I wanted to take a look at how Carnival used or abused stock options, and what effect that might have on the company's performance. Interestingly, the form on which option grants are disclosed, a DEF-14A, was not filed for 2003, though there had been such forms filed previously each year going all the way back to 1995, generally in March of the year.In 2002, however, according to its 10-K, Carnival could grant as many as 40 million option shares, though no one person could receive more than 2 million shares.After reviewing Carnival's ESO grants, it does not appear that the company is an overt abuser. In 2001, stock options represented just 0.4% of all the shares outstanding; in 2002, that had jumped to 2.2%, still not an unreasonable amount, though it's reaching the limits. I've seen Fool articles suggesting that "abuse" would begin somewhere north of 3%.What I noted with the 2002 figures is there were two separate grants, one given in January and the other given in October. All Figures in 000's 2001 2000 ---- ---- Shares Outstanding 586,202 599,663Options Granted 13,169 2,857ESO % of Shares 2.2% 0.5%To determine the number of options granted, I went to http://edgarscan.pwcglobal.com/servlets/edgarscan and typed in "Carnival" and chose "Carnival Corp" from the list. The site then gave me a complete list of every form filed by the company, going back to 1995. Using the DEF-14A, I found the stock options table where it told me that CEO Micky Arinson had been granted 240,000 shares, representing 3.65% of the January options, and 120,000 shares representing 1.82% of the October options. To determine the approximate total number of options issued in 2001, I divided the shares by their percentage of the total (240,000 / 0.0365) to arrive at 6.575 million for the January grant and 6.593 million for the October grant (120,000 / 0.0182), for a total of 13.169 million.Determining their percentage of the shares simply required me to divided the option shares by the total shares outstanding. If the diluted and basic shares had not been the same, I would have used the basic number since the diluted figure represents an estimate. I then performed the same math for 2000.According to those same tables, the exercise price for the 2001 shares was $29.8125, and $43.5625 for the 2000 shares. The impact of the unrecorded option expense on Net Income for the two years would have been $392.6 million and $124.5 million respectively. This would have had a significant effect on earnings per share as seen in the table below: All Figures in 000sReported Net Income: $926,200 $965,458Adjusted Net Income: $533,600 $840,958 Reported EPS: $1.73 $1.58Adjusted EPS: $0.91 $1.40Difference: 47% 11%According to Carnival, however, in a section of its 2002 10-K entitled "Stock-Based Compensation," they used the "intrinsic value method" for calculating the costs of options, pursuant to SFAS 123. Using those calculations, would have valued the shares at $12.67 and $13.31, respectively, for the years 2001 and 2000. It would have reduced Net Income only to $904 million in 2001 (for an EPS of $1.54, or 11% dilution) and would not have materially affected the results for 2000.As the basis for my calculations, I referred to a series of "Fool on the Hill" articles from 1999 by Warren Gump and a Matt Richey FOTH article from 2002. It's quite likely I lost something in the translation, so I hope you'll pick it apart and let me know where I can improve it.In Part II, I mimic what Chin did with PAYX to see what would occur.Rich
All Figures in 000's 2001 2000 ---- ---- Shares Outstanding 586,202 599,663Options Granted 13,169 2,857ESO % of Shares 2.2% 0.5%
All Figures in 000sReported Net Income: $926,200 $965,458Adjusted Net Income: $533,600 $840,958 Reported EPS: $1.73 $1.58Adjusted EPS: $0.91 $1.40Difference: 47% 11%
Hi. I was looking at the 2002 10-K and the numbers from there don't match up with your numbers. I've listed your numbers after the ones from the 10-K.(in 000s)Options granted (2001): 6,580 vs 13,169Options granted (2000): 2,911 vs 2,857Wgted exercise price (2001): $26.44 vs $29.8125Wgted exercise price (2000): $35.92 vs $43.5625I'm not sure what's going on here, whether these numbers were just restated or if there is suspicious accounting going on here.Larry
Larry,As I noted, my numbers came from the DEF-14A statement so there may be some discrepancy as the 10K has several amendments filed after it. As to your 2001 options granted figure, I believe the 6,580 number is from the January grant; the additional ones from my number are from the October grant.As to the weighted exercise price, I see what my problem was there: the January grant price was $29.8125 but the October price was only $22.57. I completely overlooked that price and it is what undoubtedly brought down that number to your $26.44 amount. At least for 2001. Now why the 2000 price would be off so much I'm not sure, except that Robert Dickinson had received 80,000 shares (totalling 2.8% of all options given) for a price of $18.90625. His were the only ones lower that were listed, but I wonder if they amounted to so much as to lower the weighted price almost $8 per share.Thanks for looking and pointing out the discrepancies. I hope you'll be able to look over the rest of the post as well and find (or not!) any inaccuracies there too.Rich
