Welcome back to the Conglomerate Junior Scholars Workshop. Today's paper is The (Not So) Puzzling Behavior of Angel Investors by Darian Ibrahim. Darian is an Associate Professor of Law at the University of Arizona James E. Rogers College of Law and teaches and writes in the fields of Business Associations, Law & Entrepreneurship and Contracts. Today's expert commentary will be provided by Larry Ribstein, Barbara Black, George Dent and David Hoffman.
We invite readers to comment on the paper (and the comments) in the comments section of this post. In the interest of running this workshop like a physical world conference, no anonymous commenters, please.
The abstract for the paper is here:
Angel investors fund start-ups in their earliest stages, which creates a contracting environment rife with uncertainty, information asymmetry, and agency costs in the form of potential opportunism by entrepreneurs. Venture capitalists also encounter these problems in slightly later-stage funding, and use a combination of staged financing, preferred stock, board seats, negative covenants, and specific exit rights to respond to them. Curiously, however, traditional angel investment contracts employ none of these measures, which is a marked departure from what financial contracting theory would predict. This article explains this (not so) puzzling behavior on the part of angel investors, and also explains why some angels are moving toward venture capital-like organization and adopting venture capital-like contracts in the process.
TrackBack URL for this entry:
Links to weblogs that reference Conglomerate Junior Scholars Workshop: Darian Ibrahim's The (Not So) Puzzling Behavior of Angel Investors :
1. Posted by Jeff Lipshaw on July 11, 2007 @ 8:12 | Permalink
I thought Darian's paper was beautifully composed, and did not give it the time it deserved. Nevertheless, I have some thoughts about it and somewhat less so about the series of comments.
Some preliminary is probably in order. I spent two years of my life (1998-99) in the last throes of the bubble, in which anybody who had an internet content idea thought he or she was going to be a gazillionaire, trying to develop a law practice in Ann Arbor focusing on what I called the "dead zone" of entrepreneurship - the pre-angel or angel period before there was substantial VC funding that could at least pay the lawyers involved. I was involved in a number of organizations that encouraged entrepreneurship - the local venture club was called the New Enterprise Forum and I sat on its board.
So with that, here are a couple observations:
1. One of the differences that nobody points out, and you find references to it in the literature, is that VCs invest in industries and industry segments, and angels invest in enterprises. That is to say, VCs will make a pick of the best technologies and management teams, but they are betting on pharma, or telecom, or wifi, or whatever.
2. The legal treatment of this is double-dipped "looking for rationality in all the wrong places" (apologies to Johnny Lee). I propose the thesis that we are imputing rationality, as observers from the outside, to a system that is largely random or serendipitous. It's one thing to do that as economists, and it seems to me clear that we quickly have to move off rational actor theory and into some form of behavioral theory to get at what angels are doing, even if we stay in economic models. Lawyers double dip this by then thinking there is something about the contract (or lack thereof) that purports to embody the otherwise not quite rational investment that is the dog rather than a barely attached tail. Take a look at Douglas Hofstadter's vision of the "careenium" - by which when we step back we see a system, even though at a sub-micro level, it's all just ball-bearing bouncing off each other. It seems to me that the most important data here is that angels are very rich people, many of whom got lucky themselves. It's appropriate I'm up here in northern Michigan, where the Little Traverse Bands of Odawa Indians just opened the Odawa Casino and Resort, because I think the model of the angel's expectation of return on the angel investment is more akin to how I feel about the return on my $100 worth of chips at the blackjack table.
3. The reason there's no robust literature on the LAW of angels is that there is no law of angels. At least with a significant VC investment, you might have litigation on the down side, a la Benchmark or Equity Partners (see Gordon Smith's article on this). But my thesis is that an angel is no more likely to want a legal remedy for her investment than I am for my losses in the casino.
