No. of Recommendations: 0

I am an AT&T shareholder and recieved AWE stock as a spin-off. They automatically sold my fractional share of AWE and gave me the cash in lieu. Since I have bought several lots of AT&T (and AWE now) as well as all the dividends that were reinvested, I am unsure what lot should I use to consider the tax cost basis for the fractional shares that were sold off. Typically, it should be the first lot that I bought, but I heard that since it is a spin-off or because it is a fractional share, I should just cost average all my lots to determine the cost basis. What is the correct way?

No. of Recommendations: 0

You're not going to want to hear this, but the "correct" way to calculate your AWE basis is to consider each lot of T you purchased (including dividend reinvestments) separately. You will have to calculate the number of fractional shares of AWE spun off from each purchase and the corresponding cost basis for each of the lots. You can then total the number of shares and the total basis.

This is the strongest argument I know of against DRP plans -- the bookkeeping is exceptionally burdensome.

In any event, you cannot cost average lots for a single security under any circumstances. Average cost basis is only allowed for mutual funds.

Ira

No. of Recommendations: 0

*What is the correct way?*

I think I have seen, on this board, the suggestion that you simply use a basis of zero on that fractional share, leaving the basis of your remaining whole shares *very* slightly inflated - perhaps by a few dollars. This is certainly the easiest thing to do.

But I don't think there's any justification for averaging costs - perhaps some of the resident experts will say otherwise. I think to be absolutely precise, you have to do the FIFO thing - that is, identify the earliest shares of T and proceed from there. I've been the victim of numerous spinoffs over the years, and that's what I've always done. In fact, I'm a long-time T shareholder, and have suffered through LU and NCR spinoffs and the 3:2 split in 1999, never mind the sale of a fractional share (when I stopped with the DRIP business) and the subsequent sale of part of my position - how I now wish I'd sold it all! In last week's spinoff, I wound up with a cash-in-lieu payment of $4.21 for 0.26 share of AWE. And via my spreadsheet, I worked out that:

0.54 shares of T, purchased on 5/1/97, basis $17.96, became

0.81 shares of T via 3:2 split on 4/15/99, basis still $17.96, and that generated

0.26 shares of AWE, basis $4.01 (since AWE inherits 22.33% of T basis)

And there you have it - FIFO basis is $4.01, so I have a capital gain of twenty cents. Fortunately, it's long term.

No. of Recommendations: 0

*You're not going to want to hear this, but the "correct" way to calculate your AWE basis is to consider each lot of T you purchased (including dividend reinvestments) separately. You will have to calculate the number of fractional shares of AWE spun off from each purchase and the corresponding cost basis for each of the lots. You can then total the number of shares and the total basis.*

Well, ok - but it's *entirely equivalent* to calculate the number of shares of AWE received, and their basis, from the *total* number of shares of T that you have, and their total basis. It's only when you have a sale that you have to start messing around with separate lots. Consider this example, using a fictional company X:

lot 1: buy 100 shares of X, basis $1000

lot 2: buy 200 shares of X, basis $2500

lot 3: buy 250 shares of X, basis $2800

In all, you have 550 shares of X, and your total basis is $6300

Ok, now suppose that X spins off Y as follows: for each share of X that you own, you receive .316 shares of Y. The basis adjustment is 25% to Y, 75% to X.

Doing the lots separately,

lot1 -> 31.60 shares of Y, basis $250

lot2 -> 63.20 shares of Y, basis $625

lot3 -> 79.00 shares of Y, basis $700

Adding these together, you now have 173.80 shares of Y, with a total basis of $1575. And you'll get the same numbers if you pretend that you have a single lot of 550 shares, with a basis of $6300.

The original question was, "what's the basis of that fractional share?" In this case, Y will issue 173 whole shares, and pay you cash-in-lieu for the fractional (0.80) share. Where did that share come from? Using the FIFO rule, it came from that first lot of 31.60 shares of Y. Those shares have a basis of $7.91/share ($250/31.60 shares), so the basis of your 0.80 cash-in-lieu is $6.33 ($7.91 x 0.80). And now here's the messy bit, and the reason that DRIPs are so damn tedious: *forever after, you have to keep track of the fact that you've used up $6.33 of the original Y basis of $1575.* That is, the basis of your 173 whole shares of Y is $1568.67. If you subsequently decide to sell 100 share of Y, here's where they come from: 30.80 shares from lot 1 (remember, you've already sold 0.80 shares of that lot); 63.20 shares of lot 2; and 6.00 shares from lot 3. I'll let you do the basis. Hey, this is why spreadsheets are so useful!

*This is the strongest argument I know of against DRP plans -- the bookkeeping is exceptionally burdensome.*

I agree, and this is why I abandoned DRIPs several years ago - the bookkeeping became just too onerous. I now own whole shares only, and just take the dividends in cash. (Though to be perfectly honest, all those fractional shares are *not* a problem if you sell off your entire position. It's a partial sale - which, after all, is what you get with a cash-in-lieu for a fractional share - that's such a pain in the neck.)

