Blair Hull pushed max pain due to his belief that the public was net-long atm calls; as expressed in the stock's P/C ratio. Public sells calls as expiration approaches -> locals buy those calls and short stock to get neutral. Hull felt strongly that the net-effect was a convergence to the near-strike, CLOSE TO EXPIRATION, but felt that there wasn't necessarily any relationship to the high OI strike. Low-vol shares will often expire near the the high-OI strike simply because the limited vol made it so. IOW, it's coincident. The belief was pervasive throughout the firm, until a top-trader at HTC produced analysis which showed the majority of atm calls were traded on the bid; suggesting that pensions (and others) were doing buy-writes. The case was also made on the vol-surface (25d risk-reversal spreads). Further, the convexity is such that the locals would lose the most by being pinned to the strike if being forced to buy gammas as expiration approaches. Hull discarded his earlier theory in favor of the buy-write, which was a complete 180. HTC's profits in the earlier strategy were likely due to microstructure edge, or simply being net-long flies across the active strikes. The trader that brought the analysis to Hull left HTC and started a fund. By 2005 his personal wealth exceeded $1B. Yass at SUSQ made a case for pinning due to the unwinding of conversions/reversals close to expiration so that locals can AVOID pin-risk. He also believes that it no longer has any impact.
Calling the numbers I posted analysis would be an insult The only purpose of posting them was to refute the claim by many who post that max pain causes stocks to gravitate to the strike that causes the most pain. If that were the case, at expiration, one would expect to see a clustering of stock closes at strike numbers or mid strike and month after month, that's just not the case. As I alluded to earlier, there was a "scientific" study that concluded that there was an expiration effect of about 15+ cents. Some chalk it up to market manipulation. Others to market forces or just chance. My opinion is that market and sector movement, and individual news are more relevant to direction and certainly more reliable for trading.
How many option positions have you held on expiry day in the last two years ? This establishes some relevancy ( or lack of ) to your opinion.
Any of the proponents should publish pinning candidates to this thread for Jan expiration. Say 10 candidates on the Wednesday close before expiration. My guess is that nobody will take the challenge.
well i do not have time to do the work for others. but we could have a look at these 25 stocks - they should be more inclined to max pain manipulation per above... let's also remove those 5 or 6 above, say, $2.5bn market cap. I will be interested myself in the results... http://seekingalpha.com/article/240...s-with-bullish-options-sentiment?source=yahoo
Hi@all, hope the thread is still active. I dabbled around in the max pain theory for some time now and created initially for myself a calculator (as most of the free versions online were very feature limited). I just wanted to share my creation with you and also ask for feedback. I am thinking of adding now a "unusual activity" too, to easily track identify these symbols (currently there are over 500 symbols in the database which are updated daily; symbols that are not in the database but are user requested will be added in future updates). here a quick summary of the features: - Option pain/max pain calculation for current and future expiration dates - Option pain/max pain Put/Call Chart - Detailed grid for further analysis (fields vary between iPhone and iPad) - Historical chart for expiration month (Calculated option pain price vs. last price updated daily) - Over 500 default stocks in database (DOW, SP500, NASDAQ) - Ability to add new stock symbols by users feel free to check it out under: www.option-calc.com and let me know what you think or have ideas for features or improvements. cheers, Alex
nice job indeed! some suggestions. 1. graph: use less similar colors for puts' pain and calls' pain. 2. calculations: each option is actually for hundred shares of the underlying. therefore all pain values are in fact 100x higher than what you show. after this is fixed show them in '000,000s (of the underlying currency) with precision to 2 decimal places. in general avoid too many decimal places in the table (and show numbers in e.g. '000s, '000,000s etc.) 3. layout: split horizontally (not vertically). graph at the top, table at the bottom. 4. add more tickers. think about other markets, e.g. Canada. start to calculate historical pain for all of them. 5. hide info page by default. 6. when finished and you have a robust tool then start to charge something for it - it could be a nice product. maybe you even manage to get some sponsors/advertisers to allow some reduced functionality for free. obviously more "live" the data and calculations are (i.e. more servers you buy for the task) the more you can charge. 7. and most importantly, don't forget to give me a discount cheers
I am a big seller of the whole max pain concept. Mainly because it is meaningless unless you know what strikes are delta hedged and what strikes are kept naked. Also, it would only matter if the open interest in a given strike is meaningfully larger then ADV. There was a concept of "put bombers" in the hedge fund community a few years ago (before the Max Pain of 2008 ). A fund would identify a lowish liquidity stock with high implied vol and sell a lot of puts to the dealers. Since the dealer would be forced to monetize the vol and trade a large percent of ADV, realized vol would fall significantly. You could only guess what happend to these people in 08...
At the moment I only trade ETFs, futures options and OTC in a very different asset class, so there is very little i can offer to you there. As I said above, the general idea that the stock will expire around the areas of highest open interest is based on the assumption that long option buyers will delta hedge at high fraction of ADV. Looking at historical data I could not find any evidence either that there is a significant decline in volatility nor that there is increase in autocorrelation.