Information Capital (in the 21st Century)

I am still not finished reading “Capital in the 21st Century” but the following introduction to Chapter 11 really got me thinking:

The overall importance of capital today […] is not very different from what it was in the eighteenth century. Only its form has changed: capital was once mainly land but is now industrial, financial and real estate.

One of my main intuitions about the coming changes is that “information” will become more important than industrial, financial and real estate assets. Or put differently “information” will be the capital of the 21st century and beyond.

This isn’t very well thought out at all yet but it is tantalizing because unlike land, machines, money or buildings information is non-rival. You and I can’t simultaneously farm the same plot of land. We can’t simultaneously use the same machine or occupy the same home or office. We can’t simultaneously spend the same dollar to buy something. But we can simultaneously access and make use of the same information.

It is not that there wasn’t information historically. It is just that its volume paled compared to what we have today stored in digital format making duplication and access free on the margin. There is some controversy over just how much information production has increased but among other things we know that every minute over 100 hours of video are uploaded to Youtube. And millions of people write blog posts such as the one you are reading (or upload songs to Soundcloud, or stories to Wattpad). And so on.

This raises some interesting possibilities. First and foremost among them is that we can overcome many of the issues of wealth and income inequality that I have been writing so much about recently by investing heavily in the information commons. One great example of that came to my attention yesterday thanks to a link Andy posted to about Open Source Seeds.

I will likely be writing more about the implications of “information capital” in the coming days as I just decided that this will be the topic for my talk at DLD NY next week.

Posted: 24th April 2014Comments
Tags:  information capital wealth inequality

The Aereo Case

Living in New York City and having cut the cord years ago, Aereo has been a terrific service for us. Yesterday the US Supreme Court started hearing arguments over whether the service is legal. Timothy B. Lee has an outstanding summary of the background of the case and what’s a stake over at Vox.

Based on the early reactions by Supreme Court justices it appears pretty clear that they don’t think Aereo is legal but want to come up with a narrow ruling that wouldn’t have implications for other cloud services. Even if they were able to come up with a narrow ruling, I still think it would be a bad outcome for consumers to have Aereo be illegal.

First, a narrow ruling that doesn’t have broader implications for cloud computing is possible. The judges could make the ruling specifically about the television broadcast signal and require that any enduser antenna equipment (Aereo’s argument) has to be within some very small distance of the premises of the recipient.

Second, it should be clear why any kind of broader ruling against re-transmission of content could be a nuclear bomb for the web and for cloud services. For instance, accessing the web through a proxy service might wind up being illegal, which includes any form of cloud browser.

Third, what’s really at stake here are the retransmission fees that existing cable systems pay to broadcasters. That whole market structure is antiquated and is harming innovation in the US. Even a narrow ruling against Aereo will keep that structure in place.

Fourth, there is an opportunity here — one that the Court is unlikely to address — to reclaim bandwidth which is a public resource. TV broadcasters could be allowed to make good on their threat to remove signal from the airwaves in return for giving up any rights to spectrum (which could then be used for data transmission).

Fifth, if the right market structure came to pass Aereo’s business would cease to exist purely by market forces. Imagine a situation in which you have high last mile bandwidth at a reasonable price and all TV content is available for streaming. In that world, why would anyone pay $8/month for a DVR? It would be called Dropbox or Google and either be free or part of a storage service you already pay for or would be provided as a value add by the broadcaster.

The bottom line then is that this whole case exists only because of an outdated industry structure. Any ruling against Aereo, however, narrow further entrenches that structure.

Posted: 23rd April 2014Comments
Tags:  aereo supreme court

More on Wealth and Randomness

Yesterday I wrote about startup wealth and randomness and got a lot of good comments. One of them led me to an important way of summarizing my argument about the role of randomness:

If the world were a simulation and you ran it many times, you would see a company like WhatsApp emerge many times. But only a few of the runs would have the same founders (ditto for the name of the company, the exact implementation, etc)

While I wrote the post specifically about startup founders, employees and investors, I believe the same is true for many other top earners including corporate managers, artists, athletes, etc. For instance, someone takes on a difficult project inside a company and succeeds, setting off a steep career ascend culminating in a CEO position. If you looked at multiple “runs” of the world there are some (possibly many) in which the project fails and the person is fired — even if they are the same person and put in the same effort.

