Unintended Consequences of a Delta/Northwest Merger

On December 31, 2007 Southwest's (NYSE: LUV) share of the $28.4 billion quarterly revenues generated by nine domestic air carriers was 8.4%. Question: What was LUV's share of market cap on that date?

Even if you have a large position in the stock, you may not know the answer. Why? Market share is a marketing metric. Does market share of group capitalization make sense?

SOME BACKGROUND

The overarching purpose of my book Competing for Customers and Capital is to forge links between finance and marketing. Specifically, to show how the markets for customers and capital interact in generating revenues and creating value. But before I jump into that, take a look at how market share most often is used in the press.

In their February 7, 2008 article, "Northwest and Delta Talk Merger," Andrew Ross Sorkin and Jeff Bailey make just one reference to market share, one which is not share of capitalization:

United, with strong Pacific routes, a big Chicago hub and a large market share on the West Coast, would complement Continental, which has big hubs in Houston and Newark and gets more of its revenue from international flights than any other big domestic carrier.

Similarly, in Jeff Bailey's January 17, 2008 NYT article,"In the Math of Mergers, Airlines Fail," he makes only one reference to market share, and again, it's one which is not share of market cap:

The second theory used to justify airline mergers is that combining would increase revenue because a bigger route system would help take market share from competitors.

In fact, references to market share in The New York Times are unusually frequent compared with other sources, including financial blogs. They may have a gut feeling about the importance of market share. The markets for customers and capital interact in many subtle ways that affect stock price. Which in turn affects revenue. The purpose of this new series on airline mergers is to show the importance of tracking that interaction.

THE BASIC DATA

The following table presents the basic data on market value and revenues. The first column lists the ticker symbols of the nine airlines used in this analysis. The second column lists the closing price of each stock on February 29, 2008. This is followed by the number of shares outstanding (in millions) on December 31, 2007. Next market value in billions, the arithmetic product of price and shares out, appears in the 4th column. Finally revenues (in billions) are listed in the last column. The shares out and revenue data were downloaded from the EDGAR Online I*Metrix financial database.

By the way, notice the group's total market value ($28.84 billion) and revenue ($28.36 billion) are nearly equal. That is the group's value/revenue ratio is about one, which is exactly what I found to be the long-run expectation.

VALUE-REVENUE DIFFERENCES

It's simple to convert the basic data into share of value [%SOV] and share of revenue [%SOR]. That's done in the following table for the same nine airlines. The value-revenue [V-R] differential is calculated in the 4th column. For example, American Airlines' (NYSE: AMR) value-revenue differential is -8.7%. The company generated 20.0% of revenues in the last quarter of 2007, but it created only 11.3% of the group's market value. Is this a bad thing? Yes, its market churn was nearly twice its value creation.

The answer to the question I asked at the top of this post appears in the table above. LUV created 34.3% of the group's market value with only 8.8% of group revenues, or, the Southwest Airline employees share of value was about 4 times the share of revenue. Is this a good thing? Yes. If you're interested in the theory behind this metric and how it behaves over the long-run in the airline industry check out my audio slide show Y'all Buckle That Seat Belt.

HIGH-FLYERS AND BOTTOM-FEEDERS

Does data in the table above shed light on a possible merger between Delta (NYSE: DAL) and Northwest (NYSE: NWA)? Begin with their V-R differentials: DAL has a negative differential of -2.7 points while NWA has a positive differential of +1.3 points. The volatility [Risk] in these differentials is measured by their standard deviation over the 23 quarters from the Q2-02 through Q4-07: 4.3 and 4.1 respectively for DAL and NWA. Dividing the differentials by their risk creates a risk-adjusted differential [RAD].

Since it is a standard normal variable, with mean 0.0 and sigma 1.0, any value of RAD greater than +2.0 identifies a company as a high-flyer. Any company with a RAD greater that +2.0 is performing two standard deviations above the expectation. The chance of observing this extraordinary performance is about 5 in 100. In the last quarter of 2007 LUV was approaching this upper limit with a RAD of +1.7.

