Warren Buffett has reportedly been trying to reduce "breakeven" levels of the puts he has risen, from 70% to 15%, without putting new money into the deals. The price that he needs to pay is to "shorten" the average time horizon of his holdings. Suposedly, he now breaks even if S&P prices rise 15% by 2018 than 70% by 2028.
This is almost certainly the worse bet. That's partly because rational investors do deals based on the law of averages, and partly becasue the longer the time frame, (i.e. the more trials), the more likely that the law of averages will actually apply. Put another way, time (in the form of an extra decade), is a form of leverage.
The shorter you slice the time periods, more likely that the result will deviate from the statistical mean. That is to say, there are more ways for a 15% bet to go wrong over one decade, than a 70% bet to go wrong over two, even if the latter amount is more than twice the former amount.
The whole purpose of trading derivatives is to gain leverage. Time is a form of leverage. By "shortening" the time period of his derivatives, Mr. Buffett has given up this leverage, albeit to reduce his exposure. But if he didn't like the exposure, he never should have taken it on to begin with. (As President Harry&a... said, ("If you can't take the heat, get out the kitchen).
I advise anyone applying for the job of Mr. Buffett's successor to bet against him on this one. If you win, the boss will probably be impressed (see my piece on Rich Dennis). If you lose, you probably weren't going to get the job anyway.
Disclosure: Long BRK/A, BRK/B shares.