>> @JackRab: Well that sounds very far-fetched, since you're unwilling to back up your bluster... Well, I didn't want to post the link to the article in the open since I don't want my name easily linked to the drunken ramblings that @zdreg takes great pleasure in re-quoting for me... >> JackRab: Found it. Why do you think Antonie Kotze is wrong with his paper on delta hedging with futures? ... but with a little detective work it can be found, as you're proving here Actually he's not "wrong" but using an approximation which he doesn't give in his article but I've deduced it myself based on options math. When trading options there are two things that matter: the price and the delta, if you do delta-hedging. Same price can have different deltas and his approximation is one of them. Mine is not an approximation but the correct (complete) formula for the case of using futures to hedge stock options. I've got a draft paper that better explains how it works, with visual simulations and less formulas, which I intended for a trader / finance audience. But I had to strip it down at the request of my math supervisors (which are not very much into the business side of the things and, you may guess, love formulas ). I can send it to you if you've got a mail address (by PM, please no open doxxing).
Just do it. And post back here on ET after you've Accomplished The impossible and Divine~ Good luck and Mazal tov and May the Farce be With you, (but of course, complete failure is always a possibility as well...and that deserves a post too to documenting it)
>> I reckon OP has some ideas about dynamic hedging and working out a way to capture real vs implied vols... am I correct @Aquarians? Yes, dynamic hedging is essential. >> The main purpose of market making is capturing bid/ask-spread through high volume, for which you need high flow information... especially in index. On top of that... in a low vol situation, where bid/ask-spreads are very tight and often volumes are not that great... you will need to think out vega/gamma/theta strategies and run positions. That's what I build my backtests system for, so I can research complex strategies and figure out how to limit my risk while still making a profit. >> That's not really market making, but it's necessary to turn a profit... you need some experience for that Well nobody ever taught me jack squat, au contraire, did everything in active and passive aggressive ways to fault me, put me in my place, ignore, isolate, threaten and eventually fire me. And applying to other places as @ajacobson is kindly advising me, from a bottom feeder perspective as one does through recruiters or cold calling, well I've said it on another thread that I've better chances doing a reverse Russian roulette (a single bullet out instead of a single one in) than passing through that. And if by pure dumb luck I get by, I'll find myself in the same crap a grunt is at these corporations, a position of no authority but full responsibility, while my superiors are free to take the credit and pass the blame. All in the meantime teaching me absolutely nothing and me being in the same situation, if there's any experience I'm getting it's all by myself anyways. At my age and experience, my only chance is to eat what I hunt and I'm actively training myself into that. >> ... and capital. All I can raise is by working a day job. At least the moment I've quit finance and got instantly hired in a different industry, my salary got 50% up while my work load halved, all with much lower stress and without absurd claims over intellectual property on my financial side project.
Just open a private conversation with me... Trading options, delta is the first hedge... but there's a lot more going on than just straight forward delta... With delta calculation, the problem will be in the further OTM where IV shifts impact your price significantly and a standard delta will be mostly wrong...
Not really an options guy here, but always interested to learn more so I wouldn't mind reading your draft paper either. (Similar to JackRab I also had a quick look at your published paper - if you want 'proof' of my detective skills, its on pages 214-219)
>> @lawrence-lugar: Just do it. Good luck and Mazal tov. Thanks, I'll definitely document either outcome of my endeavour (hopefully the happy one). @h20: sent you a message using "Post" on your profile.
I've done this before ( had a market making system created for me) when I headed an arb desk. Can you tell me where you will be trading? I may have some input
Retail option market making via IB is possible, but difficult. 100 data lines gets you a single name and a single expiry (e.g. SPY171117 has ~250 strikes with bid or at least ask but you can ignore most of the ITM ones and some of the way OTM ones). I suggest SPY to start as it is a fairly efficient market and you won't lose much as you work the bugs out. Fit a parabola to the [single expiry] mid-price vol- curve and put bids in on the OTM strikes most below the curve. Once one of your bids is hit put in a couple of offers in OTM strikes most above the curve. When you need to do a few cancels, hedge off whatever residual delta you've accumulated, IVV - SPY (IVV minus SPY). Your biggest problem will be managing your margin and your cancel fees. Concentrate your efforts on this module -- the cancel/margin management module. The rest of the system can be snapped together from open source (github) components. Since latency is not really an issue here (you will never be able to compete head-to-head with real market makers), you can write the whole thing quickly in Python yourself and be testing by next week. I know at least one person still doing this at IB with moderate (but high variance month-to-month) success using code written circa 2006. Good luck. Let us know how it goes if you actually follow through, though based on your prior posts, I am a bit skeptical that you will.
>> Good luck. Let us know how it goes if you actually follow through Thanks, I'll let you know. >> though based on your prior posts, I am a bit skeptical that you will. Well, today I laid down the tasks (7 points, some of them with up to 5 subpoints). From this point on it's essentially routine. Routine work is still hard but it's 'steel factory worker' hard compared to 'find a cure for cancer' hard. I'm working on my options trading project only after a full day job, which usually means about 12 hours after waking up. I pull 3-4 hours on average, 5 days a week even though I've a strong feeling that most of if it's Sisyphus work. Routine work such as what follows next is a rare treat since I can ignore the interruptions and fatigue, just push through and actually get things done. Worst thing (not now, but in the majority of other cases) is when I have to think. Then, I can blow a full evening without the slightest bit of progress at least in terms of some written code. And worse things worse, if I'm getting close to actually writing some code to start validating my idea, I have to interrupt it for full conscious 12 hours (sleep doesn't count and 12 hours are an optimistic average) before I get a chance to resume. Knowing that it takes 1-2 hours to get back into 'flow' and anyways, with 95% probability I'm fortunate if all I get is not a complete failure but a chance to learn something and pop up another idea (which will most likely fail). And most of all knowing this is my only chance of breaking, after 17 years already, out of the slavery of a job which pays a bowl of soup for 12 hours of my life. Which in a world who largely hasn't gotten over it's primal fear of inanition, is seen as the uttermost achievable goal and my questioning of status quo is at least indecent, when others don't make even that bowl of soup, not to say correctable by making sure I'm not making it either. So yeah, you're right to be a bit skeptical. Just a bit, though.
you maybe in luck, i was a quant for the world's largest option trader. the fund was closed years ago and the principles in the firm went separate ways. i am setting on all the intellectual property that managed globally known clients. the track record is verifiable two decades long. now the catch you really need 100mm cash 10mm to be used as risk capital 90mm to be parked. however this commitment delivers you a never lose iron clad algorithm if we work out a deal. you do not have to pay to develope failure, and you will fail i promise. plus you will lose years of your life and may pass away penniless if you wade too deep. so think on it.