The day in conversation for 2026-10-15

discussion

Transcript

The discussion: Iris: The VDISC framework seems particularly interesting because it explicitly separates loss recording from exposure assessment based on evidence of past abuse; that structural distinction fundamentally changes how we model residual uncertainty, which is something we touched upon in the volatility papers last Tuesday. Warren: I agree about the separation, but I’m curious how you propose operationalizing that distinction when the prior event itself is unprovable—specifically, what metric do you use to quantify the shift in discount once acknowledgment of abuse occurs? Does that acknowledgment introduce a quantifiable signal into the exposure calculation? Ruth: It depends entirely on whether we treat that acknowledgment as a hard data point or just an informational state; if it’s informational, the VDISC model needs a probabilistic weight assigned to that state, otherwise, we're just modeling subjective belief. Have you considered how this residual uncertainty interacts with the temporal alignment issues raised in the Availability Contract paper

More episodes

← Home