Is correlation really rising?

In a calm year the correlation between weekly returns of a stock and the Nifty is 0.3. In a crisis Nifty's variance rises fivefold ($\delta = 4$). (a) Under the model of Proposition 2.15, what crisis correlation would you measure? (b) In another crisis, with the same $\delta$, you measure 0.7. What calm-period correlation would explain this with no change in the underlying relation? If the actual calm correlation was 0.3, what do you conclude? (c) Why is the Forbes–Rigobon adjustment not a reason to ignore crisis correlations when sizing positions?