Because the economy is increasingly global with less friction for commerce (virtually none for say apps), we are winding up with power law distributions for outcomes. This is quite different from when local businesses dominated. If you made shoes, even very good shoes, in a pre-industrial world you might make more than the next shoe maker but not a lot more. So the distribution of shoe maker income was likely close to a normal distribution. Compare that to say the game app economy — if you write a successful game app it can be several orders of magnitude bigger. And so there we have a power law distribution.

Why am I writing about this? Because it is an important component of the transition from an industrial society to an information society. And because the old solutions, such as high inheritance taxes, won’t work well in a world of increased mobility and I also don’t believe in feeding existing government structures. More of this in future posts.

Posted: 22nd April 2014Comments
Tags:  wealth randomness

Startup Wealth and Randomness

I have been writing a lot about wealth and income inequality and I am reading Thomas Piketty's Capital in the 21st Century (review forthcoming as soon as I am finished). So it seems appropriate to ask how I feel about the wealth being created by tech startups for entrepreneurs, early employees and investors (which of course includes VCs like myself). I have hinted at this here on Continuations in the past in a post about Twitter but it is worth clarifying. I believe that a large component of this wealth is random — or if you so want: luck.

Maybe the perfect illustration of that random component is Brian Acton, the co-founder of Whatsapp. He famously tweeted about being turned down by both Twitter and then Facebook as an engineer in 2009. If either one of those jobs had come through it seems highly unlikely that Brian would have joined Jan Koum in the fall of 2009 to work on Whatsapp. Which in turn would have made it pretty unlikely that Brian would have made hundreds of millions of dollars when Facebook acquired WhatsApp.

My personal story is similarly full of randomness. I spent all of 2003 trying to buy a traditional software company with the goal of then Internet enabling it. Together with a friend and partner we got super close to buying the leading trucking software company in the US which was headquartered in Cleveland. The deal fell apart at the 12th hour over a sales tax liability (that later turned out to be trivial — I will at some point blog about the details). In any case, had that deal happened I would have been in Cleveland instead of teaming up with Joshua for the wild ride that was delicious and subsequently joining USV.

If you want to argue that this entrepreneurial or investor wealth creation is not random to a large degree, then you would have to believe that people like Brian (or myself) have some skill or ability that is thousands or even millions of times larger than that of others. I personally find that preposterous. Just to be clear, I am not denying an element of risk taking, smarts, etc. — clearly you have to work for a startup or become and investor to begin with if you want to have a shot at this (as in the joke of the person praying for years to win the lottery only to finally hear God speak “buy a lottery ticket”). Beyond that though much depends on being in the right place at the right time.

At the moment  randomness is amplified significantly by the winner-take-all characteristic of many markets. The leading service or app in a category with network effects can be an order of magnitude or more larger than the next competitor and can do so on a global scale. Or put differently, the potential size of the lottery tickets has increased substantially. We have had a similar period in history during early industrialization.

PS Some further clarification on the randomness aspect. I am not arguing that the emergence of a company such as WhatsApp is unlikely (in fact the opposite). I am talking about the probability of a particular startup becoming that and of a particular person to be a founder of or investor in that startup.

Posted: 21st April 2014Comments
Tags:  wealth startups randomness

Liberating (Innovation in) Mobile

Chris Dixon recently wrote a post about the decline of the mobile web and the shift to apps and how that is hurting innovation. Fred expanded on that theme in his post on the mobile downturn. So it is worth thinking about what we need to get to a model of permissionless innovation in mobile. Here are some changes that might make sense

  1. An unencumbered open source mobile operating system (Android is not that due to the control exercised by Google)

  2. Ability to side load apps / have competing app marketplaces

  3. No private APIs that are accessible by Google/Apple only but not available to other developers

  4. Requiring IPV6 on phones to facilitate direct routing between phones

  5. Requiring ability to mesh network between phones

  6. Allowing resale of last mile residential bandwidth

Only #1 on this list could be accomplished without government help and once it succeeds would make #2 and #3 unnecessary. But given the current concentration in the OS market these should be considered as potential regulatory remedies.