To determine the approximate total number of options issued in 2001, I divided the shares by their percentage of the total (240,000 / 0.0365) to arrive at 6.575 million for the January grant and 6.593 million for the October grant (120,000 / 0.0182), for a total of 13.169 million.:Hi RichDoes your figure of 13 million options come from the 10K? Weighted Average Exercise Price Number of Options Per Share Years Ended November 30, 2002 2001 2000 2002 2001 2000Outstanding options- beginning of year $28.95 $26.80 $22.70 12,774,293 8,840,793 6,517,168Options granted $26.54 $26.44 $35.92 33,000 6,580,250 2,910,575Options exercised $14.35 $11.70 $13.43 (404,615) (2,218,075) (244,850)Options canceled $32.80 $35.15 $35.91 (573,720) (428,675) (342,100)Outstanding options- end of year $29.26 $28.95 $26.80 11,828,958 12,774,293 8,840,793Options exercisable- end of year $28.71 $25.96 $15.82 4,775,894 2,972,498 4,042,452It seems like you calculated it from adding grants. Was there a reason not to use a table like the one above?According to those same tables, the exercise price for the 2001 shares was $29.8125, and $43.5625 for the 2000 shares. The impact of the unrecorded option expense on Net Income for the two years would have been $392.6 million and $124.5 million respectively. This would have had a significant effect on earnings per share as seen in the table below:To arrive at adjusted net income with options, how many options did you use? Granted, exercised or total?Thanks, I am always interested in other peoples approaches to valuing these horrible forms of compensation. Nice job!>^..^<
Weighted Average Exercise Price Number of Options Per Share Years Ended November 30, 2002 2001 2000 2002 2001 2000Outstanding options- beginning of year $28.95 $26.80 $22.70 12,774,293 8,840,793 6,517,168Options granted $26.54 $26.44 $35.92 33,000 6,580,250 2,910,575Options exercised $14.35 $11.70 $13.43 (404,615) (2,218,075) (244,850)Options canceled $32.80 $35.15 $35.91 (573,720) (428,675) (342,100)Outstanding options- end of year $29.26 $28.95 $26.80 11,828,958 12,774,293 8,840,793Options exercisable- end of year $28.71 $25.96 $15.82 4,775,894 2,972,498 4,042,452
mskk,I think I'm beginning to believe I got in way over my head on this stock option stuff. I got my figures from the DEF-14A. I was trying to follow along from a Fool article on how ESO's impact net income.Was there a reason for not using a table like the one you provided? Sure! I didn't realize it was there. The table I got my figures from included headings "Number of Securities Underlying Options Granted" and "Percent of Total Options Granted to Employees in Fiscal Year." So it showed CEO Micky Arison receiving 240,000 (representing 3.65% of the total) in the January grant and another 120,000 (representing 1.82% of the total) in the October grant. I calculated the 13 million options from those totals.Please advise me of the defects in my work. I'm interested in learning more about this.Rich
I think I'm beginning to believe I got in way over my head on this stock option stuff.Hi Rich,Welcome to the club:)The main thing for Carnival, as well as a number of the other companies out there is what happened in 2003. "If" Carnival returns to the habits of the past, as opposed to the direction they headed in 2002, then the stock options will be a problem. If not, then exec compensation will be included in the expenses. I haven't found anything on this yet. Chin
Hi again Rich,I took a look at the stock options as compared to the net income for CCL, and it doesn't look like stock options are a problem. Let me offer a simple way to do the options. Take the number of granted options, multiply the price, and divide it by 3. This will give you a close enough number to work with. You will need to average the last 3 years in most cases. Chin
Chin,Thanks for taking a looking. Would that formula be akin to Buffet's way of valuing options?Rich
Thanks for taking a looking. Would that formula be akin to Buffet's way of valuing options?Perzactly,If you don't know, and in most cases I don't, why not take the words of the best as your guidelines? I don't know ESO will not continue to be a problem, but I think the attention to ESO recently will slow it some, at least for those companies that at least feel they owe something to the shareholder:)Chin
Please advise me of the defects in my work. I'm interested in learning more about this.Hi RichI don't think your work has any flaws. I am so sorry if this question implied that. I was just curious if your method gave any greater accuracy as far as declaration of options granted and exercised. The figures are very close.I see so many ways of calculating options expense, that I am always a little confused as to the best approach. Of course the best approach would be to give us the real expense on the financial statement or get rid of them.>^..^<
I see so many ways of calculating options expense, that I am always a little confused as to the best approach. Of course the best approach would be to give us the real expense on the financial statement or get rid of them.Hi Mskk,Like you need another way of looking at ESO. Huh?My thoughts on this is that the company offers the actual expense from exercised options. Take this expense (granted price minus exercise price * # exercised options minus tax benefit), and reduce net income before adjusting further, and you have the net income available to FCF and reinvestments. Like WC, you would have to look at maybe a three year period. If the company is reducing ESO, odds are the execs are taking their compensation as simple cash bonuses, or salary increases. I have my doubts the big boys are going to reduce their income going forward. A little attention to ESO shouldn't hold them back:)Chin