4. This lengthy comment, as might not be surprising, is not the first time I have thought about this. I have two working papers out there (in low levels of coherence). The first, Why the Law of Entrepreneurship Doesn't Matter, was a quick reaction to Gordon's paper on why it did. It poses the question whether there is a difference in the orientation to rule-following between the entrepreneurial mind (or culture) and the legal mind (or culture). The second, Aboutness, Thingness, and Morphosity: A Pragmatic Ontology of Formal Systems in Law, is a broader approach to the way lawyers (versus other humans) see patterns in data and ascribe form and meaning to them. There's an implicit teleology going on when social scientists want to theorize and the data doesn't quite fit. We end up with terms like "incomplete contracts" as though there were any complete contract.
Sorry for the long comment. I love the topic and the discussion.
2. Posted by Darian Ibrahim on July 11, 2007 @ 8:14 | Permalink
Thanks to Christine and Conglomerate for hosting this great workshop and for including my paper. I am most grateful to Larry Ribstein, Barbara Black, George Dent, and Dave Hoffman for reading my paper and offering these incisive comments. All are excellent and offer much food for thought. I’ll be back soon to respond to each commentator. Following Sasha’s lead, I’ll use a separate comment for each response, and be brief.
3. Posted by Jeff Lipshaw on July 11, 2007 @ 8:47 | Permalink
Having read the commentary more closely, the view that most closely approaches my own grenade-tossing is George Dent's "second,...."
4. Posted by Darian Ibrahim on July 11, 2007 @ 10:30 | Permalink
In response to Larry Ribstein:
1. As might be expected, while I spend pages describing where angels fit into entrepreneurial finance, Larry is able to do this in two sentences: “What do smaller entrepreneurs do before the VC firm arrives but after they’ve maxed out their credit cards? They look for angels, of course.” Exactly right. This is the space that angels fill.
2. One of my explanations for traditional angel contract design is that non-aggressiveness facilitates later VC funding. Larry asks an interesting question: Why does this constrain angels but not later-stage VCs, who have investment bankers/public shareholders following them in the case of exit by IPO? I should address this in the paper, which treats the later-stage VC, a bit myopically, as the last investor. Having said that, it doesn’t seem that VCs are similarly constrained because IPOs are designed to unwind their preferences, the most notable example being automatic conversion of the VC’s preferred stock to common. In other words, public shareholders are not forced to live with VC aggressiveness, just as VCs don’t want to be forced to live with angel aggressiveness.
3. A big one: Larry, like the other commentators, wants to know more about AIOs. He asks: “Are AIOs the next evolutionary stage of angel investing? Or do they handle a different set of firms?” The truth is, I don’t know either. To jump ahead and agree with one of George’s comments, this paper involves more conjecture than would be ideal given the lack of data on angels (particularly on AIOs). For traditional angels, the data-collection task has been difficult because of their preference for reference-led investment and, consequently, the desire to remain anonymous to everyone else who might wish to empty their pockets. It’s telling that when Van Osnabrugge and Robinson were writing their book on angels, some angels interrogated the authors as to how they were found and then refused to be interviewed. AIOs, on the other hand, do not share the same penchant for secrecy (which is also understandable – they need a steadier deal flow mechanism) and therefore should be easier to study. Still, there are few studies to date of the newish AIOs (a creation of the last decade), which forces me to theorize from mostly anecdotal accounts – for now. I’m delighted that Larry wants to know more about AIOs, because George Geis (Alabama, visiting Virginia) and I have been talking about studying them. We would hope to compile all of the excellent questions asked in this workshop, along with some others, and find the answers empirically.
4. I am, however, able to somewhat speak to another of Larry’s questions about AIOs at this time. He asks about their organizational structure: Are they non-profits or for-profits, and if the latter, what kind (e.g., LPs, LLCs)? From what I gather, the vast majority of AIOs themselves are probably non-profits, but very little happens here. In other words, most AIOs are not the investment vehicle – they don’t pool member funds and invest in the group’s name. Instead, the member angels who are interested in a particular start-up will get together and invest separately. So, the AIO is the vehicle by which member angels organize to screen investments (and perhaps where dues are paid and administrative expenses billed), but the actual investing occurs outside the AIO. It would be very interesting to know what organizational form these side investments take. Do the angels invest in their own names, or do they form a new company for the investment (and is it an LP, LLC, or something else)? From what I gather, it’s a little bit of each, but this is another question ripe for empirical study.
