SHSSM · IIT Bhubaneswar

Econometrics you can
run, see, and trust — live

Ask hard questions about financial markets and watch a real research engine answer them on live data, in seconds, in your browser. Built to a publication standard for researchers — and legible to anyone in thirty seconds.

▶ Try it now Read the research About the lab

Is the stock market predictable?

One click. The engine tests it on today's S&P 500 — live — and explains the answer.
No economics background needed — press the button.
Don't trust the numbers — regenerate them
We published a financial-contagion result. Watch the engine re-run it live on real data and match the published figure, scale by scale.
▶ Reproduce the paper
From first principles to published research
One methodological substrate flows up through five frameworks into a live 17-method engine — and out as peer-reviewed papers and open-source packages. Scroll the pipeline; every layer is inspectable.
01Substrate8 method primitives — the shared toolkit
Long-memory · DFA/GPHMODWT waveletsTransfer entropySurrogate inference Community detectionNetwork formationStructural attributionROC / DeLong validation
02Frameworksfive research programs, one substrate
NAMHMCPFM arXiv:2507.08065Contagion-Channels arXiv:2604.26546 WaveQTECommoditySOCH fully productised →
03Live engine17 methods · run any of them now
Unit rootVAR + IRFDFA HurstGARCHWavelet variance WQTE spilloverSOCH profileVECMGranger netPanel unit root ConnectednessNetworkDCC correlationLive stationarityWavelet coherence Rolling spilloverQuantile VAR
▶ Open the Workbench
04Published researchpapers · packages · every result citable
MCPFM arXiv:2507.08065 ↗ Contagion Channels arXiv ↗ SOCH sochcontagion ↗ r-universe install ↗ Compute engine MIT ↗
Every analysis carries a permalink and a one-click BibTeX citation — a number becomes a citable artifact.
About the lab
Econstellar is the research platform of the financial-econometrics group at the School of Humanities, Social Sciences & Management (SHSSM), IIT Bhubaneswar — led by Dr. Avishek Bhandari with Ipsita Parida. We build research that is open, reproducible, and runnable: every claim traces to a live computation on real data, and every result carries a permalink and a citation.
Real data · real engine · reproducible.
Live analyses run on the sandboxed compute engine (shssm-compute…run.app) over public market data. Workbench (17 methods) · Reproduce the paper · GitHub · SHSSM, IIT Bhubaneswar.