Hi ChinAnother method of ESOs is only going to confuse me, but I guess I couldn't be much worse off then I already am. I have been having an interesting dialogue with Iceberg at the projects board on this very subject and he /she has tried to help me understand an ESO time value model. It appears to be quite interesting. Here is what iceberg sent me:I've pieced together the conclusion that if ESOs were issued at tangible Book Value then there is no economic cost to the company - just diluted EPS in the future. The cash then that goes in at exercise and the current value of the disbursement from the treasury are equivalent in value. But this doesn't happen. ESOs are granted at market price. Sooooo....Step 1, I find the tangible book best I can and subtract that from Buffett's 1/3 rule total cost. This part of the exercise then gets you the dilution - I assume all the antidilutive options are converted and add those additional potential shares to basic shares as the share count as those option owners are just as good as equity owners. But this ignores the economic time cost. The fact that the ESO was issued at market price looms. So, step 2, I take the remaining part of Buffett's 1/3 rule and subtract that per share total from annual owner's earnings as below. This is good enough for the combat finance stuff I do - it would penalize the companies issuing ESOs at tang. book correctly plus the ones issuing at market (the normal practice). So for GMCR FYE2003, Tang. book is 35148000 - 1446000 = 33702000/6,967,756 = 4.84/shr out If they granted ESOs at 4.84/shr in FYE2003 and paid them at hopefully increased tang. book in 10 years then I'd figure there is no economic cost just the dilution which for 2003 would be 999,602, since none of them are antidilutive (I like that last word and try to use it at cocktail parties.) Well, like all companies, they don't do this, they granted them at wtd. avg. $15.20/shr, way more than tangible book. So I'd subtract from FYE2003 owner's earnings Buffett's 1/3 rule set up like this: 0.9*100,550*(15.20-4.84)/3 = $329,871 and 329871/(6,967,756 + 999602) = $0.041/dil shr. It is interesting that this comes so close to the treasury method, isn't it? The 0.90 is just to assume there are 5% cancels and 5% less value for ESOs than straight warrants. Again, this is combat finance, not the detail Damodaran might go to, but I like to look at many company's financials and his methods, used to the last decimal, would severely restrict my ability to go thru 10-K after 10-K for companies passing an initial screen from the Value Line data. Like Charlie Munger, I think Warren Buffett does it all in his head and makes estimates too. He would go deeper into the ones where he puts 40% of his net worth. Probably even beating out Damodaran for depth of detail. I'd then divide my total owner's earnings figure using the diluted share count to get a per share amount to subtract from total intrinsic value for the ESO valuation of cash flows (& then subtract any Senior Claims like debt). Some have argued this is double counting but I have yet to find anyone or any papers prove it to me. (NB: I don't do 5, 10 or whatever year DCFs as I find I cannot predict growth rates and their duration - I just use the perpetuity formula and, at purchase, assume I'll hold the stock until judgement day which hopefully will be far enough away that the market cap. will meander up around IV over the years and I haven't purchased shares of a buggy whip company right before Autos are invented, which we should have been smart enough to do or at least go short horses as a hedge.) I'd add to your thread to find an owner's earnings figure to get the current return on GMCR if the stock was to be held forever, but I have no special insight into the economics of the coffee roasting business and how it might grow. So I haven't spent the time. I'd need to see how they are able to reinvest retained earnings and keep beating their cost of capital but I don't think I can do that w/o way more work. Their incremental abnormal earnings on book value need to be maintained at the current market price, that's for sure. As you point out the current price has a first look return to me that is not very interesting. <<...deduct from an intrinsic value?..>> Yes, I'd expect the ESO program to keep going forever at current 5 or 10 year smoothed averages. The real time value & time cost is determined by setting up a 10 year account FOR EACH ROLLING YEAR and then keeping track of cancels, converts, outstanding in each of the 10 years ( I think I learned that from Damodaran but not sure). No way I can do that, so I just penalize year by year for the whole program. Pretty easy that way. Any thoughts?>^..^<
Hi Mskk,Somewhere in Dr. D's book, he does say you can reduce revenues or operating earnings by the percentage from ESO. We have already been through this, but I personally don't feel you need to count the outstanding ESOP on top of reducing net income, or operating income. As far as expense to the firm, this doesn't really matter. The reduction in cash flows, “owner's earnings,” or “possible dividends” is what we are concerned with as an expense. Where WEB's 1/3 rule of thumb comes in is that it works out as 11% CAGR with tax benefits considered, so using his 1/3 rule, earnings would have to be adjusted after taxes. The 11% is adjusted out further in future cash flows, so just reducing the earnings by the percentage the first year adjust out well on out. For a high growth stock, the cash flows will not be adjusted down but about half as far as they need to be between the 11% and their CAGR, but if you take into consideration as growth slows, the P/E investors will pay for the stock will also depress accordingly, it adjusts out well. I agree, all else equal, easy's for me. I don't think it is possible to come up with a perzact ESO expense, so ballpark is going to be just as close, and possibly even more accurate. I'm a bit on the conservative side when estimating the cash flows, so I automatically build in a margin of safety into the intrinsic value. I doubt this helped, but there you are?
Best Of |
Favorites & Replies |
Start a New Board |
My Fool |
BATS data provided in real-time. NYSE, NASDAQ and NYSEMKT data delayed 15 minutes.
Real-Time prices provided by BATS. Market data provided by Interactive Data.
Company fundamental data provided by Morningstar. Earnings Estimates, Analyst Ra