Alternatively, any company with a RAD less than -2.0 is a bottom-feeder. Its performance is two standard deviations below the expectation. The change of observing this weak performance also is about 5 in 100. In the last quarter of 2007 AMR was approaching this lower limit with a RAD of -1.5.

DAL+NWA = BOTTOM-FEEDER

The following table summarizes the implications of a DAL-NWA merger using this theory.

The markets have had plenty of time to anticipate effects of the merger and incorporate them into the prices of DAL and NWA. Both are valued at a little over $13 and both have a similar number of shares outstanding. So, a back-of-the-envelope valuation of the combined companies is the arithmetic product of the mean of their individual prices ($13.39) and shares outstanding (280.73 billion). This leads to a merged market cap of $3.759 billion. Taking a neutral point of view on the revenues of the combined companies, I added their Q4-07 revenues.

NOT A PRETTY TAIL

Under these assumptions the merged companies would have a %SOV of 15.0 and a %SOR of 27.4. Since the value share of combined companies historically is less volatile than the separate organizations, the standard deviation in their V-R differentials over the 23 quarters is 2.6. This returns a risk-adjusted differential of -4.9. Almost 5 sigmas into the negative tail of the distribution of risk-adjusted differentials.

Will this analysis prove to be accurate? Probably not. But it gives DAL and NWA management food for thought. Perhaps even more important it also gives them a new way to study the unintended consequences of a merger. What do you think?

Comments

[Get a Clue!:]

Excellent article! However, 99% of "investors" only look at the hype written on MSNBC and would be lost looking at the reality of the airline industry.

Everyone says Southwest will be in the same boat as all the others when their hedges run out.... HOWEVER..... Southwest is still making money and buying more hedges all the time. As long as the rest of the industry follows Southwest's pricing lead, Southwest will continue to make money for their stockholders while buying more hedges to make more money in the future.

2008 Mar 03 09:06 AM

[Victor Cook:]

MyAmerica,

Thanks for your kind words and update on LUV's future. I understand those hedges are getting to be pretty expensive. Do you have any knowledge on their cost/returns at Southwest?

~V

2008 Mar 03 10:46 AM

[Bill @ Sabre:]

This is a very good way to look at the likely consequences of an airline merger. Typically in this industry, the financial press tries to make the case that a merger represents 2+2=5. The truth is that airline mergers more often look like 2+2=3. The after effects of mergers leave a drain on customer care, labor relations and ultimately the bottom line for many years. [See Republic-NWA]

2008 Mar 03 12:01 PM

[Airplane99:]

One question on the methodology, why do you assume the market value of the combined company would only equal the average of the market values of DAL and NWA as separate companies. Similar to the assumption that revenue would be equal to the sum of each company's revenue, wouldn't it be a reasonable assumption that market value would also be equal to the sum? This would yield a V-R of -1.3 and a RAD of -0.5. While not great, very different than what is presented above. Additonally, if the presentation above is correct, wouldn't this also result in a DL-NW merger improving the V-R's and RAD's of every other airline included in the survey, since the combined market value would be lower across all the airlines, based on a combined DL-NW only being value at roughly 1/2 of what the separate entities were valued? Please educate me if I am incorrect in my assumptions.

2008 Mar 03 12:59 PM

[Iggnignot:]

wouldn't resulting profit ratio from each companies revenue, whatever it may be, change, as the combined companies operate differently after they merge, therefore making those averages merely guesses?

2008 Mar 03 04:39 PM

[Victor Cook:]

Airplane99,

The methodology becomes clear when you think about how market cap is created: the closing price (say $13.35) on a given day (say 2.29.08) times the number of shares outstanding on that day (to keep the math simple say 1M). The market cap of this company on that day is $13,350,000.

