Alphabet and Google

Peter Klein has an interesting Rand Journal piece (pdf) on conglomerates:

This paper challenges the conventional wisdom that the 1960s conglomerates were inefficient. I offer valuation results consistent with recent event-study evidence that markets typically rewarded diversifying acquisitions. Using new data, I compute industry-adjusted valuation, profitability, leverage, and investment ratios for thirty-six large, acquisitive conglomerates from 1966 to 1974. During the early 1970s, the conglomerates were less valuable and less profitable than standalone firms, favoring an agency explanation for unrelated diversification. In the 1960s, however, conglomerates were not valued at a discount. Evidence from acquisition histories suggests that conglomerate diversification may have added value by creating internal capital markets.

In other words, today’s Google announcement isn’t as crazy as it may sound.  Here is further positive evidence on conglomerates, and Glenn Hubbard also thinks the 1960s conglomerates were largely efficient.  Here is some evidence, however, that conglomerates tend to be less innovative.  Scharfstein and Stein are less positive more generally.  Here is some evidence that the non-Google divisions will receive favoritism in the allocation of capital within the conglomerate.  That all said, conglomerates are understudied in microeconomics, in part because they are hard to study.

What do you all think of the news?


peak ad-tech. With ad-blocking coming in in with iOS 9, the value of these ad tech companies is going down.

Great name

The greatness of the name will be decided based on the success or failure of the conglomerate. If Alphabet does well, in five years we will read sentences like: "Once again Page and Brin proved prescient, this time in selecting a name that reflected the varied strengths of the vast conglomerate."

If Alphabet does badly, we will read sentences like: "The origins of its failure lie in the very name chosen by Page and Brin; they created an alphabet soup of disjointed companies with no cohesive vision."

I think that the name does not mean much and it would be good if they spent $1M on some small branding company to come up with the name as it would have spread the wealth a bit.

I'd take even money odds betting against the combined Alphabet subdivisions beings worth anything more than 25% of the Google division by 2025. Assuming Google Cars becomes Alphabet cars that's probably the best hope.

I think the bets Page aims to take might implode or explode. I'd put the chances of a bigger-than-Google bet as a possibility, but it will be really random.

In tech, the bets are always long shots, with massive rewards, with the ability to scale massively quickly usually in areas with no direct competition where you get to create the market. AirBnB, FaceBook and Pinterest all had no direct competitors (see Peter Thiel's monopoly thesis).

Traditional business, lets use mining, Zara and Coke as examples, take years to build capacity. The maximum size a company can grow to is limited by physical constraints, like building a widget, building stores, investing in infrastructure. BTW, that is the path Amazon has taken, to own as much infrastructure as possible -

Alphabet is a decision by Page to take more longshot bets where the risks are massive, the rewards equally so, and the competition is not head on. These sorts of bets either make m/billions (Android) or go nowhere (Google Wave). Google Cars I would wage will be sold to car makers like windows was to PC manufacturers. Having the ability to leverage existing car manufacturers infrastructure would make it global in scale very quickly. This massive risk taking makes Alphabet less Berkshire Hathaway, with solid portfolio investments, and more gambler gone wild, doubling down on cray wild shots day after day.

If one of Google's crazy ideas, from space elevators to wifi balloons to driverless cars, takes off, it will likely at least match in revenue. By 2025? Probably not. By 2050? Likely.

How the market reacts to constant failure is going to be interesting to watch, as will Page's determination to see it through. Because while 10 more Google Waves before a single Android might spell the demise of the concept and sacking of Page, if Page retires/dies/has health issues/ decides being a billionaire is enough any time soon, the idea will almost certainly fall apart.

The two strategies google follows are:

1. Extract as much as possible out of the firehose of search ads
2. Search for the next firehose

My hope is that Google will spin-off its non-search business lines in pieces to Alphabet. If you want to bet on Sergi Brin and Larry Page, own Alphabet. If you want to bet on Google Search, own Google. If you want to bet on Google car, own Google Car Inc., etc.

The proper viewpoint is that of a company’s shareholders. Today, investors can hold stocks in a diversified portfolio of public companies very easily and cheaply. In the 60's, it was more difficult. Shareholders can adjust their aggregate exposure to individual stocks to whatever level they wish. In an world where portfolios can be adjusted with a mouse click, there's no reason for any conglomerates.

One of Google's biggest assets is its tremendous library of internal software tools. Even if the spinoffs are linked in a Zaibatsu structure, its gone be orders of magnitude less seamless to integrate that IP across the structure. I'd want to bet on Google Car, because it can easily utilize BigTable, Deep Dream and Bazel. Otherwise I don't see what its competitive advantage is.

