Sean’s blog

When I was blathering on in my long and poorly written post a week ago about alternative financing in software / internet I missed something obvious:

There is a huge opportunity for SaaS companies with subscription revenue to borrow based on the security of their subscriptions - much bigger than the (bad) example I chose of advertising-supported websites.

Some examples Veronica Belmont, Eric Olsen, Brad Feld.

The compare people app just asked me if my aunt smells better than my classmate and if one of my former co-workers is hotter than my cousin. I personally am 0-2 on “pretty eyes” which is kind of a bummer.

Um…..yah!  Wouldn’t it be nice if this kind of logical distinction were made rather than the current silly platforms of the democrats and republicans.

 Libertarians and communitarians (to continue this unjustified generalizing) are different character types. Communitarians tend to be bossy, boring and self-important, if they’re not being oversweetened and touchy-feely. Libertarians, by contrast, are not the selfish monsters you might expect. They are earnest and impractical–eager to corner you with their plan for using old refrigerators to reverse global warming or solving the traffic mess by privatizing stoplights. And if you disagree, they’re fine with that. It’s a free country.

From Libertarians Rising (TIME).

They aren’t hip, but old-school bulletin boards (powered by old software like vbulletin) have a considerable audience.  It would be interesting to buy lots of them, add some social networking goodies, improve the software (ajax, etc) and sell ads on the resulting site(s).  Why this would work:

1. They are likely cheap to acquire
2. Better software could increase visits
3. They are mostly hobby sites and don’t have professional ad sales, leading to low CPM
4. Monetization could be further improved with behavioral and contextual targeting where appropriate

Often the founders of such boards are an important part of the community - they would have to be offered some amount of continued control as well as an ongoing financial stake, but this could be figured out.

Here is a list of the largest bulletin boards on the net.

Is that it smells like chocolate. Of all the awful things that a big city can smell like, chocolate is pretty great. The reason is the Blommer chocolate factory, which is fortunately just a few blocks from my apartment.

The smell is particularly strong in the winter, it seems, and most noticeable on the west side of downtown, or along the river to the east (because the wind blows up the river).

View Larger Map

The EPA has complained in the past about the amount of particulate matter being released from the factory.  This is a violation of the same rules from which coal power plants are granted an exemption.  I would much prefer the exemption not be granted to  coal plants because they smell bad and contribute to the greenhouse effect, rather than chocolate dust, which does not contribute to the greenhouse effect and also allows me to wake to the smell of brownies each day.

Four recent investments by Bill Miller, the legendary value investor at Legg Mason, are pretty interesting.  This guys has traditionally picked common publicly traded stock positions and has an incredible record of beating the market in his fund Legg Mason Value Trust.

The investments were in early-stage startups Ning, Zillow, Sermo and online currency trader Oanda.  Each was for a significant amount of money $25-$50M at a high valuation.

The investments were made out of the Legg Mason Opportunity Trust which “Invests with a flexible strategy and is not restrained by investment style, type of security, industry sector, location, size or market capitalization, investing primarily in U.S. common stocks” and holds pretty big stakes in Amazon and IAC. About 10% of the 8B fund, or about 800M is in private companies and hedge funds.

I suspect that we will see more non-traditional venture investors investing in technology and consumer web companies in the near future for a few reasons:

1. They have lots of money, which only keeps increasing since many institutional investors place large weight on historical returns
2. Buyout returns are starting to dry up
3. Hedge fund returns have proven to be less predictable than expected (somehow every 10 years investors forget that hedge funds have risk associated with the large negative skew of their returns)
4. Venture investments have large positive skew, which makes them appealing as a small part of a large investment portfolio

Too bad Bill Miller doesn’t blog.

Here is the most data-rich story on these Legg Mason investments that I have been able to find.

Most VCs have very concentrated portfolios with 7-10 companies / VC partner.  Thus, they need to be pretty careful when they make an investment.  In a sense the evaluation is wasted effort since there are at least a handful of VCs all doing the same work.

It seems like there exists an opportunity for a new investment strategy of funding smaller amounts to a greater number of companies based on a more standardized investment criteria.  One could value a company based on some multiple of traffic, do a background check on the founders, write a check and then move on.

There is potential for adverse selection, however, with the more qualified companies going to a traditional VC leaving only the less qualified companies for this new strategy.

Not all startups are hard to value and sell. Take a startup 7-11 or Subway, for example. In fact, there are entrepreneurs that start new franchises of that sort and sell immediately upon opening, having done a lot of the hard work of establishing the business and making a profit in the sale.

