Fintwit hive mind: What's the best way to forecast 30-day realized volatility for the S&P 500? Any good review articles of the latest-and-greatest methods? Thx.
-
VIX futs not that useful as they are forwards on 30-day realized variance, so include (i) vol risk premium (forward vol an upward biased estimate of future realized vol) and (ii) variance risk premium (var is convex wrt vol) and both of those are time-varying
i.e. there are two sources of bias and unless you have a good model for the size of those biases they add additional noise to your vol forecast
-
-
Binned data would be fine for intraday realized variance, e.g. 5 or 10 min bins. Much smaller than this and your estimates will be biased upward due to microstructure noise (the std deviation equivalent of the Epps effect)
-
https://www.jstor.org/stable/27638860?seq=1 … Bid/ask bounce if you sample to small. Vol signature plots for different assets help find the cutoff. Advantage is this method converges on a estimate of vol faster than GKYX or close to close.
tells all - Show replies
New conversation -
-
-
Of course, you can remove one of those biases by using ATM IV instead of VIX (which is derived from the var swap) This *can* be useful cos it "knows" about future events, whereas your realized estimate/forecast won't unless you adjust it yourself.
- Show replies
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.