Is the lack of direct shareholder pressure a feature rather than a bug though?

Long live the Sausage King of Chicago

That sort of flexibility has to be the point of this whole thing, right? Just adding another layer of management won't be the way forward.

Matt Levine on conglomerate discounts:

Excellent analysis, although I think it gives insufficient credit to the possibility that dumb money acting irrationally is driving the undervaluation of Yahoo, and to the possibility that one or more of the firms has been overvalued by analysts.

Well Alibaba IPOd well above the break even point and Yahoo Japan is public, so that can't be the case. Also, dumb money being that big a driver of valuation for one of the most liquid stocks in the world? Highly doubt it.

I'm reminded of a time when I was a teenager when the shares in Valueline were trading for less than the book value of the fund. Because all Valueline did was hold shares of stock, I asked my dad how that was possible. If you had enough money, you could just buy Valueline and liquidate it for a large profit.

My dad said "That would take a huge amount of money. Nobody thinks that's going to happen. Everybody thinks they're just going to keep on doing what they're doing."

Do conglomerates provide opportunities to pass money around from one company to another for tax purposes? Just speculating, especially since I have never heard this charge wrt Berkshire Hathaway. Or is Berkshire Hathaway a different animal because each (or most?) of its constituent companies is publicly traded?

A conglomerate is a firm that owns all or substantially all of the equity of multiple firms with unrelated lines of business. While a conglomerate may by itself publicly held, its subsidiaries are, by definition, closely held firms, usually with one (or in rare cases of conglomerates that own parts of joint ventures, a very small number of) dominant owners.

One of the big issues in the case of a conglomerate is that there is no market based valuation of individual firms that it owns, making it hard to compare the firms it owns with others in the same market by traditional financial metrics. Often one does invest in growth companies with current losses to offset current profits from other firms in the group for tax reasons.

One question that randomly comes to mind: Does this make it easier for anti-trust action in the future to split them up?

I see this as at least partially a way to find a promotion for Sundar Pichai (both to keep him and to increase his output). Another factor is that this might give more confidence towards potential acquisitions that they can maintain some level of independence (bullish for Twitter, Box, Dropbox, and other potential consumer internet targets?).

This also makes it easier probably to acquire various more tech-like companies that are currently being invested in by Google Ventures, but not a fit under the google consumer internet products group. Some possible targets include:
-23 and me
- Home Away
- Blue Bottle Coffee
- Optimizely
- Git Hub
- Zen Payroll
- Zenefits
- Stripe
- (some company I've never heard of in the medical research space)

Certainly. Refundable tax credits are often a money-losing firm's most valuable assets if it wants to convince someone to buy it.

There are several really good comments on the unrelated conglomerate phenomenon further down, but here I am commenting on the 50s and 60s, which were its heyday. During that period the DOJ actively intervened to prohibit almost any mergers and acquisitions in related business lines, consequently M&A activities tended to be diverted toward unrelated businesses. Why M&A activities? Growth through acquisition was a means of avoiding high corporate and personal income taxes (a tax deferred is a tax avoided, not to mention lower capital gains rates and the advantage of acquiring carry-forward losses). Because of tax effects, the market rewarded unrelated acquisitions even where they were associated with overall slower growth in total turnover from operations. The evidence cited by Klein can be interpreted as consistent with this story, which is about tax shelters; it may have little or nothing to do with productive efficiency as that concept is usually understood.

ITT was an exemplary conglomerate in its hey-day. The basic question is whether good generalized indirect management is better than more or less autonomous self-perpetuating managers subject to stock market discipline, or perhaps more accurately, if middling to mediocre generalized indirect management is better than more or less autonomous self-perpetuating managers subject to stock market discipline, because sooner or later good managers are replaced by less good managers.

Whatever its faults, the stock market is relentless in killing firms that have ceased to be profitable for the indefinite future. Mediocre conglomerate management, like government managers, frequently fail to meet this core imperative. But, conglomerates are frequently better a turning around and improving companies with good management than a lackluster incumbent self-perpetuating management team that can maintain an unimaginative status quo and suck economic rents out of the company that they are minding without adding any value. Ill managed companies that remain profitable that squander value like Microsoft come to mind. Yet, the importance of a strong product specific knowledge, rather than general management skill is key to business success in companies like Apple or Google or Facebook, casts doubt on how much of the general indirect management talent necessary for a conglomerate to thrive is present in the companies most likely to become our next big conglomerates. One of the keys to success in conglomerates that have done well is to have a very well honed back office bureaucracy that handles accounting, HR, legal, tax planning, finance, etc. far better than the average bear so that the original product specialists in the acquired companies can focus on the core business of the acquired firm while assimilating the rest to the conglomerate-borg. This might be a good plan for the cash rich company that processes Visa credit card transactions, or ADP which does payroll. But, it isn't obvious that Apple, Google or Facebook excel at those tasks for reasons other than the ability to throw lots of money at them without having to worry about becoming unprofitable due to the quality of their core products.

