Ron, two observations, it doesn´t seem to be as much of an issue in RUT and in SPX if I keep the combined call/put skew box checked it largely negates the major effect of altering strike preferences and gives a more realistic value. Also I´m seeing a large effect on iron flys, the M3s are ok. So it seems to be a call issue. I don´t mind minor variations in position delta but huge discrepancies like I´m seeing, e-g., 150 vs 50, are impossible to work with ($300k notional vs $100k )
Ron, although EIOIO is weaker than the variable model for predicting changes in greeks in response to changes in price & volatility, it does provide the gold standard of the current position delta doesn´t it? And if that´s correct keeping our trades within delta limits is best performed in EIOIO mode whereas variable mode can be used for predicting where it will go (i.e, greek trends)? Not sure how all this is compatible with observations on how TOS (using EIOIO) was providing inaccurate delta values as compared to OV but perhaps greek trends (variable model) are more important in analysing high gamma positions and or positions value in response to big moves, but much less so otherwise. Have a position that´s about $2k south that shouldn´t be based on modelling, but should be based on reality
Yes, your P&L comes out of EIOIO. Your modelling (predicting changes in P&L based on price and IV changes) is best done with variable variance (CEV). In OV, notice that there will almost always be a difference between the black dot (based on EIOIO) and the T+0 (which you should have set to variable variance). This implies that you can't just use the price shown via the magic wand in OV in order to figure out your P&L based on a move up or down in price. You need to take the difference between the black dot and the T+0 and then project that same difference from the point you are now wanting to look at. For example, the black dot may be at a level of $500 profit. The T+0 shows (via the magic wand) a profit of $1200. This implies that the theoretical greeks are overestimating profit by +$700. You then look at your T+0 for a potential move which is 50 points higher that where you currently are. The magic wand tells you that you are now at a profit of $1000. Therefore, since you knew that the theoreticals were overestimating your profits by +$700, you need to subtract this from $1000, giving you a "real" projected profit of $1000 - $700 = $300. If using TOS, and using the volatility smile approximation for your greeks, then you would need to do a similar process as mentioned above.
Great explanation Ron, thank you! For the record I´ve narrowed the whole issue down to the call strike preferences, not EIOIO vs variable or combined call/put skew. Set the calls on "moderate" or "large" and my position is delta minus 130 to 150. On "small" it´s minus 50. The minus 50 is close to correct (cross checked in TOS and with synthetic equivalent). So something about the call values are messing the modelling up. Wrote to OV, will see what they say.
As a newbie, trying to follow this whole discussion. Current summary: --Follow mentor of various trades (Locke, Larson) in checking or unchecking Combined Put/Call Skews --Set AutoStrike ranges for Calls to Small for both ITM and OTM --Use EIOIO to best determine *current* delta number. Use Variable (1-day) as best model of greeks in response to future price movements --Add or subtract difference between Graham dot and T+0 line to OV's projections for P&L of future price movements --Use TOS "Individual" and "Volatility Smile" selections in same way as OV's EIOIO and Variable selections Any corrections would be appreciated. Jay
Hi Jay 1) use the same OV settings as whichever mentor you follow - agreed 2) setting autostrike ranges I would suggest also using the same settings as whatever mentor you are following John Locke uses "Large" for all 3) using EIOIO for current delta number well, that's what DavidF seems to prefer, but I disagree if you're using OptionVue, then I would strongly suggest that you set it to use "variable" variance and not EIOIO (the black dot automatically uses EIOIO so there's no need to change anything) 4) yep, agreed 5) yep
Regarding the strike preferences, OV have said they recognise there´s a problem with the calls in ´moderate´or ´large´ and are working on a solution. They suggested to temporarily stick to ´small´until it´s fixed. Since this was the whole issue with the crazy deltas I was getting, the advantage of EIOIO will disappear when this is resolved (Ron knows this stuff 100x better than I do).
I´ve been recording where EIOIO vs. variable would predict my P/L would be and where it actually ends up and have found that EIOIO comes out much closer in almost all my trades. For example yesterday with the down move OV had both my rhino and M3 as up 2 to 3k respectively. EIOIO predicted they´d be flat/small loss. The latter was correct, both trades flat/loss. Same for a Weirdor. Admittedly a very small sample but frustrating. On variable if my call strikes are set to ´small´my SPX is a neg. 60, on moderate it´s neg 10. Not good. Understand Len Yates´s explained this but currently uncertain what my notional risk is. I know others have found the exact opposite so no idea what´s going on
I understand there is a difference between theoretical value and actual value and I definitely want to see that information, but doesn't make a lot more sense to have the main curve of your T+0 line show the ACTUAL value of what you will be able to sell your position at, then maybe have a dot above or below that line to indicate theoretical value? (Or maybe even two separate lines) Isn't knowing what your real P/L will be 30 points up from now a lot more valuable than knowing what the theoretical P/L will be? I trade in actual dollars, not theoretical dollars. If my stop gets hit on my real position value, then I need to get out, regardless of what the theoretical value is.
Isn't knowing what your real P/L will be 30 points up from now a lot more valuable than knowing what the theoretical P/L will be? The issue is with option prices which are all based on theoretical models, whats the real impact on your position if the underlying moves by 30pts, whatever P/L you come up with will always be theoretical, if the model can tell you exactly what the IV will be with a price move, then you will have a real P/L, unfortuntely no model exist that can do that.