#4 - 6 are all potential regulatory responses to the concentration of bandwidth providers in the US. An interesting entry point to #4 and #5 would be from an emergency response perspective — it could maintain network connectivity even in very large disasters with power outages. #6 would allow a phone mesh to easily connect to the net at the closest onramp.

Posted: 17th April 2014Comments
Tags:  mobile innovation regulation

Wealth and Speculation

At the moment I am writing about why I think the industrial system is breaking down as its main components are no longer functioning properly and sustaining each other. In particular I wrote about the growth and increased concentration of wealth and its relationship to our present day democracy. Today I want to suggest another aspect of the changes in wealth and how it relates to the allocation of capital.

Let us start with a look at what the components of the financial asset side of private wealth look like overall (based on Federal Reserve data, in 2013 Billions)


What stands out are the growth of equities and pension claims which together account for almost 3/4 of all private financial assets. Financial assets in turn make up 70% of all personal assets with the remaining 30% consisting primarily of real estate.

Now as we saw in the previous post on wealth and voting, the bottom 90% in the US have only 25% of the total wealth and the bulk of their holdings are pension claims. Conversely this means that almost all of the equity ownership is held by the top 10% of the population. Within the top 10% in turn the bulk of ownership is concentrated in the top 1% of the population (in fact, the top 0.01% alone seem to own as much as 10% of all equities).

Now as I have argued separately, we are living in a deflationary world largely due to the impact of information technology on the economy. Let me be quick to point out that by this I mean a world in which “real” prices are declining based on supply and demand for goods and services. This does specifically not consider some massive potential monetary change, which one cannot rule out given the tremendous amounts of money created.

In such a world we should expect to see many more speculative bubbles. Why? As it gets harder to make money by investing in productive capacity, returns are sought out in temporary price movements even when those are known not to be sustainable. This is entirely rational behavior as each of these bubbles is a redistribution opportunity between those who have better timing and those who don’t. Each bubble itself is a zero sum game though.

Now there is strong anecdotal evidence of this. We had the dotcom bubble. Then we had the housing / mortgage bubble. As it turns out this increased volatility shows up very clearly in the overall private wealth data. Here is a chart that I put together based on the same Federal Reserve data with a measure of 10 year volatility

The volatility measure is the sum over the absolute year-to-year changes for the preceding 10 years divided by total wealth that year.

For the time being I expect more volatility ahead given the reasons above. My point here is not that we should address speculation or the problem of wealth and voting each in isolation. Quite the opposite, I am suggesting that all of these are connected to each other and are part of the breakdown of the industrial system.

Posted: 16th April 2014Comments
Tags:  finance wealth capital speculation

Writing Software Applications In-House

There is something very exciting going on in the world of software: a shift towards developing new application software in-house (as opposed to relying on third party offerings). For instance, at USV we have developed the linksharing on and our new analyst application process. Why does that make sense now? Because writing small applications has become so easy that it is today’s equivalent of writing an Excel macro.

Chris Dixon has a good post yesterday where he describes some of the trends that have made this possible, including on demand infrastructure, lots of API based services, open source libraries and scripting languages. I would add to that a database layer that makes it easy to deal with semi-structured data, such as MongoDB (we are investors).

As a result the tradeoff has shifted from using a complete third party application to building it yourself for many projects. What you gain is fine grained control over design, functionality and integration with other systems you are using. What you pay is having code that you need to maintain. So if you are going that route there is a real premium on keeping your code short and super readable.