Now, if the price of the stock falls to $12.98 the next trading day, it's worth $12,980,000. If you owned this stock you lost $370,000. Since market cap is a stock variable, you cannot add the value at one point in time to the value at another point in time. But you can average the two. Now, this is a very simple approximation. As Iggningot says, profit ratios may change since the combined companies will operate differently. So an average value is just a guess. But, so is every forecast future market value. Is this guess less reliable than a detailed analysis of how investors will value the combined company? Not really. As Bill@Sabre says, one analyst's guess of the discounted present value of combined earnings will be $2+$2=$5. When the actual post merger earnings turn out to be $2+$2=$3. These numbers are additive because earnings is a flow variable.

If the company had earnings of, say $6M in the last year and earnings of $8M in the current year, its earnings in the two years were $14M. We can add these two numbers because earnings is a flow variable. Earnings represent an accumulation of value over several points in time. While market cap is a store of value at a give point in time.

Thanks for your comments,

~V

2008 Mar 03 06:14 PM

[tyroni:]

your presumption "The markets have had plenty of time to anticipate effects of the merger and incorporate them into the prices of DAL and NWA." is complete trash and meant to skew the results of your analysis. The values of both companies at YE 2007 were near the $13 mark as is now and before specific consolidation talk. The stock values then RAN up to approx. $18 each as aniticpation of a merger grew in the market, which has now cooled in the past 9 business/market days and the price has dropped back down to $13. A representation of these values would only be accurate if you had used stock values as of their February highs. Do you think everyone here is stupid or what?

2008 Mar 04 03:44 PM

[frm airline...:]

Maybe the worst analysis I've ever seen. Two companies of roughly comparable market caps merge, and the combined market cap is the average!!! How about the combined mkt cap is roughly double? The error comes from averaging the currently outstanding number of shares instead of adding them! If every NWA and DAL shareholder got one share of the combined company, there would be roughly twice as many shares outstanding. Thus, even if the share price didn't rise, as it would be expected to do in light of the presumed synergies of the combined company, the market cap would be double that of each stand alone company. Thus, on a risk adjusted basis, the merged company has a better RAD; the opposite conclusion than the one reached by the author!! How does this stuff get published?

2008 Mar 04 07:18 PM

[Victor Cook:]

Tyroni,

Your right, if instead of the most recent closing prices I had used the stock values of DAL and NWA at their February highs, the combined market cap would have increased from $3.759b to $5.390b. This would have significantly increased their risk-adjusted differential from -4.9 to -1.7. This in turn would have placed the merged companies just inside the bottom-feeder upper limit of -2.0. Still uncomfortably close to the tail end of the distribution, but in a much stronger position.

But markets don’t use the highest prices in a relevant series to anticipate the effects of a merger. Typically the most recent closing prices best reflect investor assessments of stock values. That’s why I used the closing prices on Friday February 29, 2008. I did not, as you imply, intentionally “skew the results of my analysis” using the lower prices. If the prices on that date have been $18.00 I would have used them instead. If the prices had been $25, reflecting investor sentiments that the merger was a good thing, I would have used those prices instead.

My belief in writing posts published on this site is that SA readers are both smart and well informed. Thank you for your comments.

~V

2008 Mar 05 03:35 PM

[Victor Cook:]

Frm airline exec,

Of the nine airlines in my analysis, the M&A transaction costs of AMR, DAL, NWA, UAUA and LCC over the period from 1979 through 2007 exceeded their market cap on December 31, 2007. By a large margin. The combined transaction value of their M&As totaled $29.6b. Their combined market cap at the end of 2007 was $15.5 billion. That’s a ratio market cap to M&A costs of 52%.

But, then you were talking about a DAL and NWA merger. The combined M&A costs of DAL and NWA over the same period were $16.6b. Their combined market cap at the end of 2007 was $7.5b. That’s a ratio market cap to M&A transaction costs to of 45%. As a former airline executive, you must have access to inside information on presumed synergies of the combined company that I don’t know about. If so, please share it with me. If not, you may have confused a fatal error in my analysis with a wake up call to Delta and Northwest management.