Some of the weaknesses of conglomerates are a function of weak GAAP standards and weak shareholder voice. Financial accounting rules are quite exacting for ordinary single line of business firms, but lack rigorous business segment financial data reporting requirements and this undermines transparency. In recent times, even minimal conglomerates, like Dardan Restaurants, which owned merely Olive Garden and Red Lobster, have felt pressure to divide to get that transparency and to make clear what investors are really investing in. Yet, if a conglomerate has five good companies whose upsides exceed the downsides of two bad companies, it is very hard for the market can do to convince its management to drop the losers.

In the case of Google, Chipotle (which was at one critical point a McDonalds subsidiary), and surely before long Facebook and Apple, the issue is that a successful business has mountains of money, can't distribute it to shareholders without triggering substantial shareholder level taxation, and have invested everything that it makes sense to invest at the moment in the core business. There are 23 companies in the S&P 500 with zero debt and substantial cash reserves. So, it is a question of investing in subsidiaries, or investing in treasury bonds.

In the previous generation, companies like GM, GE and Sears responded to that situation by expanding into financial services (integrating the business in the case of GMAC, entering new lines of business as well as supporting existing ones in the case of GE, and sabatoguing their core business by not accepting any credit card except their in house Discover card for many years in the case of Sears).

Today's movers and shakers would like to invest in more original and more promising ventures. But, the risk is that those more sophisticated and complex and unique ventures are more than they can handle.

I hope it works out.
I am tired of hearing the phrase "Super Excited".

I'm looking forward to "We are frantic like dachshunds about this opportunity!"

My reaction:

PageRank, Maps, and Android have created immense value, of which Google has captured far less than its fair share. Google does not get enough credit for making the Web work. That said, Google bought YouTube. Chrome is Firefox without the technical debt, based on the same open source engine as Safari (after Apple adopted it). Maps was a port to the web of an acquired desktop program. Android was an acquisition going nowhere fast until they cloned the iPhone. Gmail was a non-obvious solution to a problem, felt (and "dogfooded") intensely by Google engineers, that someone was going to solve. I am thrilled to see Page and Brin spend their surplus on moonshots. They have landed one of their own creation, or two if you count the creation of modern cloud infrastructure 2002-2008. It's a record to be enormously proud of, and one dominated *in breadth* by incremental improvements and failures.

"Chrome is Firefox without the technical debt"

This one I take the most issue with. Making a successful product that seems exactly like an existing product and dominating it is no small achievement. Using this same line of reasoning you could claim that Apple has done nothing interesting at all.

That being said, your argument is actually well in favor of the new structure. If google's main wins are through acquisitions, restructuring the company to make it easier to make big acquisitions (which it has), should be very bullish for their future.

I think it is good. Right now Google is like Microsoft, a big, old, profitable but not so excited company. There is no way you can make it young again. Page and Brin want to change the world through technology, just like Elon Musk. Buy reorganizing, they can focus on those small but exciting companies.

Page and Brin just activated their ejector seats.

The fat Google blimp descends a little more quickly now. Eventually it will land in a swamp of mediocre software which most engineers find repellent.

Someone should focus on running Google better.

Interesting that (the slimmer) Google and Microsoft both have CEOs from South India (as does Adobe).Long live nerdiness !

I think they realized they fucked up their motto so they are making a new one.

The guys at the top think that "search Google" is pretty much complete and want to focus their energy elsewhere. Putting it under one guy allows them to do so.

Hi Tyler,

After UCI, I worked briefly for a Mexican conglomerate with activities spanning autoparts, realestate, petrochemicals, and food. I would suggest, that like the 1960's and 70's, the low hanging fruit in the economy enabled large companies to leverage scale, access to capital, human talent and other considerations to develop seemingly unrelated opportunities. As the economy became more developed, the ability to focus and strengthen core competencies may have become more important. This may be that period for Google and the internet. The opportunity set for Alphabet, may be defined by its organizational imagination. I just wished I owned more GOOG!

Muchos saludos professor!

Their ability to game international tax regimes just skyrocketed above the already impressive level of gamesmanship they currently achieve.

