IB – CL options delta discrepancies

R

risknav

I’m wondering if any other CL (Light Sweet Crude Oil) option traders at IB have noticed rather large discrepancies in the delta being reported in TWS, compared to a third-party, such as QuikStrike.

I have to believe QuikStrike is correct, after all I’m using a version provided free by the CME.

For example, I’m seeing the Feb 56 Call with a delta of 0.13 in TWS, but as 0.20 in QS. Another, the Feb 40 Put with a delta of -0.12 in TWS, and -0.15 in QS. The underlying was within mere ticks when both deltas were calculated, so the underlying did not move to produce vastly different values, if that’s what you were thinking.

The problem is once you start trading multiple contracts of each (or any if you review the chain, they are all off somewhat) your total position delta can become “exceedingly off”.

Furthermore to this, many variables related to risk management are based off this total position delta, and other times off single option position deltas. So if IB is seeing everything wrong, that would mean in theory it’s gameable, or they are seeing all the correct values but TWS is spewing out garbage numbers.

Any thoughts of this guys?
 
I haven't noticed a new problem in CL but NG was *way* off most of last week (maybe longer) - both in the Options Chains and in Risk Navigator. I think they fixed NG Friday or over the weekend.

During the trading session a discrepancy between IB and QuikStrike, similar to what you describe above, is not unusual. IB seems to use a midpoint and your examples are well OTM where the Bid/Ask can be wider. Also, QuikStrike is using the previous session's vol curve rather than a mid of B/A so that can affect things as well (that's the way I remember it being explained to me). Bottom line: With OTM options, and especially when IV is jumping around, you're not going to get a super precise figure from either firm.
 
Brighton said:
I haven't noticed a new problem in CL but NG was *way* off most of last week (maybe longer) - both in the Options Chains and in Risk Navigator. I think they fixed NG Friday or over the weekend.

During the trading session a discrepancy between IB and QuikStrike, similar to what you describe above, is not unusual. IB seems to use a midpoint and your examples are well OTM where the Bid/Ask can be wider. Also, QuikStrike is using the previous session's vol curve rather than a mid of B/A so that can affect things as well (that's the way I remember it being explained to me). Bottom line: With OTM options, and especially when IV is jumping around, you're not going to get a super precise figure from either firm.
More...

Thanks, all good feedback.

Now with regards to QuikStrike using the previous day’s volatility curve, that’s applied to present (settled) day prices?

If so, then IB TWS “should” be the more accurate delta for something that definitely has intraday volatility changes, such as CL and NG, but as you mention this also assumes the platform isn’t having systemic issues.
 
To follow up on my own post, during today’s session the same as witnessed in my first post was occurring. When trading reopened at 6PM EST tonight everything “looked” more realistic based on what I believed it should be. No surprise to me, it now matches “exactly” what is being reported in QuikStrike.

So there you go, issues last week with NG (as reported) and this week (and longer back) for CL with regards to option delta.

I hope IB has finally taken most of the bugs out of the broker-side option calculations. The real pain to this is it was all working perfectly fine before, but development resources seem focused on things like the “Portfolio Builder” and “Mutual Fund Replicator” to name a few.

Obviously those are very important to active, sophisticated traders that IB says it prefers. As for accurate option IV, delta, etc. – nope, the above category of traders never need to use this.

“Who needs delta!?”

o_O
 
Sometimes the variation in option Greeks is a result of different volatility inputs when the market has a pronounced skew in the option prices. Good options software will use the actual measured skew in the options price series. Not-so-good options software will simply apply a model to calculate a volatility estimate based on a theoretical options pricing model.
 
Trader13 said:
Sometimes the variation in option Greeks is a result of different volatility inputs when the market has a pronounced skew in the option prices. Good options software will use the actual measured skew in the options price series. Not-so-good options software will simply apply a model to calculate a volatility estimate based on a theoretical options pricing model.
More...

Yes, and I understand what you’re saying. The problem, and I don’t think I’ve actually addressed this yet, is that it doesn’t matter what the “real or more accurate” delta is when it comes to margin (or risk, etc.) when your broker has it wrong.

The best and easiest to explain example is; let’s assume we have an “accurate” delta neutral position. This will, under normal circumstances be margined conservatively under SPAN (and PM). However, in the vast majority of situations once it loses that neutrality the margin will increase, based on the increased risk associated with one side of the position.

I’ve had larger than normal margin changes recently, it’s not related to exchange margin changes (I’m on the CME performance bond mailing list) so I have to assume this is/was the reason.

My margin, as of right now, is the lowest it’s been in the past 7-10 days. This also happens to align with correct delta outputs in TWS.

I guess this is just my beef of the week with IB.
 
You must log in or register to reply here.
Back
Top