30 September 2006

Long-term or Short-term?

This post is a quick stroll down algebra lane to figure out whether holding on to something for the (more favorable) long-term capital gains rate really makes sense or not. I’ve discussed this sort of problem with a couple people and I’m recording my ideas for reference, posterity and criticism…

This evening I was thinking about the cliff in taxation between short-term capital gains and long-term capital gains. Something is “long-term” if held for a year or more, or if you’re forced to sell it because you’re hired into a position in the Bush administration. (No, I’m not making that part up.)

Short-term gains are taxed at the normal income rate which is 25 or 28% for most software developer/system administrator folks I know, and long-term is taxed at 15% in most cases. (These numbers may change at Congressional whim. And while I’m disclaiming things here: I am not your tax or investment consultant, you should consult one if you want advice you can hold someone liable for. Beware of dog. Slippery when wet.) I’m going to pretend the AMT doesn’t exist for purposes of this exercise. (If your capital came from stock options, the IRS sadly will not let you pretend the AMT away.)

This leads to a problem of deciding if the risk of waiting longer is worth the 10-13% difference in tax rate. The rate only applies to the gain, not the full amount. The amount the captial is worth in-pocket is the price minus whatever capital gains tax applies.
So a useful application of algebra:

Net = Price – (Price – Paid) × Taxrate

The net right now is probably going to be at the higher short-term rate. The market goes up and down a lot, and it’s no good if by the time the long-term rate applies the net is going to be less than it is now. You can’t necessarily time the market and know where it’s going, but some industries have pretty obvious seasonal trends. What price can the capital drop to before it’s a better deal to sell now? We know the net we can get right now because of the equation above. So now we need to find the price for another tax rate, so solve for Price:

Net = Price – (Price – Paid) × 15%

Net = Price – (15% × Price – 15% × Paid)

Net = Price – 15% × Price + 15% × Paid

Net – 15% × Paid = Price – 15% × Price

Net – 15% × Paid = (1 – 15%) × Price

(Net – 15% × Paid) ÷ (1 – 15%) = Price

So if I paid 6.50 for something that’s now worth 20.00 short-term, that’s equivalent to 18.41 long-term. That is if it is likely to drop in value more than about 2.50 by the time the one year “long-term” timer runs out,

In just 6 months time, you’ll need to net 2% more just to keep up with the current rate of inflation.

((102% × Net) – 15% × Paid) ÷ 85% = Price

So now I’m at 18.80 long-term vs. 20 now. If I think there’s an even money chance that it’ll lose 1.20 in 6 months, I’d be better off to sell now and invest in something that has a more favorable trajectory.

“Timing the market” is a foolish adventure, but blindly waiting for the long-term is an unnecessary risk if the value moves in a seasonally predictable way that’s not in your favor.

I’ve setup a Google Spreadsheet with this math in it for the curious.

10 September 2006


In the front yard.