Saturday, July 12, 2014

Calculating an Expected Day's Move Using Implied Volatility

In order to figure out if being long (or short) a weekly day straddle/strangle is in the cards, I have to figure out what the implied volatility is saying.

For example from February of 2014, Facebook announces it wants to buy some stupid messaging company for a stupid $19b, or >10% of FB's market cap. I realize I know nothing about the intricacies of social media businesses but assume that other people do and they'll re-price Facebook's stock accordingly. I see Facebook's back month vol is around 36% and the weekly vol opens at 39%. I pull up a vol chart and see that it's IV typical hovers around 55%. This stock news is more of an earnings event than a boring day in the life of Facebook stock. On news like this, I don't even check twice, I just buy the implied vol. But I digress, few situations are that easy.

What I do next is figure out the expected "day's move". I do this by taking the implied volatility, divide it by 1,590* then multiply that by the stock price. That gives me the expected dollar move for that day.

Example 1, Facebook:
Weekly IV: 39
Stock: $68.00
 (39/1590) x 68 = |$1.67|
(expected percentage move of 2.4%)

So, 68% of the time Facebook will move |$1.67|. My guess is that this is not going to be a 68%-of-the-time kind of day.

Furthermore, FB pre-market/extended hours absolute value range was $4.00, or 5.8%, yet the stock opened almost flat. Here's how I work that percentage change backwards to get the implied volatility indicative of such a move:
5.88 x 15.9 = an implied volatility of 93%

Here's a screenshot of what got me long IV in KORS after they announced earnings back in May, which earned a quick 25% in less than 10 minutes thanks to a spike in IV:

Getting back to the original Facebook example, IV opens at 39% within 10 minutes it's trading at 48% and my strangle position is up 20%. The stock opened at 67.73 hit a low of 65.73 (-2.00, or -3%) then hit a high of 70.11 and closed at 69.63. In other words, weekly IV was way under priced off the open.

Maybe it's just dumb luck, but this simple math has proven to be highly effective and profitable for me.

To see a more concise version without examples, visit the CBOE's writeup courtesy of Russell Rhoads here:

 * Why 1,590? Because volatility is an annualized number. In order to break it down by the day I take the square root of the number of trading days in the year (approximated at 252). The square root of 252 is 15.874. I round that to 15.9 then multiply by 100. Multiplying by 100 saves me from having to divide the stock price by 100 to figure out the implied move.

No comments:

Post a Comment