internet companies vs. state-owned enterprises in China

Recently Beijing announced it wants to take equity positions in the major internet companies in China and place Communist Party officials on their boards of directors.

What’s going on?

I see two general possibilities.

Some background first.

Deng’s economic reform

In the late 1970s, Deng Xiaoping realized that the Chinese economy was too big to be controlled through central planning.   To grow it had to adopt Western economic (but not political) methods.  So he began to allow the market, not doctrinally-correct political cadres, to dictate the direction of expansion.

A major issue he faced in doing so was that, say, three-quarters of Chinese industry was owned by the state.  These companies were rudderless, and hopelessly inefficient–but they employed tons of people.  If large numbers lost their jobs all at once, the ensuing social instability might threaten the rule of the Party.  Therefore, economic progress had to be tempered by the need to avoid this outcome.  And this in a nation without sophisticated macroeconomic tools to control the pace of growth.

The result over three+ decades has been a Chinese economy that lurches between boom and bust, depending on the temperature in the state-owned enterprises.  The strategy has generally been successful, I think, with the state-owned sector now representing less than a third of China’s overall output.


–China’s internet companies have become large enough that their actions, intentional or not, can accelerate the speed at which state-owned companies shrink.  So they need to be monitored much more carefully than in the past.  This is the benign interpretation, and the one which share prices suggest the market has adopted

–China’s internet companies have become large enough to generate “creative destruction” in large enough amounts to threaten the economic control over China exercised by the Communist Party itself.  If this is the case, then the oversight over domestic internet conglomerates will be much more draconian than the consensus expects.  That would presumably result in considerable PE contraction for the firms being controlled.

My guess is that the first possibility is much more likely to be the case.  But I think we should watch the situation closely for new hints about Beijing’s intentions.

how online ordering is shaking up the food business

I’m less and less a fan of the Wall Street Journal as time goes on.  Writers seem to be more interested in filling up the page than providing astute analysis.  But there is an interesting section on “The Future of Food” in today’s edition.

What strikes me:

–the threat to supermarkets isn’t simply the shift of dollars to online vendors.  A disprortionately large portion of supermarket profits come from two sources:  house-brand goods; and impulse purchases from endcaps and, more importantly, the shelves along the checkout line.  Even if the online order goes to the supermarket, the chance of selling for $2 during checkout the soda that’s $.40 as part of a six-pack in the beverage aisle is lost.  Also, at least in these early days, online purchasers choose many more national brand items than house brands

–consumers in general, and online orderers in particular, are increasingly gravitating toward healthier foods.  This means less sugar.  The difficulty for food manufacturers is not only switching sugar for some non-sugar thing that tastes the same.  There’s also the volume that must be replaced and the physical properties of sugar that are lost–like that it makes ice cream soft and bread less prone to mold.  Sugary foods are on the way out, but its not clear that the traditional brands will be able to hang on to all their customers during the transition

–ex Amazon, the food ordering app business is complicated by the fact that customer expectations/behavior can be far different from what, say, a fast food conglomerate or coffee chain expects.  Again, the risk of losing customers during a transition period is there

investment implications

Of course, everything has a price.  So at some point traditional food manufacturers and supermarket chains will be cheap enough that all the potential bad news will be more than baked into the stock price.  But I suspect we’re not at that point yet. The issue is operating leverage.  The arithmetic of distribution company profits is such that a 2% drop in sales can mean a 5% fall in pre-tax income.  If the sales lost carry double the average margin, however, the negative effect on profits will be multiplied by at least twice (most likely more).

Silicon Valley backlash?

I think we may be at a watershed moment in terms of the acceptability of the corporate behavior of tech/internet-related companies.

Up until now, it has been enough for investors that Silicon Valley produce increasing profits.  Institutionalized poor behavior on the part of the firms’ managements–whether that be violation of some employees’ civil rights or less-than-ethical treatment of customers or shareholders–has made little difference to their stocks’ performance.

Uber is perhaps the poster child for this phenomenon, which has also been, aptly, I think, characterized as “fratboy” behavior.  But now Uber appears to be losing its license to operate in London, which holds 5% of the worldwide active users of the taxi service, because of its not being a “fit and proper” operator.

The case of Facebook (FB) is just as interesting.  Founder Mark Zuckerberg announced plans last year to give a large amount of his stock in FB to a charitable trust that he and his wife would run.  In order to preserve his majority control of FB despite divesting a large chunk of his shares, he proposed that each A share, the ones with super voting power that insiders like him hold, be “split” into one A + two C shares, the ones with no voting rights.  Zuckerberg could then give away the C shares, representing about 2/3 of his wealth, without any decrease in his 53% voting control of FB.