Now they can, and will, pre-design networks of subsidiaries that do not violate required economic substance minimums, but allow for huge secured cash transfers and IP displacement. Getting around the transfer pricing rules under s. 482 and transfers under s.367(d) using their meta-corporate structure, they have effectively become an independent corporate network, the top of which sits above any particular nation (more so than was already the case).

Even if there are precedents, given the underlying technology in this particular case, it's likely fair to say we've evolved a new type corporate entity.

Does Alphabet also promise "don't be evil"? With all that AI under the hood, they really ought to promise "be good."

Zarathustra has returned from Mountain View, and is beyond good and evil.

Does Alphabet also promise “don’t be evil”? No. According to the press release, the new entity ("Alphabet") will have one division whith moto "Don't be evil!" (Google), the motos of other division will be dispered on the scale from "Be moderately evil!" to "Max Evil!".

This doesn't look to me like a conglomerate assembled by M&A (Buffett), more like a reorg with goal to give more autonomy to various current and future businesses (is Calico now an Alphabet company?) . I wish them luck.

I love the tax spin. This may actually turn out to be the biggest technology oriented, private equity firm in the world! Apple may be next... And all pre-tax! :-)

Conglomerates work when they can scale their core competencies. Often in the developing world you will find companies that cover a huge range of operations, this is because their core competencies are navigating their countries complex regulations and government controls, the actual businesses don't need to be efficient in themselves, the business owners can either copy or franchise or partner with developed country businesses for that. Other reasons conglomerates work - like GE for the last 30 years - they can access financial subsidies or, like the Korean Chaebols their sheer size allows them to get protection or extract rent from Governments.

Absent these drivers, conglomerates are usually prone to agency risks. There is a string of companies in the US and the UK that grew in the 1980s due to their share being premium priced. They would take over dull low PE businesses and then the business would be magically rerated at a higher PE. This gave the impression of constantly growing earnings supporting a constantly growing share price, resulting in rich rewards for their management. Hanson was one of these companies. Eventually however they would run out of acquisition targets and suddenly the whole thing would crash. I sometimes wonder if Berkshire Hathaway is simply the most successful example of this.

Is the economics of conglomerates the right way to analyze this? I associate conglomerates with M&A activities and industry consolidations. The directions Google is going seem more like venture capital on a very large scale. So let's think about the limitations of capital markets for the allocation of capital, economies of scale in venture capital, and the problem of what do you do if you have a creative organization that's produced a hugely profitable cash cow, and now you have the problem of what to do with all that cash, in an environment where your stockholders likely have poorer prospects for reinvesting the profits than you do.

it's like uber for dodging international tax regimes

Conglomerates can be successful in the short term when existing management and successful products remain intact. However, long term they tend to break down as central executive management don't know anything about the business they now own, and have corporate goals that may ignore the market situation of their various subsidiaries. Competition is dynamic, and companies need to constantly innovate and respond to market changes. This would explain why initially successful conglomerates of the 1960s (buying successful companies) failed in the 1970s (as markets changed and original owners/management left) and were divested in the 1980s. Management needs to be separate and empowered. Having independent minions is not what most empire builders want.

I think Berkshire is a poor example to use. It is more of a stock holding company. Buffet's ownership of much of Coca-Cola doesn't mean Coke execs are running IBM, P&G, John Deere, or another unrelated company. Even the businesses owned outright seem to be kept separate other than Buffet and Munger serving on their boards. The businesses aren't entangled like in a true conglomerate. Buffet also has a unique investing strategy (that many people claim to follow, but actually don't). While Berkshire is publicly traded, Buffet's prestige allows it to tolerate market fluctuations that would cause shareholders to force them to divest.

I think the comments others have made here that in developing countries, conglomerates may make more sense as their size allows them to better navigate treacherous regulatory schemes or provide easy source of capital to exploit new markets. In developed countries, this isn't a decisive advantage against your competitors.

The original claim of conglomerates is that they reduced shareholder risk by investing in unrelated industries. This can now be done cheaply through Mutual Funds, especially index funds.

Alphabet seems more like a venture capital firm to place money into new start ups that will be run as independent companies. It just happens to own Google.

Quick question: does anybody here think that the public capital markets would have funded any of Google's longshot projects if they had come about as independent startups?

Public capital very rarely funds any startups. Usually they progress from life savings/friends and family money to angel investor money to venture capital money and go public only once they are much better established. It also helps that most startups aren't in industries that are as capital intensive as say, building railroads or toll roads or whole industrial complexes to build complex capital intensive manufactured goods. They need money, yes, but a lot of their expenses are pay as you go.

Dr Klein should spend some time working for one.

Comments for this post are closed