Add training workflow, datasets, and runbook

training_data/curated/text/64e770baa9edaf891085d3d88d8d879c8c457eec8d8cadacacb7fedf51fbff61.txt

@@ -0,0 +1,25 @@
+IV is a common cause of time-spread failure for market makers. When i
+in the front month rises, the volatility of the back-months sometimes does
+as well. When this happens, it’s often because market makers who sold
+front-month options to retail or institutional buyers buy the back-month
+options to hedge their short-gamma risk. If the market maker buys enough
+back-month options, he or she will accumulate positive vega. But when the
+market sells the front-month volatility back to the market makers, the back
+months drop, too, because market makers no longer need the back months
+for a hedge.
+Traders should study historical implied volatility to avoid this pitfall. As
+is always the case with long vega strategies, there is a risk of a decline in
+IV. Buying long-term options with implied volatility in the lower third of
+the 12-month IV range helps improve the chances of success, since the
+volatility being bought is historically cheap.
+This can be tricky, however. If a trader looks back on a chart of IV for an
+option class and sees that over the past six months it has ranged between 20
+and 30 but nine months ago it spiked up to, say, 55, there must be a reason.
+This solitary spike could be just an anomaly. To eliminate the noise from
+volatility charts, it helps to filter the data. News stories from that time
+period and historical stock charts will usually tell the story of why volatility
+spiked. Often, it is a one-time event that led to the spike. Is it reasonable to
+include this unique situation when trying to get a feel for the typical range
+of implied volatility? Usually not. This is a judgment call that needs to be
+made on a case-by-case basis. The ultimate objective of this exercise is to
+determine: “Is volatility cheap or expensive?”