One little puzzle I have been pondering as of late is why some hedge funds whose core business is "special situations" - ie, the broad range of sins not covered by investing in public securities sometimes choose to have "deal guys" or "sourcers" in house and others do not, and under what circumstances either model makes sense. The reason I ask this is that the group of people I work with had a great deal of success in their previous place of work with a large group of sourcers in-house though the quality of sourced deals that I am seeing from them are, by and large, no better than the stuff that either randomly gets sent my way by old contacts or than I find in the public markets. The people who run my firm are from stupid or lackadaisical on costs so I have come up with a model of this particular decision and what might have changed in Asia over the last 5-10 years that could have made our current configuration sub-optimal.
For this model we have a few pieces:
The Firm (naturally): Seeking to maximize profits. Profits are a function of the return on assets as per the usual hedge fund or private equity structure.
The Expected Return of Private Market Transactions: Treated as a normally distributed random variable.
The Expected Return in the Public Markets: Treated as a normally distributed random variable.
The Sourcer and their compensation function: Compensation is equal to some share of returns at the fund level with a weighting towards the private deal portfolio and some fixed deal fee per private deal done. As such, the Sourcer makes more when private portfolio returns are high and more deals are closed.
The Analyst/Portfolio Manager: Cares strictly about returns at the fund level. Is indifferent between private and public transactions. Analyst/PM skill is defined by the returns on their portfolio, and with respect to private deals it is the expected excess return given that they do the deal. This means that a really smart guy can still have some private deals in the portfolio mix.
The Sourcer's acces to deals: Some constant - ie, the Sourcer will see x deals per year. Following the network literature, I'm assuming that "deal flow" or guanxi or what have you is not normally distributed - indeed, I'd opt for a
Zipf curve distribution. Ie, you can expect that a really good sourcer sees an order of magnitude more deals (and therefore more good ones) than a number of lesser ones. Think Levin Zhu versus your average investment banking Director in his/her early 30s.
First things first: why do a private deal? The only reason the analyst/pm would have any reason to do a private deal vs buying an equivalent public equity is if the returns on private transactions were significantly higher or lower risk. As can probably be gleaned from the rest of this blog, the agency problems of private deals do not make them lower risk in most asian countries than buying vanilla listed equity. Our assumptions of normality may be overly generous here but we will keep them. Now, what drives the private of capital between private and public markets? Some level of market efficiency no doubt, represented by institutional or regulatory inefficiency, or the inefficiency of the banking system as well as a plain old liquidity premium. Needless to say, in less developed countries where it is harder to list or where the banking system requires hard collateral for loans you'd expect this gap to be bigger. So, we get a few stylized facts here:
1) Private market returns should be higher in less developed markets.
2) Private capital returns relative to public market returns should be eroded away as financial and regulatory liberalization takes place.
A current example might be the ChiNext board in Shenzhen: it should, ceteris paribus, reduce the expected excess return for doing private transactions in China. This in turn would reduce the optimal portfolio mix between private and public transactions within a fund.
Now, in terms of choosing to have the Sourcer in house or outside as someone who gets paid a larger deal fee but no base salary one has to consider the costs of the Sourcer. These are both compensation costs but also time costs for looking at their deals - all of which is a cost to the analyst/PM in terms of time. Sourcers are generally less pushy when they've got something done recently so these costs are some variable decreasing in recent deal activity. It is my experience that no matter how stupid a deal looks and sounds you will not hear the end of it until there's a 5 - 10 page summary on why it's awful. So, one key saving is not just base pay but that when the Sourcer is outside you can turn down the deal and he can go and waste someone else's time. Some firms I know have outside Sourcer's who have to show everything to a particular firm first but after that can show it to whoever they like. This is a great way of making these savings.
Now lets compare the case of the big, heavy hitter who brings in a number of deals a week versus the Junior Varsity ex bank-VP/Director hustler. The big hitter will bring in so much stuff that unless the PM is incredibly bearish on the market there are going to be some deals getting done and won't be a high cost to analysts - they will, however, command a likely bigger share of returns and may be a partner in the firm. In contrast, having 4-5 guys who do the same work will be hard because they'll be much more burdensome and not necessarily bring in any more than the important guy. In a market where the divergence between private and public prices (China 2002-04?) it isn't much of a problem: deals are cheap compared to public deals and everyone is happy, regardless of whether you've got a big hitter or a bunch of hustlers. However, at a time when things are not particularly cheap and the private/public spread has collapsed you have a bunch of very upset Sourcers who can't get stuff done, and, needless to say, the big guy is less of a burden than the small guys and can bring himself to think big picture ("I am pissed about not getting deals done but my share of the GP is worthless if my what my nerdy partners are telling me is going to happen to China/US Real estate / etc").
The key point is this: Sourcers do have serious value but it depends on the private/public return difference. I would argue that due to this Zipf distribution it pays to have a couple of very senior or serious Sourcers rather than an army of time-sucking hustlers trying to find proverbial needles in haystacks.
As it pertains to Asia today, with the growth in credit and greater accessibility to markets one seriously has to wonder whether the old special situations model of having 4-5 smart guys and a small army of hustlers might no longer be a great model because the game is now quite different. I think firms are increasingly going to have to face the facts that some Sourcers are going to need to get paid, and paid well and the rest are probably better off sitting outside the firm.
