GARCH INSTITUTE

Independent Quantitative Research Institute

[GARCH INSTITUTE @GARCHInstitute]

https://t.co/Z7kDm4VYWK

[GARCH INSTITUTE @GARCHInstitute]

https://t.co/TxBD0ugZlz

[GARCH INSTITUTE @GARCHInstitute]

Similar forecast accuracy, different market outcomes. Our new video examines belief correlation, order synchronization and an execution-timing placebo in a controlled market simulation.

Watch the full 13-minute video:
https://t.co/UgyiyuUdTm

[GARCH INSTITUTE @GARCHInstitute]

https://t.co/OX9kYm5BgH

[GARCH INSTITUTE @GARCHInstitute]

Forecast quality is not one-dimensional.

Two questions:

Can volatility forecasts be statistically different yet decision-equivalent?
Can cross-model disagreement act as an ex-ante signal of conditional fragility?

Video companions:

Different Volatility Forecasts, Same Portfolio? Decision Equivalence Explained
https://t.co/u3hkEW3Pg7
Can Model Disagreement Warn of Volatility Forecast Failure?
https://t.co/DvwENNwJzq

Papers:
SSRN 7422079
https://t.co/x2uvYHb9gS

SSRN 7423900
https://t.co/wPagbjG1o6

[GARCH INSTITUTE @GARCHInstitute]

https://t.co/Yb8tCiG3CR

[GARCH INSTITUTE @GARCHInstitute]

https://t.co/Lw1k8cLYyX

[GARCH INSTITUTE @GARCHInstitute]

https://t.co/QG0LGTd5sk

[GARCH INSTITUTE @GARCHInstitute]

https://t.co/REyI80FdH1

[GARCH INSTITUTE @GARCHInstitute]

https://t.co/k6RJSKENgq

[GARCH INSTITUTE @GARCHInstitute]

https://t.co/zYe7TLs2fv

[GARCH INSTITUTE @GARCHInstitute]

https://t.co/o0gt1EOdJ2

[GARCH INSTITUTE @GARCHInstitute]

https://t.co/JAlS3JOsZ9

[GARCH INSTITUTE @GARCHInstitute]

https://t.co/1eCkjNk240

[GARCH INSTITUTE @GARCHInstitute]

https://t.co/oEm4UR0mgg

[GARCH INSTITUTE @GARCHInstitute]

https://t.co/6FB60PqoTw

[GARCH INSTITUTE @GARCHInstitute]

https://t.co/8peMHHpvZg

[GARCH INSTITUTE @GARCHInstitute]

https://t.co/IpMI4ypFh6

[GARCH INSTITUTE @GARCHInstitute]

https://t.co/bUZSOtk2li

[GARCH INSTITUTE @GARCHInstitute]

New paper on SSRN.

Volatility model comparison is conditional on the evaluation criterion, the volatility proxy, and the market environment.

In finite samples, a superior model set may be more defensible than a unique winner.

https://t.co/q5DT4YhgCk