I am fully aware that we have gone through phases like this before. For instance in the early days of the PC that’s how I first earned money — writing in house software for the personnel department of the local Siemens branch in my hometown.

Still it feels like this time the in house approach will be here to stay and if anything be extended to non-programmers through services such as IFTTT. One of the key drawbacks of the in-house move during the PC days was the fragmentation of data. That is largely eliminated today as the data resides in the cloud in any case.

In addition to the cost and complexity of in-house development having come down massively there is another crucial reason for the shift (which also makes me believe it will be longer lasting): software plus data is increasingly the key competitive differentiator.

PS Back to my posts on the economy tomorrow

Posted: 14th April 2014Comments
Tags:  software in-house strategy

Wealth and Voting

Earlier this week I started writing about wealth and financial capital as part of examining how the various components of the industrial system are beginning to break down. I will have more to say about the financial capital portion of wealth in upcoming posts but wanted to first provide more data and more context. As was pointed out in the comments Thomas Piketty provides a lot of data that he and others collected and form the basis for his book “Capital in the 21st Century" (which I just started reading). So here is a wealth graph that goes all the way back to 1770 and also shows population growth

What this comparison shows is that much of the total wealth in existence today has been generated post 1900, which supports my claim that the industrial system wasn’t really fully in place until then. In fact from 1920 to 2010 the population grew by less than 3x and wealth grew by more than 15x.

The distribution of this accumulated private wealth is extremely uneven, with just 0.1% of the population controlling 20% of all private wealth in the US today. And the top 10% of the population account for 75% of wealth. But in terms of ability to use wealth the picture is even more lopsided than that when you consider what the wealth for the remaining 90% consists of as the following chart by Emmanuel Saez and Gabriel Zucman shows image

Essentially the wealth of the bottom 90% consists of pension claims and housing both of which are highly illiquid (something we will revisit in the context of changes to income).

Now why does all of this matter? One crucial way it matters is that the governance portion of our current system is based on voting with an original idea of equal representation. This was written into the United States Constitution, Article One, Section 2 which sets up the apportionment of seats in the house of representatives and establishes the decennial population census. The initial exclusionary definition of who was to be counted was later expanded with the Equal Protection clause of the Fourteenth Amendment. As late as the 1960s the Supreme Court upheld these basic tenets with several “one man one vote” rulings.

More recently though two changes have come together to undo this principle, resulting in representation based on money instead of individuals. First, the cost of attention has been rising steadily and has exploded with the advent of the Internet. The cost of Super Bowl advertising is a good proxy here as it measures roughly how much it costs to get a minute of attention in the US. Here is a chart the Business Insider put together


Second, changes in legislation have made it easier for money to be spent by politicians and individuals / groups supporting them to buy attention. This goes as far back as Buckley vs. Valeo in 1976 which took much of the bite out of the Federal Election Campaign Act. It really gathered steam with the Citizens United decision in 2010 which removed limitations on corporations and the recent McCutcheon case which did away with any aggregate limitations.

Do the effects of this show up in the election data? The following chart put together by shows what happened in 467 district races. The headline sums it up: the candidate who spent more money won 91% of the time.


For one possible legislative attempt to counteract the power of money you should check out the American Anti-Corruption Act.

There is also an interaction between the distribution of wealth and income that impacts the nature and size of the demand for products and services. I will tackle that in another post. Before that though I may dig into the composition of wealth to show the role of financial capital.

Posted: 9th April 2014Comments
Tags:  wealth statistic voting

Heartbleed: Social Change and Decentralized Identity

While gathering up additional data for the next post on wealth, here are some thoughts on the Heartbleed security vulnerability. At USV we sent out a notice of the issue to all the heads of engineering as well as all of the CEOs. This was easy to do since as part of the USV network we have these groups pre-defined. It was immediately clear based on the severity of the issue that this requires CEO level attention. For an example of an excellent response from one of our portfolio companies, here is a detailed disclosure by Twilio.

