Huberman and Adamic's group over at HP Labs have come up with another of their just-nails-it empirical studies of behavior on the net. Get your PDF here, loaded with equations and real life data all neatly graphed. (Via Howard Rheingold)

I would particularly like to draw the attention of Mr. Ross Mayfield to Figure 4 on page 3, which would seem to show a Power Law pertaining to distribution of connections in e-mail activity. Might that, perchance, be some anomolous behavior in a medium that you'd surely pick to support flat, convivial, egalitarian social networks? And if even there, why not in other social networks? Give it up while you can, Ross. You're doomed, I tell you. Doomed!

8:56:54 PM    

My previous piece on venture capital funding was already long enough, so I left one obvious question hanging: If you guys are forced to pass up deals that may be good businesses because your funds are closed end, why not structure them as open-ended, so you can let the bets ride? There are reasons...

First, most of the money going into venture capital comes from so-called institutional investors. Think pension funds, insurance and annuity companies, university endowments. These guys are using us as part of an asset diversification strategy, that will typical encompass asset classes all the way from T-bills to speculative items like venture capital and new IPOs. Part of running such a strategy is that you need to periodically rebalance assets, as various classes under- or over-perform, and the overall portfolio grows or shrinks.

These guys won't leave their money with us for an indefinite term, and we really want to them to keep investing. So if we wanted an open-ended partnership agreement, we'd have to grant on-demand redemption rights (like your conventional mutual fund does.) A bit of a problem when your underlying assets are illiquid! Even more of an issue when knowing what funds you will have available to invest in the future is crucial to deciding what commitment you can make now to an early stage company that will need multiple rounds of funding later.

The only way to pull this off would be to keep a large fraction of the fund's committed capital in the form of cash equivalents. However, our limited partners don't want to pay us our management fees just to put the money into T-notes or repos. Remember, they already own lots of those, and they hired us to diversify away from them. And holding onto that low yielding paper also hurts our own measured rate of return, which is how the limiteds judge us the next time we come knocking and asking for a commitment to a new fund.

This is one other approach that's been tried: get a different type of investor, specifically, retail investors. There's a structure called a closed-end, publicly traded fund. The fund itself has a fixed size determined when it begins, but its shares are traded on a public exchange. If a share owner need to get out of the investment, you can sell the shares to someone else, although it may be at a discount to the underlying asset value. Meanwhile, the profits on the portfolio are plowed back indefinitely. This is a venerable, though somewhat out of favor, way to structure a mutual fund whose underlying assets are also publicly traded. Back at the peak of the bubble, the well-known VC firm Draper, Fisher, Jurvetson organized such a fund with venture investments as the underlying assets. It was called meVC. Probably the ability to sell such a thing to retail investors should have been a sign of impending doom, because shortly thereafter the decline began, and things got rather unpleasant at meVC. Now there are lawsuits all over the place and the plaintiff's attorneys are clustered in flocks around the carcass. It will be a chilly day in Hades, or the peak of the next bubble, before that's tried again.

Lastly, while it might be attractive in some situations to keep the money in for a longer time, it's in no way a free lunch. Remember that return (discount) rate? It just keeps compounding. And the risk discounts get bigger as you go further into a murky technological and economic future.

6:26:39 PM    

One of my partners found this while searching for something else and thought I'd enjoy it. He was right. Business relevance? Well, if you can homebrew such a gadget for under $500 BOM cost over at Fry's and Target, it defines a tough benchmark for PC/CE convergence products.
10:19:04 AM