5. Posted by Darian Ibrahim on July 11, 2007 @ 10:54 | Permalink
In response to Barbara Black:
1. Barbara makes a number of very interesting comments, but I want to key in on one in particular. That is, that Barbara found my answer to the traditional angels puzzle more successful than my answer to the AIO puzzle. I suppose that, in a way, I am pleased by this. As noted in my response to Larry’s comments, there is still much to discover about AIOs, and I’m confident that a richer story will emerge with more data. Each commentator asks excellent questions about AIOs, which are important to explore both in connection with this paper and in thinking about future work. But Barbara’s comments give me a chance to stress that the genesis of this paper, and its primary focus, is on traditional angels. Traditional angels still account for some 70% of angel investments (roughly $17.5 *billion* dollars annually) and get an undeserved rap for imprudent investment behavior. As Linda Loman would say, attention must be paid to these investors. Also, primarily because the trend is toward AIOs (a point emphasized by George that I’ll elaborate on later), it seems clear that these angels and this model will move center stage in future work. Which makes it all the more important, I think, to explain the traditional angel model now. Also, perhaps it is too soon to sound the death knell for traditional angels. While the trend is toward AIOs, it remains to be seen whether a stable balance between the old and the new will result (and where that balance will be). After all, I stress that the angels differ in terms of their non-financial motivations, which suggests that not all traditional angels are ready to join AIOs. Of course, empirical work would shed light on whether the field is expanding, or whether the same angels are simply moving from one context to another. I note that the total amount of angel investment appears to remain steady, around $20-$30 billion each year, which bolsters the same angels, changing contexts story.
2. Regarding Barbara’s two minor points at the end of her commentary:
a) The evidence suggests that rapid growth start-ups will be organized as corporations rather than LLCs. Two excellent papers on this are Joseph Bankman, The Structure of Silicon Valley Start-Ups, 41 UCLA L. Rev. 1737 (1994), and Victor Fleischer, The Rational Exuberance of Structuring Venture Capital Start-ups, 57 Tax L. Rev. 137 (2003).
b) Regarding how VCs address overvaluation in previous rounds, I assume that they simply decide for themselves on a proper valuation when they invest, but that this produces some friction with the entrepreneurs, whose more optimistic valuation was bolstered by the angels’ willingness to accept it. An interesting sidebar is that while VCs as a general rule disfavor complicated angel rounds, one "unclean" term that VCs might be willing to accept is convertible debt. Why? Because it allows valuation to be deferred until the next round of investment. So, in essence the VCs might accept one thing they don’t like (complicated terms) to avoid another thing they don’t like even more (overvaluation).
Many thanks to Larry and Barbara for their terrific comments, which I hope these responses have done some justice to. I’ll return to address the comments of George, Dave, and Jeff soon.
6. Posted by Barbara Black on July 11, 2007 @ 11:23 | Permalink
Thanks for your response to my comments. Re (1): I thought about suggesting that you divide the paper into two and in this one simply focus on the traditional angel, given the importance of the traditional angel. That I didn't make the comment in the first instance shows my ambivalence about it, but you might give it further thought. Then you wouldn't have to deal with comments like mine, which boiled down to its essence is that I liked the first part so much that the second part was, by comparison, disappointing.
7. Posted by Jeff Lipshaw on July 11, 2007 @ 11:36 | Permalink
One other thought on AIOs. The boundaries between FFFs (friends, family, and fools) on one end and traditional VCs on the other are blurry at best. I'd suggest that organized angel activity is more akin to very early stage VCism than angel activity. There ARE some very early stage and seed money VCs, an example being the Enterprise Development Fund in Ann Arbor (one with which I am familiar), but as we know, unlike angels, these funds are using OPM (other people's money), and unlike in angel investing, there will be accountability to investors, and selection and screening expertise that is the early stage VC's stock in trade. What are the key differences (other than "own money" versus OPM) in that blurred area between very early stage VC activity and angel activity? Here are two off-the-cuff guesses at generalizations:
1. VCs still are investing in industries, so the hot areas get early stage VC money, and entrepreneurs plodding along in the more mundane areas look for angels.
