Monthly Archives: January 2013

It is not uncommon to find companies with a sleepy quasi-monopoly: the product has become overpriced for what it is, but is has remained dominant on the market, or benefits from network effects.

In theory, if there’s no essential intellectual property in the way, someone should pop up with a fair priced competitor, and arbitrage the overpricing away. But often the barrier to entry, while not insurmountable, is sufficiently high that nobody bothers entering. Public companies with such “moats” tend to trade at a premium.

Dry castle moat at Carcassone, France

A dried up moat that may be difficult to defend (credit: user quinet on flickr)

Fund startups by shorting incumbents

If financed the old school way, the payback of launching a competitor may take a lot of time, because the large costs of building the product are incurred upfront, and building market share may be loss-making for a number of years before reaching critical mass, and then become only slightly positive requiring more years to get up to full speed, and eventually recover the initial investment.

One could imagine a model where the launch costs for the competitor are funded by the anticipated failure of the incumbent: that is before starting the product, the funding entity shorts (directly or via derivatives) the incumbent. They still need to finance the development, and provision for failure of the short (if the prices goes up for unrelated reasons), but if the product’s potential is recognised by the market soon after it becomes available, the payback will come much faster, because not only will the new company be on the slow road to profitability, but the upside from the short will come much sooner if the market is any good at discounting the disappearance of the moat and possibly losses due to the sleepy incumbent being run with high costs.

This is almost “free”, because the upside risk of the short is in theory neutral to this — the failure of a previously unpriced attempt at introducing a competitor should not change the valuation of the incumbent.

Ideal for open source…

This could work particularly well in the case of creating an open source replacement for proprietary software. Once a project is bootstrapped and used by enough entities, the benefits of open source come in full swing, and an ecosystem of support organisations can easily be sustained, at a much reduced costs to users of the software. The hard bit is to bootstrap the project, especially if it doesn’t scratch a programmer’s itch so as to spontaneously appear through community efforts. There is no obvious way to recoup the significant bootstrapping cost of the initial development, as anyone can start a support organisation once the product is available on an open source basis and so support should end up being priced based on costs that do not include the initial development.

But if you short the incumbent, the very appearance of the open source solution, which can be cheaper by an order of magnitude and thus see relatively quick adoption, may sink the incumbent sufficiently to recoup the project bootstrap costs by closing the short position relatively early in the process.

… or free internet services

It could even work for things like creating a competitor to a dominant social network. Let’s imagine a social network that has 500 million users and a market cap of 50 billions, that is a valuation of $100 per user. The service is free but it harvests monopoly profits on advertising and the commercialisation of user data, but nobody knows how to lure users from the incumbent to a more tightly run ship. The cost of reproducing the service (hosting and software) could be relatively small, because the value of the incumbent lies in the network effect, not the product itself.

Less advertising and less data resale (back to normal profits) is a soft proposition that is not enough to get everyone to move en masse, unless the incumbent really grossly abuses their position. One way to work around this would be to short the incumbent, estimate the reduction in value that would result from a loss of dominant position, deduct the set up costs for the service, and use the rest as bribes to break the network effect, that is pay the users, once, to switch services (e.g. by paying switching users a bonus after a few months of regular usage).

If we imagine a social network valued at 50 billions, with say 500 million active users and that an exodus of 20% of them would be enough to bootstrap their loss of dominant position and take their valuation to say 10 billions. If we short 20% of the stock we basically can pay each user something close to (or commensurate with) the valuation per user of the incumbent. The bribes can be targeted at the most active users who will bring their less active friend with them and for free. It may not take that much of a bribe to break the quasi-monopoly of the incumbent.

Where’s the catch?

The idea seems a bit too good to be true. Of the problems I can think of, there might be issues that a manageable short position would rarely produce enough profits, even in a realistic success scenario, so as to remain too risky to be worth attempting.

There might be legal problems with shorting a company while building a competitor, though probably not insurmountable with good lawyers and early disclosure, assuming a disclosure ordering that doesn’t kill the proposition can be devised. Besides it’s generally in the public interest for new entrants to be able to challenge incumbents, so regulators could possibly be convinced to welcome such practices.


This Slate article about demographics is a relatively good take on the topic, apart from perhaps excess pessimism about a possible population shrinkage.

It eschews a common mistake, where people think that modern societies have dramatically lower birth rates than in the pre-technological era, sometimes arguing some moral implications out of that. While it is trivially true if you look at gross numbers, it is also misleading, and ceases to be true when looking at birth rates in net terms, the number of children who reach childbearing age per women, which is a much more useful metric to understand demographic trends.

In net terms, we have merely moved, as noted in the article “from high death rates and high birthrates to low death rates and low birthrates”, that is pretty similar net rates. Without knowing the exact numbers, it can trivially be inferred from the fact that humanity grew pretty slowly for millenia, if at all during some periods and geographies, that the net rate has been between 1 and 3, and rarely much above 2 during humanity’s almost entire existence. Net rate bumps where healthcare was improved rather suddenly and before the rest of a developed economy’s features arrived, as happened in the late twentieth century in some developing countries, are a short lived historical anomaly.

It does seem that people adapt to the available technology to get to a rate within sustainable boundaries (there is evidence, for instance in medieval and renaissance Europe, that people were already actively operating below the then available capacity). Given that it seems pretty universal and not dependent much on local culture, historical time, or on the degree of technological development, one can ponder why is it so? Do resources constraint or some other perennial equilibrium forces play a role, or is it just a curiously persistent coincidence?