There is a good section in one of the Market Wizard books about option expiration and why there tends to be a lot of volatility. I will try to recreate the blurb here…

Imagine you have a very active option market and two speculators are on either side of a very large position. Trader A is short 3,000 call contracts of XYZ with a strike of $30 expiring tomorrow. XYZ is trading at $29.75 and Trader A is starting to worry that he will be in the money and have to pony up some big bucks if the stock shoots to the other side of 30.

Trader A may decide he needs to close his position, but the option market isn’t liquid enough to let him buy back anywhere near the number of contracts he needs to cover. He might be able to hedge by buying some $27.50 calls, but liquidity is still an issue. If he does manage to buy some $27.50 calls, the people selling the calls will likely try to offset their exposure by buying — you guessed it — XYZ shares, causing them to go up.
Trader A might buy underlying XYZ shares himself so that they can be called away on option expiration. But here’s a catch — what happens if XYZ doesn’t end above $30? Now he’s long a lot of shares of a stock that he wasn’t bullish on in the first place! On top of that, we’re talking about almost 300k shares ($9 million worth) to fully cover the position — and his buying would force the price up, probably enough to make the calls in the money!

Enter Trader B, the other side of the large call position. Does he want the calls to be in the money? If they are, he has to have the margin to be able to inherit all 300k shares. If he does get exercized and ends up with shares in his account, he would probably face a margin call on Monday morning and the shares would be sold at market — creating a huge amount of selling pressure.

Trader B (same logic applies to Trader A for that matter) probably didn’t just buy 3k call contracts… this is where it gets interesting. He probably entered into a spread, for example a calendar spread: he’s long this month, but short next month. Most likely, there are a few more layers to their motivations and context.
So both Trader A and Trader B are going to start doing weird things. They might buy or sell options to try and hedge their positions or wind down their trades. They may try to buy or sell the shares to affect the market, or they may release a few rumors to try and get others to stampede into or out of the stock.

The result? Volatility. And a wierd tendency for stocks to get “pulled” towards their nearest option strike price on expiration days. Now that’s what I call a witchy.

I apologize in advance for massacring the original story from Market Wizards.