Today in Data Mining/Penis Jokes


It's not easy to generate alpha as an analyst at a quant fund these days. Most of the signals that are likely to make money have been discovered, so you find yourself sitting around running a lot of back-testing regressions on anything that could, however implausibly, predict financial results. It's a stressful, lonely job, and if you're low on good ideas your mind may wander to whatever is closest to hand.

That appears to be how Tatu Westling of the University of Helsinki came up with this:

The question is whether and how strongly the average sizes of male organ are associated with GDPs between 1960 and 1985? It is argued here that the average size - the erect length, to be precise - of male organ in population has a strong predictive power of economic development during the period. The exact causality can only be speculated at this point but the correlations are robust.

What could possibly go wrong? We took a look.

1. Data quality

Not everyone lives by Bridgewater's culture of ruthless honesty. Could the data be skewed upwards a bit? Nah, it has to be right - we found it on the internet:

The data regarding the physical dimensions of male organs is openly available online and has been compiled [by an unknown party] from an extensive number of sources. Large part of the data has been collected by health authorities but some observations are self-reported. Due to the sensitive nature of the subject matter, self-reported data might be biased, supposedly upwards. However, a moment of reflection with the global penile length distribution map and anecdotal 'Internet-sourced evidence' reveals that the self-reported figures are in-line with anticipated patterns. Still, measurement errors can not be ruled out.

2. Correlation vs. causation

So you found that penis length explained 15% of cross-country GDP variation in 1985. But you're looking to make investing decisions and you don't just want a past correlation - you want to know if it'll hold up in the future. What's the mechanism here? Behavioral finance provides the explanation:

[I]n an evidently Freudian line of thought the notion of self-esteem might be at play. In particular, male organ size s and income y could be considered factors in the aggregate 'self-esteem production function' f and hence affect utility u. Assuming the following functional form and decreasing returns of self-esteem, namely u = f(y + s) and f'(∙) > 0, f''(∙) < 0, the 'small male organ' countries would gain more utility by expanding their economy than the 'large male organ' countries. Actually the latter populations would simply exploit their nature-given, non-disposable groin-area endowments.

3. Wait, seriously?

Even with the reservations outlined above the 'male organ hypothesis' is worth pursuing in future research. It clearly seems that the 'private sector' deserves more credit for economic development than is typically acknowledged.

So ... no, then?

Sadly, Westling makes no policy recommendations. But we see a clear investing opportunity, especially if these results can be extended from countries to companies, and we look forward to analysts prying into which CEOs are overcompensating by improving margins and growing market share, and which are just exploiting their groin-area endowments.

Male Organ and Economic Growth: Does Size Matter? [Helsinki Center of Economic Research viaInfectious Greed]