I have posted frequently in the past that crypto alone is not the answer to questions of privacy and security. Whenever I write along those lines someone shows up with a “but math defends us” argument. Well, cryptography math isn’t something abstract. It lives in code. And that code can and will have errors. There will inevitably be endless arguments about bounds checking and programming languages and so on focused on how the particular bug could have been avoided. Again this misses the point. Systems todays are layers upon layers and cryptography will never be the only answer. We need to focus as much if not more on protecting people rather than data and systems.

If we want to have a debate on the code and technology side it should be around the future of certificates and how they are issued, distributed, verified, revoked and used. There is a current mass process happening (hopefully) in which companies are scrambling to replace potentially compromised server keys which requires that they obtain new certificates. The certificates bind the public key to the identity of the company / service. This mechanism is build on a web of trust that goes back to root certificates. This seems like an area where block chain type solution could provide a truly decentralized solution. If I am being vague here it’s because I don’t know what the exact mechanism would be but this seems very worthwhile looking into. As part of that we can hopefully get a broad mechanism for tying keys to individuals (not just organizations) in a decentralized identity system. We had an early start on that with client side certificates in browsers which unfortunately eventually got abandoned.

For both the political and the technological debate I hope that Heartbleed serves as a wake up call especially to technologists. Again, the two key issues I would like folks to focus on are (a) the importance of political and social changes in preventing the abuse of data that has and will be leaked and (b) the necessity for a truly decentralized identity-key mapping infrastructure. 

Posted: 9th April 2014Comments
Tags:  heartbleed security identity

Wealth and Financial Capital

Last week I introduced my view that the various interlocking components of the industrial system are beginning to break down. The first one I mentioned was that “financial markets have been allocating capital into vast asset bubbles instead of into productive investments.” To dig deeper into this claim a good starting point is to look at wealth. The following chart shows the growth of private wealth in the United States, where private wealth is the sum of household and nonprofit organizations. The difference between the asset line and the networth line are liabilities. 


The amounts are in constant 2013 billions, using the CPI to adjust for inflation. The source data for the chart is published by the Federal Reserve (thanks to Dina Lamdany for compiling it). 

Several features stand out. First, private wealth has grown more than 8x during that time period (for comparison, US population grew roughly 2.3x over the same time from 140 to about 319 million). Second, growth in private wealth accelerated significantly around 1995. For the fifty year period from 1945 to 1995 the compound annual growth rate was about 3%. From 1995 to 2006 it accelerated to 5%. Third, the crash of the financial markets in 2008 stands out clearly. Equally clearly though private wealth has now exceeded the pre 2008 levels and has grown extraordinarily from 2012 to 2013, expanding by 12%.

One of the to dos here is to try to bring this data all the way back to the 1780s. Why? Well, at the time of the drafting of the US constitution there were fewer than 4 million Americans. The population has grown nearly a hundredfold since then. It would be interesting to know by how much wealth has changed over the same time period.

Thanks to research done by Emmanuel Saez and Gabriel Zucman we know in increasing detail how unequally distributed the growth in wealth has been in recent years. They recently published the following chart showing the percentage of wealth held by the top 0.1% of the population in the US (which requires greater than $20 million in networth)


So at the same time that wealth has grown significantly overall, the share of it held by the very rich has increased substantially and is now back at levels last seen in the Roaring Twenties. Again, it would be interesting to see if one can come up with estimates for what this looked like even further back.

Here is how this translates into absolute numbers, focusing on the period of growth. In the mid 1970s the share bottomed out at about 7.5% of a total private wealth of about $22 trillion for a total of $1.6 trillion. By 2013 it had climbed to 22% of about $70 trillion for a total of 15.4 trillion (the reason I am saying “about” is that I don’t have the exact decomposition of private wealth into households versus nonprofits). That is almost a 10x increase and a staggering number overall.

I will come back to the implications of this growth in concentrated wealth in subsequent posts. But first I will start digging a bit deeper into the composition of wealth which is also intriguing.

Posted: 7th April 2014Comments
Tags:  wealth statistics

Older posts