Thanks for your comment. It motivated me to look more closely at the assumption that 1 + 1 = 1. Clearly I was wrong. In this case 1 + 1 = ½.

~V

2008 Mar 05 06:12 PM

[Independent...:]

Mr. Cook: At the end of you article, you said: "Will this analysis prove to be accurate? Probably not." That's an understatement. Definitly not. But why you are writing this? The combined airline of NWA and DAL will be number One. By the way, how does negative market share work in the real world?

2008 Mar 10 01:56 AM

[kurious:]

I think Victor is Short Dal and or NWA. 1+1= 1/2? Victor you might want to see out a new profession. :)

2008 Mar 10 08:52 PM

[Victor Cook:]

Independent E. G.

Okay, definitely not! But it’s also definitely true that the combined NWA and DAL will not be number one in share of market value anytime soon.

“But why you are writing this?” These posts are designed to test my theories in the market for ideas. Writing a book is not enough. To build a bridge between marketing and finance I think it’s necessary to do the construction in public. Seeking Alpha helps me do that. As you point out, some of the conclusions of enterprise marketing analysis are controversial. But it if makes you think, it’s a good thing.

“How does negative market share work in the real world?” It’s called bankruptcy! Theoretically it’s an IF statement: m < 0.0 is returned as 0.0%.

Thanks for your comments.
~V

2008 Mar 11 12:23 PM

[Victor Cook:]

Kurious,

Actually I’m not short DAL or NWA. And I'm not long on LUV! I’m just curious to find out why investors continue to allow senior airline management to throw their money down the drain.

I appreciate it’s a sexy business -- on a par with America’s Cup. But the scale is so much more overwhelming and the implications so far reaching I think investors would have had enough by now.

BTW I’m quite happy with my position in the ivory tower. ;-)

Thank you for your comments.

~V

2008 Mar 11 12:40 PM

[Independent...:]

Mr. Cook:
Your theory on NWA and DAL's market share is not reflecting reality now. Do you agree? I don't think they have or will have negative market share. Do you agree? What's the value of your theory in dealing with such reality? Is it > or < than 0?

2008 Mar 12 12:35 AM

[Victor Cook:]

Independent E.G.

I wasn’t clear in my discussion of the market share metric. I’ll try to explain now what I failed to make clear in the post.

DAL and NWA “market shares” in the last table above are measured in two different ways. One is share of value (market cap). The other is share of revenue. And they also are WEIGHTED differently.

1) Share of Value: %SOV. Refer back to the first table in the post. In Q4-07 DAL’s market value was $4.00 billion. NWA’s market value was $3.52 billion. Their combined market value on that date was $7.52 billion. The market cap of all nine carriers was $28.84 billion. DAL’s stand-alone share of value was 13.9% [$4.00/$28.84] and NWA’s stand-alone share of value was 12.2% [$3.52/$28.84].

Based on their history of unsuccessful mergers and acquisitions I assumed the value of the combined companies would be the AVERAGE of their two stand-alone market values: $3.76 billion. This assumption reduced the market value of all nine companies to $25.1 billion. So, DAL+NWA’s post-merger share of value becomes $3.76/$25.1 or 15.0%. This is the %SOV in the last table in the post.

2) Share of revenue: %SOR. I followed the same pattern of reasoning here, except I worked with revenues rather than market values. And, in calculating %SOR I took a neutral position by ADDING the revenues of the two companies to get a combined market share of revenues equal to 27.4%.

The negative “shares” that appear in the last table are the differences between share of revenue and share of value. For DAL alone that’s -2.7 share points. For NWA alone that’s +1.3 points. For DAL+NWA the difference is -12.4 share points [27.4-15.0]. Here’s where the confusion crops up. This result is not a negative market share of -12.4 points. It’s a negative DIFFERENCE between %SOR and %SOV.

