An options observations, given the day's conversations (long, saying in a thousand words what edamo said in about two dozen <g>):
I'm not singly out any particular poster here, because I don't even remember who's all been contributing to the discussion. But I'm seeing confusion at best, and misinformation at worst. I thought at first I'd wait for the options thread to get rolling, but it seems timely now.
Yes, option values decay with time, and yes, LEAPs tend to start to decay more quickly (ie, lose their time value component) six to nine months before they expire, basically (almost by definition) at the point they become "normal" options (which aren't any different than LEAPs, but people here seem to want to make the distinction bigger than I see it as...)
But people, it's vitally important to remember that this quick decay only applies to options that are near or out of the money. When an option - LEAP or otherwise, it doesn't matter - is deep-in-the-money, there's little or no time value left in it to decay away. So in my mind there's no point in rolling it for another LEAP that's further out in time, if all you're trying to do is capture more time. All you're doing by the roll is generating a taxable transaction, paying the bid-ask spread again, generating your broker two commissions, and most importantly, paying the option seller a lot of money just to buy a higher time value component, one that's almost guaranteed to decay away. You're not reducing any sort of time risk, because there's hardly any left in the option you're holding anyways..
Look at it this way. Say you have Jan 2001 Q strike LEAPS at a 300 strike, which are selling today at the sale price of about $215. Of that amount, right now $467-300=$167 is intrinsic (actual) value, and only $215-167=$48 (or 22%) is time value. Now let's say that next April 1st (when they have almost nine months of life left), Q has risen only extremely modestly from where it now - say 10% to 515. (I know most of us expect that rise in the next three weeks, not the next three months - I'm calculating conservatively here - if we rise more quickly it makes my point even more emphatically.) If today's same implied volatility values for this series of about 66% are still in force, on that day, these LEAPS will be selling for about $246 - so their instrinsic value is now $215 and their time value would now be down to about $31, or 13%. If Q's price froze solid in place at $515 for the next nine months after that day, at expiration you'd still have an option worth $215. Thus your total, maximum time value risk at that stock price, calculating from next April 1st onward, is only $31 or 13% of your net present equity.
But let's say you really want to do the roll anyways, and sell your Jan 2001 300s next April 1st for $246 and buying Jan 2002 300s for their estimated price of $285. So you have to spend $39 to make the transaction go, and what do you have to show for this? A new option who's instrinsic value is identical, but whose time value is about $70 (25%) instead of your old option's $31 (13%). And what will happen to that time value you've bought if Q stays where it is, or rises even more? That's right, it decays away, both with the passage of time and with a higher stock price. If Q goes to 700 in the next six months after you roll, guess what: both the 2001 and 2002 options will basically be worth barely more than their instrinsic values, or very near $400 each. The $39 new time value you bought is wasted, because now both series are so deep in the money they have little or no time value left. Ditto for the passage of time (though if Q stays where it is the Jan 2002s will of course be worth a bit more than the 2001s by the end of the year - but their instrinsic value won't have changed.)
Given these numbers, I guess I'm just not seeing why it would pay to sell the options you already have and buy ones further out just to capture more time. Your time risk is minimal anyways; more importantly, you've already lost quite a bit of the time value you probably paid for (as time has passed and the stock has risen), so why pay the option seller for it all over again? "Sell time value, buy instrinsic value" has always been the wise option player's mantra.
If you really are (1) using LEAPS as a stock substitute, (2) want to keep doing so, (3) are holding DIM LEAPS, and (4) simply want to preserve your current position, you'd be far better off doing nothing this year, but waiting until January 2001 to sell your 2001 LEAPS and buy the 2002s, because they'll have less time value left to lose at that point. (Or better yet, if you're not in a taxable account, ignore the LEAPS entirely and buy Apr 2001 DIM calls, planning to roll them forward near to their expiration - but that's a topic for another post.) You can also better 'tailor' your purchase then, buying at the 'sweet spot' in the strike price range that you personally like to find yourself at (I tend to like to buy DIM calls when the time value is less than 15% of the total price, for example, but that's just me.)
Now, if you wanted to roll out and up, say selling your Jan 2001 300s next April to buy twice as many Jan 2002 600s with the received funds, that's a slightly different story. There you're trading instrinsic value for much higher speculative leverage, a reasonable transaction (though not one I'd necessarily choose to make). However, if I just had to do that sort of purchase, I'd much more likely not sell my 300s - instead (using the example above, Q @ $515 next April 1st), I'd convert them to a fixed-price spread by selling another higher Jan 2001 call against them (most conservatively, the Jan 2001 310 calls, locking in a fixed $10 price difference if Q closes at or over 310 on January 15th, 2001, and giving me about $240 per sold option to go do whatever else I wanted with - more calls, more LEAPS, common shares, a Ferrari, whatever). And (crucially, in a taxable account) this action still preserves the possibility of getting LTCG treatment on my original Jan 2001 300's, a huge win over rolling. (Consult your tax advisor - spread tax rules are insanely complex.)
Too long already, and getting less coherent the more I write. If this isn't clear, my apologies; I'll be glad to explain this more carefully and in better detail when the options thread gets going...
Time to go Christmas-Kyoaroling <g>,
-Rose- |