The board of FB appointed one of its members, Marc Andreessen, a developer of the Mosaic and Netscape browsers, to represent third-party shareholders in this matter–to ensure that this restructuring would be fair to them.

Institutional investors sued.  During discovery, they found among other things, emails between Andreessen and Zuckerberg in which, far from defending third-party holders,  Andreessen appears to be coaching Zuckerberg on how to present his proposal to the board in the most favorable light for him.

Today, FB announced it’s dropping the restructuring plan.


If I’m correct about a fundamental change in investor sentiment, what does this mean for us as investors?

At the very least, I think it means that the business-is-what-you-can-get-away-with attitude (borrowing from Andy Warhol) of many tech companies will be penalized with a discount valuation.  It may also prevent some early stage firms from being able to list–preventing employees from cashing in on what they’ve built.  On tech/internet’s notorious anti-woman bias, I’m not sure.  After all, the investment business isn’t that far ahead of tech in eliminating this form of prejudice.



21st century retailing: my trip to Home Depot

This is another mountain-out-of-a molehill thing.

We have Toto toilets in our house.  Toto is the leading brand in Asia and has been making significant inroads in the US over close to two decades.  Yes, they’re the toilets that play music, heat the seat, double as a bidet and make fake urinating noises (a Japanese must)–but we just have plain old toilets.

The other day, I went to the local Home Depot, which, by the way, sells Toto toilets, to get a replacement part for one of ours.  A friendly employee showed me where the replacement parts were–all aftermarket brands, not Toto, but that was ok with me–and which was the right one. The replacement didn’t look much like the broken part, but the employee assured me that it would work.

It didn’t.  And, in fact, in looking back on my trip, the HD employee may, strictly speaking, have only told me that that was all they had.  If so, kind of embarrassing for me, since for most of my working life I was on the alert for verbal gymnastics aimed at papering over problems.

Rather than launch a telephone search for a plumbing supply store in the neighborhood that might carry the part I needed, I found it on Amazon.


Around the same time, I found I needed a replacement part for a Weber grill.  Same story.  HD sells Weber grills, but not replacement parts.  So, after a wasted trip to the local HD store, I ordered from AMZN.


What’s interesting about this?

In the early days of the internet, there was lots of speculation about the “long tail,” meaning that e-retailers like AMZN would make most of their money from selling obscure items that potential buyers couldn’t find in bricks-and-mortar stores.

A great story   …just not the case back then.  Just like bam, online exhibited the “heavy half” phenomenon, i.e., 80% of the business came from 20% of the items.


But maybe the long tail is beginning to come true.  It’s not because weird stuff that no one really wants has suddenly come into vogue.  Instead, I think computer-driven inventory control programs that eliminate slow-moving items from a store’s offerings may have gone too far.  Yes, carrying fewer items has the beneficial effect of requiring fewer employees and less floor space.  But at some point, the process begins to have negative consequences, as well.

For instance, it’s training me not to go to a physical DIY store, so I’m not passing by enticing end cap displays or being tempted by the sparkly high-margin junk arrayed along the checkout line.


My experience as an analyst has been that any cost-control measures always seem to go too far.  They work for a while, but the continual application of the same process somehow eventually ends up creating the opposite of the intended effect (yes, experience has made me a Hegelian, after all).  This may be what is starting to happen with inventory control programs that retailers use.

If I’m correct, this is another plus for AMZN.


21st century retail: my trip to Rite-Aid

I went to Rite-Aid the other day to get some Aleve.  I was away from home, in a rural area more than 100 miles from the nearest Costco, and not at a place where I could get same-day delivery from Amazon (270 Aleve tablets for $18 ($0.07 each).

I had several choices:

–100 generic (naproxen sodium) tablets for $9 ($.09 ea.),

–200 generic for $14 ($.07 ea.)

–100 Aleve for $11  ($.11 ea.),

–200 Aleve for $20 ($0.10 ea.), or

–270 generic for $14.50 ($0.05 ea.).

I took the 270.

What really struck me was the fact that I got the final 70 tablets for a total of $0.50.  That’s $0.007 each.  Assuming that Rite-Aid wasn’t paying me to cart them away, the most it could have paid for the tablets was $0.007 apiece.  Multiply by 270 and you get about $1.90.

Doing the analysts’s mountain-out-of-a-molehill thing, and assuming Rite-Aid buys from the manufacturer, I conclude that $1.90 is the most it could have paid for the container of tablets I bought.

The $12.60 that remains is the cost of packaging, distribution, promotion …plus profit.  (Overall, Rite-Aid isn’t making money, even though it has a positive gross margin of about 22%.  SG&A pushes it into loss, so delete “profit” from the packaging… list.)

