Thanks. I'm leaning towards leaving this post here because indeed I don't think many people would be able to pull this off, or even understand specifically what I did. While I have bigger fish to fry (using more scalable price predictive strategies) and I need a kick in the butt to move on. While I arrived at ET not long ago seeking quality discussions, so trying to contribute a little of what I can. I may be wrong here since I try not to use Greeks myself and therefore may not clearly see your point of view. But I started trading (via coding) four years ago and understood from the get-go that the only way to play is to avoid any biases and try to look at things in my own way to find things others may not see. So in terms of Greeks, I took effort to ignore them. Though let me try to separate this into two areas: a) Arbitrage: I mean I probably wouldn't do anything differently if I never heard of Greeks or Black Scholes. All I would need to understand is that things like debit spread must cost money, so it shouldn't be possible to get it at credit. Even amateur traders on Robinhood understand this much. I took it a step further by looking at option chains all days long for months at a time, analyzing only options pricing (bid/asks) between strikes, and coming up with my own conclusions. Also discovering what is "Box spread" without knowing anything about it prior, and figuring out that getting it at credit shouldn't be possible. What I do is not related to boxes, but it's just another example of basic math that allows anyone to figure out that some option combos lock-in value or must always be priced at $0 or positive value. Taking this further, one can also figure out that Calendars (selling a strike with an earlier date and buying the same strike with a later date) must cost money. If you could simply buy such spreads or calendars at credit then you'd have an arbitrage. While from there you can come up with some "system" where you're able to create variety of combos that should cost money or be worth near $0, while if you could get these at credit then it'd be an arbitrage. You may also be able to look at any couple options and figure out what other options you'd need to cap those off and end up with a combo worth near $0 or debit/positive (similar to delta hedging). While I suspect that if I was trained on using Greeks then I would try to calculate stuff differently and worry about Delta or whether I'm properly hedged, for example, even if it has no importance whatsoever. b) Regular trading. I've trained myself to derive everything I need from looking at the pricing on an options chain and things like distance between related strikes, so for example instead of some IV value, I look at distance between strikes where the price of an option doubles, while also comparing that distance between different months. Hard to explain but I'm pretty sure I'm looking at stuff differently than most. In my own code, I do use C# libraries that calculate Greeks and I do code around them to derive Greeks that I may use here and there, but in real life I wouldn't even remember how to define them or how to use Delta or Gamma (for example) because I derive values that are more clear to me, like ROI or "%decay per day" instead of looking at some Theta value that wouldn't tell me anything as a number. In coding I just use Greeks in the background to get higher-level info. And I absolutely love trading ratio spreads, while never looking at Delta, Vega, IV, or whatever. Option prices are based on IV, so I can make the same conclusions as someone looking at IV, but I may see some values that make up that IV, as well as higher-level values that I derive at the end. For example I know that OTM Put strike distance of 200 on SPX (or ~10% of its current price) should be fairly safe for 1:2 ratio spreads when buying spreads like 1x 2000P / -2x 1800P at credit. These are usually difficult to get at credit unless IV is very high, while I look at this in reverse: the price of such ratio spread tells me how high is the IV, so I immediately know how overpriced options are by simply knowing the price of such ratio spread. And I do similar stuff with various equity options. In terms of regular trades, yes. I mean I do what most traders do by looking at my margin use and various IB reports. But in terms of riskless arbitrage, "risk management" doesn't apply at all because there is no risk. The risk may show up with strategies like those used by Long-Term Capital Management where they arbitraged pricing between different instruments and simply assumed that the price difference/spread will stay within specific parameters. While my arbitrage never has any risk because, again, it's similar to buying a debit spread or a box at credit. If you managed to do this then you'd have zero risk because their value cannot change to negative (unless you miscalculate dividends or %interest). The only risk-type issue for me is that IB may reject my orders after I place thousands of orders and they (mistakenly) determine that too many orders may go through depleting my buying power. Here are sample IB messages pulled out of my db: "Your order is not accepted because your Equity with Loan Value of [($amount) USD] is insufficient to cover the Initial Margin requirement of [($amount)]." "Your account has a very large number (10000) of outstanding orders. To place more orders please either cancel some of them or wait until some of them execute." (a while back I've occasionally used two accounts to bypass that) BTW, I think that some brokers like TDA don't even let you place more orders than your BP allows. While IB lets you place many more orders while depleting your BP/margin after orders get filled. Though I may be making all this more complicated than it really is, just trying to address your questions in detail.
Some of it is probably unique and a lot of it is probably just reinventing the wheel. On a similar note, there is a group at Virtu that looks at options purely from arb and market-making perspective. Once you start playing microstructure games (especially if you have the latency to make markets in both underlying and the options) the concept of volatility becomes kinda meaningless. Interesting conceptually, but hard to monetize in practice.
Thanks for the detailed explanation. I won’t argue that you’re wrong to forget about the Greeks if it’s working for you. Further, it’s quite impressive that you came up with that list of arbs yourself. 1. With regard to a vs. b I understand the distinction. For a) you certainly don’t need the Greeks and your LTCM example holds as far as I can see. I'm not sure at the time frame you describe there's much value in the greeks at all in terms of finding alpha (not an expert here) and if you have other mental techniques that are effective to manage risk that's just as well. 2. Perhaps of interest: Options Trading by Euan Sinclair has a whole chapter on arbitrages. It seems like you have them down though. Boxes, spreads, butterflies, put call parity, calendars. Maybe of interest at your timeframe would be his chapter on market making too. I think that a clever person who watches the screens could probably master everything in that chapter without reading it, but it's nice to get another person's mental framework around these observations.
Their are a lot of opportunities in what you do. If you are this advanced a trader, I think you would be better off moving away from IB and using software that will allow you to view orders on each exchanges COB (complex order book) and allow you to route spreads to specific exchanges. Another problem with using IB for this strategy is they may determine that an order is not marketable and not send it to an exchange at all, which is a particular problem with this strategy.