dfwjf92jfd

Looks potentially overfit to me... It would be interesting to backtest this strategy for 2006-2012...I'd be happy to do it myself if I have the time later today or tomorrow... I run my own backtests so am not using any third-party tool, so I can model any trade...I'd be interested to see the results of this if I get a chance...

A word of caution here - I'm skeptical any of these backtest results are relevant to the current market. I'd be extremely careful about post-earnings condors right now since the potential for loss is so high. I've stopped doing any post-earnings condors several weeks ago after they all started showing losses. Each one had a really good looking backtest.

I agree with the above. I cancelled my subscription as well for the same reasons. I write my own code for backtesting so don't really need a nice UI. Backtesting is seductive since you can engineer strategies that look insanely profitable. But in reality you're just overfitting to the data, *especially* with directional trades - it is a fiendishly difficult problem since you *want* to believe what the backtesting is telling you. To backtest correctly, you need to correct for overfitting, which means testing your strategies over out-of-sample data, and select data from a wide variety of market conditions. It's always a rude awakening when you see your insanely profitable strategy collapse upon meeting out-of-sample data (including in your own trading account). I've definitely experienced this first-hand. I think there's definitely a lot of validity to momentum trading, but it only works until it the day it stops. I think the trick is recognizing when that day has arrived, and stopping these trades. For me, that pretty much sums up January, since it was such an unusual month given the recent past, at least in terms of the short-volatility trades like the SVXY and VXX strategies. I, like @Yowster, also did a number of post-earnings condors, which worked very well until the very recent past, so I've drastically reduced my position size with these trades. Again, they all look fabulous in a backtest... In times like these I think it's best to be long volatility and market-neutral until some of the dust settles. I've also been increasing my commodity trades, since these are not particularly correlated to US equities. The beauty of options is that you can profit from any market condition in any market that has sufficient liquidity so it gives you that sort of flexibility.

Thanks for the note - we can move elsewhere if it would be more appropriate to discuss on the SO members' forum...

Here's a thought for the new year... So imagine you've been doing VXX or SVXY diagonals, post-earnings condors, pre-earnings calendars, a short-vol SPX position, and some pre-earnings straddles. You take a look at your portfolio-wide volatility exposure and realize you've become net short...by a lot! I'm curious as to the methods (if any) people use to hedge their short volatility exposure in such a situation. What I'm hoping for is insurance (more than a normal profit-seeking trade) that you could put on quickly and easily adjust to protect against the sort of rapid, market-wide shock we've all been reading about. Of course, we'd love to set off all our short vol exposure using hedged straddles and the like, but what if this still doesn't restore a good balance... The "naive" approach would be just to buy OTM calls on VIX. That has the obvious disadvantage of theta decay but the big benefit of simplicity, in addition to being something you can add/subtract/roll at any time. It can also be quite expensive... If you go this route, is there an optimal combination of strike price and expiration people use to achieve a target level of insurance? If you anticipate closing most of your positions within 30 days, should you do a 30+ day option and roll? Or should you buy a longer-term option and simply add/subtract as needed over time. Should we do something crazy like buy the June $22 calls? Here's another thought: maybe the biggest concern is really rapid systemic downside risk, not a jump in volatility per se. Maybe what we're really concerned about is that a market-wide sell-off would happen much faster than further appreciation, making it more difficult to adjust in time... So what about buying OTM puts on SPX or RUT, while leaving the upside risk unhedged? Maybe there's another index that would be more appropriate for this strategy? Finally, I'm curious whether people have a more efficient way of achieving the hedge than these approaches...

Thanks all for sharing your thoughts. I'm most interested in selling options on commodities, as I believe it gives you more room to be wrong than straight futures - we all know the dismal record of retail commodity investors! It also provides diversification away from US equities, which is very appealing. I was particularly influenced by the book The Complete Guide To Option Selling. The authors write at length about the virtues of selling options on commodities as opposed to equities: Diversification Lower margin requirements (obviously, this is to be treated with extreme care) The ability to sell much farther OTM options for an equivalent amount of premium compared with equities I found their case to be pretty compelling. I've done a number of these trades and have read a ton of materials but am doing it in a vacuum - the ability to discuss and exchange ideas with talented traders (who actually trade, rather than just talk) is just so important and something that can't really be replaced through research.

Hi all - I'm looking for the "SteadyOptions of Commodities" - a site for discussing commodity futures and options trades (particularly grains, metals, and energy). Does anyone have a recommendation for a forum with quality discussion (and that takes risk management seriously)?

Thanks - that's a great lead on the futures data. As for options data, I *did* find a good source after a bit of searching: http://www.ivolatility.com/. I can get a year of options data on a single commodity for approximately $30, which is pretty reasonable. I'm thinking this, combined with the Quandl dataset you suggest, should have me covered...

I know we focus here on equity options, but maybe someone here can help... I'm looking for a high-quality historical dataset of options for certain commodities (soybeans, corn, etc.). I've used and really like historicaloptiondata.com for equity options: their prices are generally $130ish for a year of data, but they only have equities. I contacted the Commodity Research Bureau (crbtrader.com) - they purport to have this data at a very reasonable price (~$50 per year per commodity). Sadly, I was informed that "We don't sell the options data anymore because our database was corrupt" and that their new database will not be available for a few months. CRB has reasonably priced historical futures data for a host of commodities...just not options. I then turned to CME Datamine (http://www.cmegroup.com/market-data/datamine-historical-data.html), but they charge on order of $500 per year per commodity. Surely there must exist another reasonably-priced source for this data, right??

Hi Ophir - Thanks for the write-up! I was curious to test this out myself, and I spent some time trying to reproduce your results. I had a lot of trouble doing this until I realized that the settings you used assumed you could buy and sell at the midpoint of the bid-ask spread. I had my settings initially set to buy and sell at market...this made a rather dramatic difference in the outcome: Parameters Long AVGO 50-50 straddle 2 year backtest 30 day options Buy 6 days before earnings, sell 1 day before Fill Type: Mid-Market Total Return: 45.6% Fill Type: Market Total Return: -33% I was shocked at how sensitive the result was to this one parameter. Do you think it's a bit unrealistic to expect you can get your orders filled at the average bid-ask? Sometimes I'm able to do this, sometimes not. But because it is so critical, I'd suggest listing it as one of the parameters of the trade in the write-up. Moreover, I think this would be a great topic to do a post about since people who are used to trading highly liquid securities with tight bid-ask spreads might be totally clueless about how this is different for options... Thanks again and keep up the interesting ideas!