This is from my past observation. Their expected p/l or greek values anywhere near the values predicted by the model. From my observation EIEIO is much more accurate when there is sudden downward move.
So in messing with the models, you're right, the EIOIO model does paint a more grim picture to the downside than the NewVar model. There isn't much about the debate online, other than some other message boards talking about the difference. The consensus I gathered was EIOIO seems to be better for P&L on the actual day, as opposed to NewVar using a blend of of the position's IV and projecting it. I'd prefer to use the one that will give me a worst case rather than a better one however, so I'll have to tinker with it some more to see how it impacts the strategy/trades. Thanks for the heads up.
Keep in mind that all models suffer from imprecision. It's simply not possible to predict what vol will do in a large dislocation, which are the moves that are the most dangerous. It's better to spend your time coming to some level of comfort about what you'll do if the dislocation occurs. And then just let things play out. If you're at that point already, then go out to lunch, spend time doing something else enjoyable, and let the market do whatever. Position adjustments in response to spot are better approached from a daily review than a minute-to-minute scalper's mentality.
Could not agree more sonoma! I only check the position once at 2:00 CT, and even then, I give the position a pretty wide berth as long as major moves in either direction don't kick my butt. I get that no models are perfect, they are just that, models. I would just like to use the model that provides me the best picture of the worst case. I'd rather my P&L have surprise upside than downside...
Snoozer day again... Greeks look good: Analyze graph looks good: The downside gamma trend is a bit concerning if we do have a 2-3% drop over the next few days, but nothing that can't be handled. NewVar model shows no losses until a 3% drop in price, however... EIOIO model shows no losses until a 1% drop in price, obviously a lot less cushion. Hoping we'll see a move down in the near future to see which one appears more accurate.
Yep lookin' good so far. Hope you can keep it going and sneak out before the devil (market) knows you're there!
RUT is down 2% today and you were right. I'm in the red by about $600, which according to yesterday's NewVar model shouldn't be. EIOIO on the other hand had me in the red after a 1% drop, which is definitely more accurate. Thanks again for the insight, it saved me a few hours of work figuring it out! Now just to figure out which model is more accurate on a strong UPSIDE move.
Not such a snoozer today. RUT futures down 2%. My delta is positive 33, which is high but not critical and normally wouldn't merit adjustment just yet, however my t+0 line is showing significant downside risk if the selloff continues with the same intensity. When evaluating a roll down of the butterfly to a 1510 mid strike, my call delta actually hedges quite well as-is, and I gain a lot of protection to the downside, while still keeping decent upside protection in the short term in case of a bounce. My gut feeling is RUT wants to close the gap made back on 2/13, so I anticipate more downside, and this adjustment makes me feel better. If I'm wrong, my call should protect me in the short term. Lost some theta in the adjustment, but gamma and delta are warm and fuzzy. I should mention that as of today, I'll be using the EIOIO model until I can study a bit more about how the different models will react on a strong upmove/volatility crush. The EIOIO model called the downside much better today, so I see no reason to deviate from it until further notice. Greeks prior to adjustment: Greeks after adjustment: Analyze graph prior to adjustment: Analyze graph after adjustment:
Maybe I'm just paranoid but if I was was trading it I can envision this scenario: About a week or a few days before taking profit, things would be looking good, about $4k profit, maybe lowish delta still, but that negative gamma is creeping up... Then bam! -3-5% down day or couple days. Deer in the headlights and a chunky -$5K loss. Cripes, had a nice gainer too... Don't be me!
Yeah that's an all too common risk, which is why barring VERY optimistic market conditions, I will probably bail out of the trade around the 21 DTE mark so the gamma risk doesn't become too excessive. I'd rather miss money than lose money.