2. VCs expect (demand) sophisticated management teams within their industries. A good team in a good industry get early stage VC money; a marginal entrepreneur or management group gets angel money if at all.
8. Posted by Darian Ibrahim on July 11, 2007 @ 12:48 | Permalink
In response to George Dent:
1. George thinks that my explanations for traditional angel contracts may be incomplete. He suggests that I further explore the importance of fiduciary duty as a backstop against entrepreneurial opportunism in the absence of contract. I agree that fiduciary duty is worthy of further exploration in two possible respects: 1) as a viable legal claim (e.g., the angels’ common stock offers fiduciary protections that the VCs’ preferred stock does not), and 2) as a viable threat (because entrepreneurs are too cash poor to defend a lawsuit). I discuss in note 134, p. 24 that minority oppression claims by angels are probably losers, but the idea that angels could check entrepreneurial opportunism through the threat of suit is interesting to consider.
2. George thinks (and Jeff Lipshaw agrees) that hubris has a lot to do with angel behavior. This reminds me of Joseph Bankman’s views on VCs and a “gambler’s mentality” in structuring start-ups as C-corps rather than flow-through entities (where early stage losses could be recouped). The interplay between Bankman’s paper and Vic Fleischer’s paper on the same subject is directly relevant here. Bankman acknowledged that VCs might have rational reasons for preferring C-corps over pass-through entities, but his emphasis was on the irrationality (and hubris) of this behavior. Vic’s paper essentially flipped the argument and emphasized the rational over the irrational. When guest blogging on this issue over at Truth on the Market, I wrote:
“I think Vic’s article is instructive on the craft of writing. It’s tempting to find irrational reasons for start-ups to form C corps (we’ll never have losses!), and as Joseph Bankman documented in The Structure of Silicon Valley Start-Ups, 41 UCLA L Rev 1737 (1994), a “gambler’s mentality” probably does have something to do with it. But when sophisticated players like venture capitalists are involved, I tend to favor rational over irrational explanations for behavior. In fact, I’m taking this approach in a new paper to explain the puzzling behavior of angel investors.”
Therefore, on hubris, I do acknowledge it, but I do not emphasize it. Instead, following Vic’s lead, I favor rational explanations for angel behavior over irrational ones. While there is no Bankman-like paper for me to respond to, this paper is very much a response to the conventional wisdom on angels.
3. George’s final point is on the trend toward AIOs. I think there are three interesting items to explore here: 1) To what extent AIOs will siphon off traditional angels looking for a steadier stream of investments (and easier access to each other); 2) To what extent AIOs will begin to invest other people’s money; and 3) what the ramifications of the latter move would be. I’ve tried to address the first of these items in previous comments; I’ll save the second and third for my response to Dave.
9. Posted by Jeff Lipshaw on July 11, 2007 @ 13:25 | Permalink
Darian, I realize I have inserted myself in this discussion, but what the hell! And you haven't had a chance to respond to me yet, even though you graciously said you would. But my view of rationality does not posit irrationality on the other side. Instead it's where you interpose the presumption of rationality such that it is subject to a meaningful algorithm in a meaningful predictive model. Trying to model rational behavior that way at the very granular level of the individual angel's contract is going to be tough. But if we step back we may well see a macro pattern to which we may impute self-interested rationality (the purpose or teleology of my earlier comment). To paraphrase Richard Posner (once again), at that level it may well not be a solecism to speak of all those angels out there as rational frogs.
10. Posted by Darian Ibrahim on July 11, 2007 @ 13:48 | Permalink
Jeff, my apologies for being slow to respond. It's been a busy day on this and some other fronts! But I'm working my way through the four sets of commentary, hoping to address overlap with your points where possible. Your thoughts on what really distinguishes AIOs from early stage VCs are great, and I plan to discuss this in the response I'm about to write to Dave.