I also was stopped from using the Oshaughnessy sim/port with the new DB changeover and coincident changes in line items.
However, I found the solution, generously posted by P123’s fantastic staff asset, Marc Gerstein. Marc ‘wrote the book’ on stock selection in his excellent reads, “Screening the Market” and “The Value Connection” - both available on Amazon# He also has some very fascinating and valuable insights in his blog posts in P123’s BLOG section.
I didn’t realize that happened. I guess I was sloppy in not reviewing the formula text carefully before hitting ‘post.’ Strange - I’ve never seen that happen before with simple copy-and-paste.
I have edited the post with the proper message link and formulas, courtesy of Marc Gerstien. Credit should also be given to Dan Parquette for creating the original OShaughnessy sim on P123.
even though it is well documented, is also not satisfying.
The formula does not take into account how many shares have been issued or repurchased. Only indirectly via the fluctuating share price, which fluctuates substantially during one year.
It makes a huge difference if 100 share or 1 Million share were issued. The higher the number of shares issued, the bigger the dilution, the less of the company you as a shareholder own. The same for share repurchases.
But it seems as if there is no equivalent of indicated annual dividend yield for share repurchase? The discussion on this thread seems to focus on rules that look back at how many shares have been repurchased.
I’m guessing there is no (Compustat) variable indicating how much the company intends to purchase in the future?
Just a quick note on the whole idea of shareholder yield. I’ve created and tested some things along these lines and have seen, as probably have many of you, that there is something to it.
Note, though, the change in the interest-rate regime. If rates start rising, stay aware of the fact that this factor may stop working, or possibly even turn around and become a negative. The Street’s favorable view of buybacks is related, at least in part, to the fact that debt capital has been so much cheaper and getting cheaper year after year.
There’s no urgent need to get away from shareholder yield right now. The cost of debt is still very low and is trending sideways rather than up.
But if you are using a model that uses this, you might want to set a calendar alert for, say, every six months, to refresh your view on interest rates and if necessary, revisit the model.
But it seems as if there is no equivalent of indicated annual dividend yield for share repurchase? The discussion on this thread seems to focus on rules that look back at how many shares have been repurchased.
I’m guessing there is no (Compustat) variable indicating how much the company intends to purchase in the future?
Thanks.
[/quote]Did you mean how much they have authorized to purchase?
Many times the authorization is a bluff, to make it look like they are repurchasing shares just to satisfy pressure from stockholders. Sometimes it’s for real. How would Compustat even try to go about quantifying that?
I have looked at shareholder yield in the past and it has some mild utility for very large established companies. I also posted in another thread about how SHYield has changed dramatically from mostly dividends to mostly buybacks for the S&P 500. To Marc’s point, we don’t know the effect of interest rates because it has only really happened since 1982, and some companies have issued debt to buy shares because of low rates, and that will end in the near future.
Maybe we need to develop a SHYield payout ratio and look at that. It will focus on those companies buying back shares with actual cash flow. If I have some time over the weekend, then maybe I will do a little testing.
The formula for the prebuilt ShareholderYield factor could not be easily reproduced in a custom formula. It does not use DbtLTReduced or DbtLTIssued.
The complication is in the way it gets EqPurch and EqIssued. If the stock is a prelim (CompleteStmt=0), and either EqPurch and EqIssued is NA, then it uses the PTM values. If they are not NA, then it uses TTM values. Aside from that, the formula used looks like this: (eval(IAD>0,IAD/Price,0) + ISNA((EqPurchXXX-EqIssuedXXX)/MktCap, 1-Shares(0,Qtr,KEEPNA)/SharesPYQ))*100)
The ‘XXX’ means that it uses either TTM or PTM depending the prelim rules I mentioned above. Also notice that if EqPurch or EqIssued was NA, then it uses the change in shares as a substitute.