That Rite-Aid can’t make money despite a 600%+ markup says a lot about the company.  But it also says something about bricks-and-mortar retail, the way Rite-Aid gets its products in front of customers.

This is the AMZN success story in a nutshell:  all it has to do is deliver a $2 item to a customer and spend less than $12.60 to do it.


My trip to Home Depot tomorrow.








Wall Street Journal on internet pricing

The Wall Street Journal has an interesting, if somewhat disorganized, front-page article today on how internet retailers vary the prices and the selection of goods they offer to different customers.  The article addresses several different topics, which it doesn’t clearly distinguish:

different pricing in different countries.  Who doesn’t do this?

dynamic pricing, where prices of goods or services change depending on time and the availability of inventory.  Airline tickets  or hotel rooms are the model of this type of price change.  There have also been attempts by bricks-and-mortar stores to vary pricing by time of day, so that a gallon of milk costs 50% more at midnight than at noon–though I’m not aware of a single successful experiment of this type.

customer assessment.  Four aspects:

—-The merchant uses the IP address of the customer as a proxy for zip code and offers different merchandise based on area demographics.

—-The merchant offers different merchandise depending on the device the customer uses to access the site–phone, tablet, PC.

—-The merchant offers different merchandise or pricing based on how the customer arrives on the site–mobile app, social media, search engine.

—-The merchant varies merchandise/pricing depending on the customer’s on-site behavior.

proximity to bricks-and-mortar alternatives.  Here the online merchant varies pricing, based on how far away the customer is either to its own bricks-and-mortar store or to those of its rivals.


The first three of these seem to me to be staples of traditional retailing.  So it’s hardly surprising that once enabling technology became available these tactics would emerge in the online world.

The fourth is the problematic one.

There may be legal or ethical problems with charging higher prices in poor or rural areas where bricks-and-mortar alternatives aren’t readily available.  It seems to me, though, that this pricing behavior trains consumers not to use the sites of these merchants but to automatically go to online-only merchants like Amazon instead.

AMZN and the new (generic) Top Level Domain names on the internet


The Internet Corporation of Assigned Names and Numbers (ICANN) unveiled the list of approved applicants for a new set of generic Top Level Domain (gTLD) names it proposes to issue.  The addition is intended to expand the number of such TLDs significantly from the current twenty or so (.com, .net, .gov etc.).


The move has two goals:

–to introduce TLDs that use non-Latin based characters.  This means languages like Arabic, Chinese, Japanese or Russian will have domain names in local language characters for the first time.  This will make it easier for people whose first language is not Latin-based to use the internet.  After all, that’s where most of the future growth will be coming from.

Users may not be conversant withLatin-based characters, for example.  And they may have to take elaborate steps with their access devices just to be able to type them.

–to expand the available universe to TLD names beyond those that ICANN finds useful, and to align naming with the specific needs of internet users.

squatters need not apply

…penniless ones, at least.  One provision of the application process is that any entity bidding for the right to control and administer a specific name (that is, to say who can use the TLD and who can’t) already have the infrastructure in place to do so.

who’s bidding?

Here’s the list.

what catches my eye

–Despite the ICANN precautions, most applicants appear to be companies formed specifically to acquire and hold TLD names., which is funded to the tune of $100 million by venture capital and private equity, is an example.

–there’s little match between the 1900+ TLDs requested and the most expensive search terms–like attorney, insurance or rehab–that internet advertisers buy.

–traditional advertising “grabbers” like “Free” or “Buy” aren’t in great demand, either.

–the most highly contested names are “Apps,” with 13 applicants, and names like “Home,” “Like” and “LLC.”

–the largest companies appear content to stake out their company name and the names of their chief brands, so that no one else can control them.  Other than that, they’ll wait on the sidelines to see the process evolve.

AMZN is the one exception

AMZN has applied for over 30 TLDs in Latin script, as well as filing 10 of the 116 requests for non-Latin TLDs.

Some of the names are what you’d expect, like “.Amazon,” “.author,”  “.book.”

That AMZN also wants “.AWS,” “.cloud,” “.fire,” or “.app” probably isn’t too surprising, either.

But it is also asking for “.bot,” “.box,” “.coupon,” “.drive,” “.deal,” “.free,” “.got,” and “.now”.

I think the AMZN move makes a lot of sense, for it anyway.  The company has more spare server capacity than just about anybody, so the cost for it to corral these names isn’t high.  And this many turn out to be just like the earliest days of the internet, when ordinary (albeit geeky) people bought basic domain names like “” and “” just to use for themselves–and later were able to sell them to corporations for tons of money.

The next step in the ICANN process?  …a seven-month call for comments.  The “list” link above will take you there if you want to chime in.