All models are wrong, but some are useful

The post below, where Peter Walker looked at my simulation data on the influence of diversification versus concentration, has ruffled enough feathers to require more clarity.

TL;DR for the lazy: The implication is simply that a trend towards larger portfolios is likely to lift the median to top quartile performance of venture capital.

This is a general observation, not a recipe for success. Nobody is trying to make the case that diversification is always and strictly better, or vice-versa.

Before going further, this quote (and the others below) from Daniel Kahneman's article, "Don't Blink! Hazards of Confidence", summarises some of the push-back:

"What we told the directors of the firm was that, at least when it came to building portfolios, the firm was rewarding luck as if it were skill. [...] The illusion of skill is not only an individual aberration; it is deeply ingrained in the culture of the industry. Facts that challenge such basic assumptions — and thereby threaten people’s livelihood and self-esteem — are simply not absorbed. The mind does not digest them. This is particularly true of statistical studies of performance, which provide general facts that people will ignore if they conflict with their personal experience."

Unequivocally true: Portfolio construction is a poorly understood topic, and too many GPs (and LPs) blindly follow the common wisdom about concentration without giving it proper thought. There is so much depth to this topic beyond the usual consensus behavior.

There are factors the model cannot represent as a simple reflection of probabilities, and many of them are counter-intuitive, or require deep analysis to uncover.

For example:

• Overconfidence is the most well-studied bias in investing, and is endemic to venture capital. It is a large part of what leads so many investors to believe they can pick well and afford the risk of overconcentration. Indeed, the GP:LP interface favours investors that are supremely confident, rather than supremely competent.

• Overconcentration manifests as broadening the distribution of returns with a negative skew, as it hurts the underperformers more than it helps the overperformers. That's exactly what we see in VC: the widest distribution of returns (of any asset class/strategy) in PE and the worst mean and median performance (in PE).

• Diversification is shown to benefit fund performance by enabling investors to take more idiosyncratic risk per investment, which is fundamentally the source of outperformance. Concentration, on the other hand, drives consensus. This dynamic was well articulated by the great David Swenson, via Howard Marks on 'Behind the Memo':

"Perhaps the most powerful lesson from Marks is the idea of 'uncomfortably idiosyncratic' investing. Citing the late David Swensen of Yale, Marks emphasizes that successful investment management requires taking positions that feel uneasy because they go against the grain."

A lot of the group-think around 'concentration versus diversification' is based on the assumption that concentrated firms must be better at picking because they spend more time per investment.

This is fundamentally untrue, as is described in Kahneman's article. When faced with difficult judgements on limited information, people will habitually change the question to something simpler.

That is, if investors spend more time on decisions where that extra time doesn't actually yield access to more useful information, they're likely to construct a process that produces greater confidence rather than greater accuracy.

"The confidence we experience as we make a judgment is not a reasoned evaluation of the probability that it is right. Confidence is a feeling, one determined mostly by the coherence of the story and by the ease with which it comes to mind, even when the evidence for the story is sparse and unreliable. The bias toward coherence favors overconfidence. An individual who expresses high confidence probably has a good story, which may or may not be true."

In venture capital this typically manifests as overconfidence in founder-focused human capital, which produces predictably bad investments (and missed good ones).

The distribution of outcomes in venture capital, and the unpredictability of those outcomes, means it is fundamentally a portfolio discipline more than it is a picking discipline. The only point at which that changes is when an investor has enough genuine experience to consistently beat the odds — which is very rare.

Just as every pitch to LPs somehow demonstrates top quartile performance, every investor believes they can beat the odds. ~90% of VCs lose that fight.

Finally, there is another important element to this portfolio construction question: reserves.

I don't have anything written on this topic (yet), but I know two people who have:

First is Joe Milam's brilliant paper on 'Process Alpha' in venture capital.

Essentially, his proposition is that a diversified portfolio with a robust follow-on strategy can reap the benefits of both strategies: broad exposure to more potential outliers, and the ability to concentrate. It is represented in the diagram below.

"Poker players and statisticians sorted this out years ago, and the economics of gambling and venture capital are similar: If you are wrong (particularly at the seed stage of investing), you lose all your money. A recognized optimization framework that is instructive as to how much investor capital to deploy at each stage of the startup life cycle (or funding round) is called the Kelly Criterion, also referred to as 'The Scientific Gambling Method'"

Another great read on this topic is from Laura Thompson of Sapphire Partners. 'Dirty Secret: Venture Reserves are Not Always a Good Thing' covers multiple angles on reserve strategies, with some interesting math, but there's one particular quote that is relevant to this theme:

"Reserve dollars have to be concentrated into winners to improve fund performance and there usually aren’t many of them. Early-stage venture remains a power-law business, and many portfolios come down to one or two winners."

To conclude a post that is already too long:

"Diversification versus concentration" is a complex and multivariate question. There is no "right" answer, and certainly not one that can be reached by a simple model of distributed outcomes.

However, the general understanding of this topic and the underlying factors is shockingly poor, which speaks to the institutional insecurity of venture capital and the unwillingness to engage in self-scrutiny.

• If you are a GP, you owe it to your LPs to be a student of venture capital. Read the research, familiarise yourself with the data. Build a well-informed and rational strategy. Do not engage in group-think.
• If you are an LP, you are paying GPs significant management fees to invest your money. Learn how to find competent investors, and then don't get in their way if they have a rational strategy.
• Beware anyone who simply dismisses questions like these; they are charlatans.

To finish with another great quote from Kahneman:

"In general, however, you should not take assertive and confident people at their own evaluation unless you have independent reason to believe that they know what they are talking about. Unfortunately, this advice is difficult to follow: overconfident professionals sincerely believe they have expertise, act as experts and look like experts. You will have to struggle to remind yourself that they may be in the grip of an illusion."