Sunk Costs Don’t Matter in Venture, Until They Do

Posted in General Venture Capital

The sunk cost bias in venture capital is widely derided as a driver of poor returns in the asset class.  It’s the psychological root cause of why good money gets thrown after bad: VCs hoping for an unlikely outcome because they are both financially and emotionally embedded deep in the deal.  Unable to extract themselves because they are in so deep, they just keep funding failing businesses.  Its certainly true that a good deal of value destruction occurs because of the sunk cost bias.

We all know the Finance 101 axiom about avoiding it: we shouldn’t consider any of the prior capital invested when making a follow-on investment decision. Instead, we should channel scarce reserve capital to the potential winners in the portfolio and we should starve or walk away from the losers.

But sunk costs in venture are actually more complex than that.  And sometimes the presence and appreciation of a sunk cost creates real investment opportunities.

Fundamentally, a follow-on investment decision should not be about a deal’s absolute return multiple, but instead about the marginal or incremental return of that new capital.  Walking away from a struggling private investment almost certainly implies a total write-down of value, due to ugly things like pay-to-plays, recaps, liquidation preferences, etc…  With a major loss of capital as the alternative, investing to protect one’s prior (sunk) capital can sometimes lead to high marginal rates of return.

Take this realistic example: DrugCo has raised $20M to date from VCs, but its lead program has struggled with some major issues and is delayed.  It’s anticipated that $5M in additional funds can solve the problem, and this could enable a defensive sale of $30M (~1x) to a Pharma buyer.

The choices for the VC are: (a) walk now and write down $20M to zero; (b) invest and make ~1x on the aggregate investment.  Not a very inspiring aggregate multiple and it certainly isn’t going to impress the LPs in the next fundraising.

But in light of the realistic alternative outcome in (a), the actual marginal return on that $5M is ~5x (returning $25M).  It’s worth noting that a 5x investment has been a top 10th-percentile deal in the venture business over the past decade, irrespective of sector.

Importantly though, without the sunk cost consideration (and the associated liquidation preferences), the return on that incremental $5M would only be ~1x (i.e., getting your money back).  So in order for the marginal return to be attractive, you have to consider the presence and financial terms around those sunk investment costs.  Hence why it’s a complex issue.

Integrating this perspective at the portfolio level is also challenging, especially when considering reserves allocations between “winners” and “losers”.

Taking the example above one step farther, lets imagine DeviceCo is in the same portfolio.  Its really been hitting on all cylinders since its founding, and it just got a term sheet for a great step-up into its new round. Lets say the higher price of the new round suggests that the likely multiple from this incremental money is 2x.  Where should a fund put its incremental $5M of reserves? Into DeviceCo to make 2x on the new round, or into DrugCo to get back to ~1x but make 5x marginal return?

Psychologically most firms would bias strongly to the former (DeviceCo).  But most LPs interested in overall portfolio returns would appreciate the latter, if the defensive sale worked (and that’s always a big IF).  But since we’re in a world of unknowns, betting on a defensive sale is tough without great diligence and conviction, whereas betting more on a known winner feels safe (e.g., you can’t get fired by buying IBM).

Deciding when to walk from a worn-out, existing portfolio company vs doubling down is a tough choice.  Chasing them all for miraculous turnarounds or “successful” defensive outcomes with the hope of great marginal returns isn’t smart nor has it led to good portfolio outcomes historically.   But letting deal fatigue block the clarity of thinking around real marginal returns is also not smart.

This nuance around sunk costs and fully appreciating the marginal return concept can be especially relevant in a closed-end venture fund where total invested capital is the important denominator; recovery of large amounts of invested capital through smart defensive investments can eliminate some of the financial craters than can form from large blow-ups in a fund.

So how then does a firm prevent this being a slippery slope to throwing the good after the bad?  Its hard to do, but it certainly has to involve the healthy process of  “champion & challenge” amongst the GPs in a partnership; that’s what LPs pay GPs to do.  Active co-sponsorship and “re-diligence” of deals helps to bring multiple informed perspectives to the table.  This C&C culture demands a set of aligned GPs around a single table, willing to walk from their own deals to create the best opportunity for portfolio returns.  It’s the essence of a tight partnership committed to their LPs.

In summary, theory has it that sunk costs shouldn’t matter.  But they do in some circumstances – and the skill of knowing when to think about them is paramount.

This entry was posted in General Venture Capital. Bookmark the permalink.
  • http://www.ginkgobioworks.com Jason Kelly

    For the poker players – can also just think of it as Pot Odds:
    http://en.wikipedia.org/wiki/Pot_odds

    a lot easier to decide when you’re throwing money at a long shot vs. playing the right odds with a known set of probabilities in cards, but idea is basically the same.