Lorenzo2

No. of Recommendations: 0

Lorenzo,

I'll assume that your calculations are correct, since the logic behind them is. However, your suggestion to use the total starting basis to yield the total post spin-off basis is only useful if the entire position is sold. As the rest of your post illustrates, as soon as you sell less than the whole position (ie, the fractional share), you have to consider the individual lots.

In almost every instance, the fractional share will arise from the first lot purchased (using FIFO), so that calculation isn't too bad. It's the future sales which are problems.

Ira

No. of Recommendations: 0

**Ira:** We seem to be in complete agreement on everything. Well done.

**mcnuprin:** The AWE spinoff will be good practice. As I'm sure you know, there are two more T spinoffs in the works...

Lorenzo2

No. of Recommendations: 0

Ira,

**You will have to calculate the number of fractional shares of AWE spun off from each purchase and the corresponding cost basis for each of the lots. You can then total the number of shares and the total basis.**

Aha! This leads into a subject that I have had many a 'discussion' about. Suppose your buy 10 shares of StockA for $50 a share. Then you buy 5 shares of StockA at $60 a share. You receive a 2 for 1 split, giving you 7.5 shares for your 15 shares owned and the broker sells the .5 share. What is the cost basis of the fractional share sold?

There will be no reward for the correct answer, even if your answer happens to agree with mine.<g>

**This is the strongest argument I know of against DRP plans -- the bookkeeping is exceptionally burdensome.**

Agree completely.

Rip

No. of Recommendations: 1

**Aha! This leads into a subject that I have had many a 'discussion' about. Suppose your buy 10 shares of StockA for $50 a share. Then you buy 5 shares of StockA at $60 a share. You receive a 2 for 1 split, giving you 7.5 shares for your 15 shares owned and the broker sells the .5 share. What is the cost basis of the fractional share sold? **

Rip,

I love trick questions. In a 2:1 stock split, you would receive an additional 15 shares for the 15 shares you originally owned, so there would be no fractional share. <vbg>

To answer the question you wanted to ask: In a 3:2 stock split, you would end up with 15 shares at a basis of $33.33/each and 7.5 shares at a basis of $40/each. Since there is no way (that I know of) to designate lots on the sale of a fractional share, I would assign the fractional share to the first lot of stock purchased -- 0.5 shares at $33.33/share = $16.67 -- leaving you with 14.5 shares of low basis stock and 7.5 shares of high basis stock.

Although it seems intuitive that you should assign the fractional basis to the fractional share in the second lot of stock, it's just a coincidence that the number of fractional shares in lot 2 = the overall number of fractional shares. For instance, if the split had been 4:3, there would be 20 shares total post-split -- but 13 1/3 share at low basis and 6 2/3 share at high basis. Since the total number of new shares is a whole number, there would be no sale of fractional shares.

Do you agree? <g>

Ira

No. of Recommendations: 0

Ira,

Thanks for clarifying the 'question I wanted to ask' <g>.

**Do you agree? <g>**

No. But you probably expected that. In the first case, with the 1/2 share to be sold, there is no question where it came from, and, so, in my eyes anyway, is adequately identified as coming from the second lot. Now I know the IRS has written a lot in their regulations about what 'adequately identified' means and that the case above will not satisify the reg because you are not going to get written confirmation from the broker that that was the half share they sold. But I believe, and a subsequent court case has held, that the regs provide a safe harbor, but do not constitute the only way a stock can be adequately identified. It is an aberation to me to end up with one block of stock of 14.5 shares and another for 7.5.

In your 4-3 split, I would maintain that there is no way the blocks can be 'adequately identified' and you have to use fifo.

This is not just a mere academic question for me. With bivio [as you know a program for investment club accounting] we had to figure an algorithm for assigning cost basis for fractional shares. I consulted 2 CPA firms and a tax attorney for this question that will probably never be challenged by the IRS. We ended up using the algorithm that if the share could be easily identified, as in the first case, to use that, otherwise, fifo.

Personally, I think that in this day and age of street name stocks, and no certificates, the government is on very shakey grounds with the 'adequately identify' regs and one good court case could clear up this mess.

Rip

www.bivio.com

No. of Recommendations: 0

**This is not just a mere academic question for me. With bivio [as you know a program for investment club accounting] we had to figure an algorithm for assigning cost basis for fractional shares. I consulted 2 CPA firms and a tax attorney for this question that will probably never be challenged by the IRS. We ended up using the algorithm that if the share could be easily identified, as in the first case, to use that, otherwise, fifo.**

While *I* will probably continue to use FIFO in all cases like this, I agree that the IRS is unlikely to ever challenge your method. After all, I haven't heard of the IRS challenging the (wrong, but not uncommon) practice of assigning $0 basis to fractional shares and keeping the entire basis with the whole shares.

BTW, I always purchase stocks in even numbers of shares (multiples of 2, not 10 or 100). This way I don't have to worry about x:2 splits creating fractional shares until the second round of splits.

**Personally, I think that in this day and age of street name stocks, and no certificates, the government is on very shakey grounds with the 'adequately identify' regs and one good court case could clear up this mess.**

Agreed.

Ira