One of the hardest things to manage in early stage venture capital is asset allocation to specific investments across the portfolio. You strive to maintain diversification by investing across 20+ investments and hope that you maintain adequate reserves to support your companies, can get enough capital into your winners and avoid plowing prodigious amounts into your losers. However, things never seem to work out as planned and you end up with a wide divergence.
Let’s look at a model portfolio. Let’s say you raise a $115m fund. Off the top, you take out $15m in fees over the life of the fund. This leaves you with $100m. Your average fund will try to invest in 20 to 25 deals to get adequate diversification, so they would target $4-5m per deal. In a perfect world, you accurately assess the capital needs and take downs across each investment and you hit your target. (actually, perfect world would be you identify losers early and under-allocate). You invest 30-40% of your total allocation to a deal in the first round and hold back the remaining 60-70% for follow-on rounds.
You can always tell new investors to the business since they have not seen their share of cram down rounds and carnage. So, they put in 50-70% of their allocation up front since the valuation is low. Why not grab a bigger share of the business and then dilute down over time? They find out the hard way. At some point, they run out of dough (usually the second round). If the company has failed to hit self-sufficiency at that point, they are now relying on the kindness of strangers for capital. In this situation, it is only a matter of time before existing co-investors put pay to play provisions in, remove anti-dilution rights and such for any existing investors that don’t do their share (their "pro rata") of future rounds. They don’t want to have free riders tagging along. If you don’t have reserves ("dry powder"), all of your investment potentially gets pushed to common and all of the remaining preference comes out before you. Life is pretty ugly, especially if a couple of preferred rounds then come in on top.
There are two scenarios that are sub-optimal:
1) Big winner, not enough capital in: your company does really well, really quickly and you only get a fraction of your allocation in. So, you targeted $5m per deal, invested your 30% ($1.5m) and then the company never raised additional capital. If it does really well (say 10x), then you pull down $15m. This, unfortunately, will only pay for 3 complete losers which you may have had ample opportunity (unfortunately) to get your full $5m into.
2) Drawn out loser, too much capital in: so, you learn your lesson, you are determined not to have too little into your deal, you put 50% in ($2.5m). Things don’t go as planned and next thing you know you have $5m in the company and are staring at a cram down. You have to put another $1m to prevent a pay-to-play (support your pro rata or go to common). This deal eventually drifts sideways and maybe you get 25% back ($1.5m back, losing $4.5m). Damn, you just can’t seem to win.
In the end, you are much better off risking #1 than #2. At least you are complaining that you only made $15-20m on the deal. In the other case, you are potentially looking at getting squeezed down to nothing if you can’t defend yourself. So, as a rule of thumb, don’t get greedy. Make certain you have dry powder (or if you are an entrepreneur, make certain your existing syndicate has dry powder to see you through hard times). If you fail to reserve, existing investors will hit you with a pay-to-play and new investors will crush you on price since you don’t have an alternative. This means putting no more than 25-30% of your total allocation up front, and make certain you have reserves to meet pro rata needs. This also means that as you start to run low, you better get the firms burn down so that your pro rata will be of a smaller number (need to raise less money).
In the end, you can see how allocating capital optimally is more art than science. You set general parameters, but because of the uncertainties regarding timing, ramp, burn, etc, it is very hard to know what the ideal commitment is to each round. Given a choice, I would always lean more towards fear than greed.