Sorry, I hope this clears up the confusion. Thanks for pressing me on this question.

~V

2008 Mar 12 10:44 AM

[tyroni:]

To suggest that the stock price and subsequent market value of a combined delta and northwest would be 3.8 to 5.4 B is rediculious! The asset values of the combined companies along far exceed these figures, and to agree that the value of LUV should be 9B is far to far reaching, in a similar circumstance NO ONE would pay 9B for the assets of an LUV. To pronounce that the management of all airlines except LUV is horrible is irresponsible as well, especially considering the FAA safety inspection issue which saved LUV millions at the expense of the public's safety (this is irresponsible and a gross mis-management).

FROM SA EDITORS: THIS COMMENT HAS BEEN EDITED TO REMOVE OFFENSIVE LANGUAGE. COMMENTERS THAT ATTACK THE AUTHORS OF ARTICLES THEMSELVES RATHER THAN THE AUTHOR'S ARGUMENTS MAY BE DENIED COMMENT PRIVILEGES.

2008 Mar 12 12:48 PM

[Lisa:]

Excellent article -- thoughtful and original.

2008 Mar 12 06:25 PM

[Ralph F:]

Given that airlines often have a lot of debt, I'd have thought that you'd have to use enterprise value instead of market cap.

2008 Mar 12 06:28 PM

[Victor Cook:]

Lisa, thanks for your comment. ~V

2008 Mar 12 07:37 PM

[Victor Cook:]

Ralph,

Over the years I've aslo thought about using enterprise value instead of market cap. It really complicates the theoretical underpinnings of my analysis. So far I haven't been able to think of a good enough reason to do that. Maybe you have one or two.

Thanks for your comment.

~V

2008 Mar 12 07:51 PM

[Independent...:]

The valuation of the two airlines (NWA & DAL) is something very tangible using asset-base or units per capacity and can be decided objectively if reasonable approaches are applied. Not long ago, US Airways put a few figures out in their bidding for Delta which were countered and rejected based on different calculations. Midwest rejected Air-Trans bidding and accepted TPG/NWA's bidding instead. If any theory can not stand the reality test, that "theory in the air" can not be used for "airlines". Simple as that.

2008 Mar 12 10:35 PM

[tyroni:]

So what's the value of the new Delta going to be? 17.7B, good luck pulling your head from your arse.

2008 Apr 15 10:37 AM

[Victor Cook:]

E.G. said “The valuation of the two airlines (NWA & DAL) is something very tangible using asset-base or units per capacity and can be decided objectively if reasonable approaches are applied.”

Tyroni said my analysis is “is complete trash and meant to skew the results.”

Let’s review the bidding to this point.

On February 28, 2008 DAL closed at $13.35. With 299.47m shares outstanding its market cap was $4.00b. That same day NWA closed at $13.43. And with 262m shares outstanding its market cap was $3.52b.

On April 16, DAL closed at $8.62. On that day Yahoo! Finance reported shares outstanding were 292.22m, so its market cap was $2.52b. That’s a 35% decline in market cap since 2/28/08. NWA closed at $9.55 with 236.43 shares out so its market cap was $2.26b. That’s a 29% decline.

If you added the 2/28/08 market caps of DAL and NWA they came to $7.52b. On April 16 their market caps added up to $4.78b. Over the ensuing seven weeks investors knocked $2.3b off their combined value, for a 32% decline.

If you guys bought at DAL and NWA (based on your theories) at 2/28/08 prices in anticipation of making a bundle once the merger went through, I guess you made a bad call.

Looks like investors don’t agree that the valuation of these two airlines is “something very tangible.” And it just might be that my analysis isn’t "complete trash ... meant to skew the results."

Thanks for your comments.

~V

2008 Apr 17 10:50 AM