led. Hang Seng closed at 25650.87 HKD, up 1.74 percent. Nikkei 225 sat at 65020.94 JPY, up 1.26 percent. KOSPI printed 6687.21 KRW, up 1.64 percent. Singapore Straits Times was up 0.94 percent.
Europe barely moved. DAX was 26046.40 EUR, up just 0.17 percent. CAC 40 slipped to 8278.77 EUR, down 0.09 percent. FTSE 100 was 10831.09 GBP, flat at minus 0.00 percent.
The US side was tame. Russell 2000 was 2975.65 USD, up 0.25 percent. The VIX closed at 14.53 USD, up 1.47 percent from 14.32, still a low reading by any historical yardstick.
So on this one tape, the money that wanted to move was in Asian equities, not in Europe or US small caps.
what the bettors are pricing
Equity closes are one vote. Prediction markets are another, and they are not shy about pricing macro fear.
I pulled regulated prediction market odds (CFTC supervised) the same morning. The contract "Recession in 2027?" traded at a 25.0 percent yes price on 348415 USD of volume. The "Trump economic boom?" contract sat at 22.1 percent yes. "China overtakes USA's economy by 2030?" was at 17.0 percent yes on 128759 USD of volume.
That is the tension I find useful. Equity traders in Asia are leaning in, while event traders are putting a one in four price on a recession two years out and only a 17 percent price on a Chinese overtake this decade.
the honest part
I will not dress this up as a signal. Three limits are worth saying out loud.
First, these index closes are not a synchronized single day print. The snapshot was collected on 2026-09-06 but most exchanges last updated on 2026-09-04. You are reading the latest session each market posted, not one global moment.
Second, a single day of percent moves is noise. One 1.74 percent Hang Seng day tells you about that day. It does not prove a rotation. I have been burned before by reading one green tape as a trend.
Third, prediction market prices are speculative bets, not econometric forecasts. A 25 percent recession yes is a market price set by traders with money down. On thin contracts the bid ask is wide and the conviction is low. Treat it as sentiment, not prophecy.
I run a data hub that aggregates these feeds because I wanted one place to query equity, macro, and prediction market data without hopping between terminals. It is broad, not deep. For any single market you will find a specialist source with more history. I use it for the cross market read, not for final answers.
how to pull it yourself
The index feed is a single JSON endpoint. Here is a real curl:
curl -s "https://agentdatum.com/api/v1/d/global-indices" \
-H "X-Anonymous-Id: your-anon-id"
And a small parse to rank movers:
import json, urllib.request
req = urllib.request.Request(
"https://agentdatum.com/api/v1/d/global-indices",
headers={"X-Anonymous-Id": "your-anon-id"}
)
data = json.load(urllib.request.urlopen(req))
for x in data["raw"]["indices"]:
pc = x.get("previous_close") or 0
p = x.get("price") or 0
if pc:
chg = (p - pc) / pc * 100
print(f"{x['index']:<14} {chg:+.2f}%")
The full endpoint catalog, including the macro and prediction market feeds referenced above, is published as machine readable JSON here: https://agentdatum.com/.well-known/ai-catalog.json
what I take from it
Asia was the only region with real appetite this session. Europe was flat, US small caps were quiet, and volatility stayed low. Meanwhile the betting market keeps a quarter price on a 2027 recession. I am not acting on either side. I am logging both, because when the equity tape and the prediction market disagree, that is usually where the interesting questions hide.
This article was written with AI assistance for drafting and editing.
Top comments (0)