Miles,
Thanks - yes, the chance of coming out ahead is about equal to the chance of losing. That's exactly the gamble I want to avoid.
I like your first idea of splitting up the rollover into smaller chunks, and I thought of that too but unfortunately it's not a practical solution for me given the distribution options for my company's 401k plan.
Your second idea sounds like a nice workaround, but it starts getting too complicated for the expected benefit, in my view. And besides any transaction costs, transfer restrictions (like those imposed by Vanguard's frequent trading policy) are another impediment. But I don't think it quite mitigates the risk as much as it appears to at first blush -- if you are unlucky enough to suffer a really big loss on the rollover (multiple percentage points), it's not going to be easy to make it back up with additional counteracting transfers. If this were really a viable strategy, it would be pretty easy to do this all the time with your portfolio and make money doing it. The problem is similar to one of the problems inherent in using a martingale betting strategy of continuing to double your bets until you win -- one really unlucky string of outcomes will ruin it for you.
I don't think you would be expected to make money over long time frames with such an approach. You would be expected to lose a little bit. Of course with hedging you will lose a little bit too in the form of fees.
In the end the transaction costs are the problem with any trading strategy, this one included. More trades equal more fees and more friction.
Happily, most weeks move up or down only marginally. And this fearful event of the market taking off while you were temporarily on the sidelines is probably even less likely than you think.
In fact one final counterintuitive truism to consider is this. I will introduce this paradox with a question.
When was the best performing day of the market this year?
And the answer is that it was October 8, last Wednesday. And this wonderful day of performance was flanked by two of the worst days, and the worst week of the year. So if you missed out on this atypically good market day, the chances are you would've also missed out on a terrible week. In other words you would've come out ahead.
And it turns out that this is quite typical. The best days typically occur in the midst of bear markets,and in close proximity to the worst days.
In addition you would be better off to miss both the 10 best days in the market and the 10 worst days in the market, rather than buying and holding through all of them.
I would interpret this to mean that your chance of something really bad happening (I.e. the market is skyrocketing while you are out of it for a few weeks)is far less than your chance of missing out on something really bad.
But most of the time, of course, nothing of consequence will happen at all over a week or two out of the market.
Here is a nice article on the 10 best day myth.
http://mebfaber.com/2008/03/27/noise-the-10-best-days/