The difference between a startup technology company and a startup 7-11 is that almost anyone can operate a 7-11. As most VCs can attest, it is not that easy to replace the management team at a young technology company.

One interesting trend in finance has been the rise of Whole Business Securitization. Securitization refers to the practice of a company selling an asset to a 3rd party (usually a shell LLC created specifically for this purpose) which then borrows from a bank based on the value of the asset in order to get the cash to pay the original company (originator) for the asset. This bit of financial engineering works because the bank is attracted to the asset unencumbered by the decisions and liabilities of the company and provides debt financing at a very low rate.

In a whole business securitization, the originator sells the majority of their assets to the 3rd party LLC. In reality, the majority of “whole business” securitizations have been done on restaurant franchise companies, including Domino’s Pizza and Dunkin Donuts. The banks find lending against the security of the franchise agreements and brand quite appealing, especially when they are legally removed from the encumberances of the original company through the magic of securitization. Here, as in the case of the 7-11, significant value is unlocked by separating the owner from manager of the asset. The original company (like Domino’s) is paid by the special-purpose company that holds the assets according to a contract called the servicing agreement. If something went wrong with the original company (like bankruptcy or incompetence), the banks that lent capital to the special purpose company could appoint a new servicer.

Maybe there are assets within an internet startup that could be monetized by somebody outside the context of that startup. The two that come most readily to mind are domain names and traffic. Say there is a hot new site that has achieved some initial success (say 1M visitors / month). What is the revenue potential? Say a modest assumption of $5 / cpm, 10 views / visitor / month, that is $720K / revenue / year assuming very poor and generic monetization (viagra and dating site ads).

Assuming the above scenario if I were a lender I would be willing to lend $1M to the company through the following scenario. The company would sell the website to a special purpose entity for $1M. I would lend the SPE $1m which it would then use to pay the company. The SPE would sign a servicing agreement with the company that pays the company all revenues over a certain amount (maybe $200K - this is the principal payment) beginning in a certain year (say year 3) and over $0K (company gets everything) from years 0-2. In exchange for the revenues the company would maintain and grow the website. I would charge the SPE interest, in the first several years the interest could be added to the principal so that the company wouldn’t be drained of cash. My investment would be very safe because if something went terribly wrong I could put the loans to the SPE into default and appoint another servicer to maintain and update the site.

This isn’t a lot of money, but I built a pretty significant margin of error into the calculation and digital companies need less and less money.

I’m not sure that it would work exactly like this, but its an idea, and it has worked in other industries.

Begin trading. It is very difficult to value an asset in isolation. The efficient markets of today depend on each asset having closely comparable analogues to which its value can be judged on a relative basis (and hedged against).

With bonds investors start with a “riskless asset” – basically a very short maturity treasury bill. Then appropriate risk premiums are added to the bond to account for interest rate risk and default risk. For example, a newly floated government bond from Afganistan could be compared to the bond prices of other undeveloped countries, the appropriate premiums added.

In the language of stocks investors frequently talk about comparing P/E ratios or EV / EBITDA ratios to measure the value of a stock relative to that of another.

Having closely comparable assets allows speculative investors to engage in arbitrage, which makes the market for such assets more efficient. The liquidity in one asset can trickle into nearby assets.

Begin by establishing a market-traded index based on aggregate venture capital returns. An index already exists - it would be trivial to begin trading an option based on the value of that index each year. It would simply settle for the value of the index at the end of the year. Venture returns have some very desirable properties for portfolio managers, and there are a wide range of people in the venture industry that could benefit from some hedging, that I believe such a contract would generate enough liquidity.

Then start trading options based on the ultimate sales price of a handful of startups. Focus on those which have lots of public information available – namely those that are planning on an IPO in the next several months and therefore have already published financials in an S-1. Investment in such options could be hedged with the overall venture market contract described in the prior period.

Then trading could begin in a few very public startups such as facebook, where there is a lot of public knowledge available.

So on and so forth, once liquidity is established in a more visible asset, liquidity can be established in a smaller and earlier companies, with the prices kept in line relative to one another by arbitrageurs and the wisdom of the crowds. Eventually we will get early enough that the market can reasonably estimate the value of early stage companies.

Such an option does not need to exist for each and every startup to be useful in valuing that startup. If the market value for a comparable startup is available then metrics can be used to draw a comparison between the startp without a market value and the startup with a market value.