Ask HN: What is your example of 'theory-induced blindness'?

https://news.ycombinator.com/item?id=19799469

by bumby • 7 years ago

5 5 7 years ago

news.ycombinator.com

What examples have you come across where a person or group was unable to accept flaws in a theory because they have already built the theory into their mental model of the world?

For example, some economists do not accept evidence of the 'low-volatility anomaly' because of their belief in the efficient market hypothesis.

PaulHoule2• 7 years ago

I want to say economics in general. Some specific cases:

* the idea that small increases in the minimum wage will lead to unemployment

* lack of ability to reckon with "secular stagnation"

* Rise of China despite China being the only place that tells neoliberals to screw off (e.g. try having a protest for neoliberalism and they'll harvest your organs)

tlb• 7 years ago

Every theory in every field has this, from theology to physics.

At the theology end of the spectrum, theories can be incredibly flawed (in the sense of being terrible predictors of experimental outcomes) and still attract new believers.

At the physics end of the spectrum, while the best theories are slightly flawed, they work well enough that rational people choose to build a worldview around them rather than throw up their hands in despair.

The low-volatility anomaly is more like the physics end of the spectrum. Empirical data doesn't quite match the theory, but nobody has a definitively better theory. Economics is the study of rational actors, but some actors are irrational, so results will never match the real world exactly.

(The low-volatility anomaly is partly a principal-agent problem: the principal has a rational risk/return tradeoff, but brokers rack up more fees with high-risk hedging strategies.)

Like when I asked my first-grader what's 4+5 and she said "6?", that doesn't shake my belief in the theory of arithmetic.

bumby2• 7 years ago

I think that's why it's appropriately named an 'anomaly'. It doesn't completely discount the theory, it only provides some evidence that the theory isn't perfect.

Or, in George Box's words, "All models are wrong; some models are useful"

PaulHoule2• 7 years ago

I wonder if the low-volatility anomaly is connected with this one:

In theory people should get paid for accepting risk in their personal careers. Gig economy jobs should pay more for the benefits you don't get. Government employees used to get paid less for more security. By that measure adjuncts should get paid more than tenured faculty members.

Practically it is the other way around. One take could be that pay and security both correlate to power. Another one is that security is good for your health, career development, networking, etc. Another one is that the idea that "risk and reward are correlated" aren't true, and that risk is really like what most people (not hedgies) think risk is... A bad thing.

bumby2• 7 years ago

My understanding of this particular anomaly is:

1. The majority of market money is handled by institutional funds, which also have leverage constraints

2. Institutional funds are generally measured by their ability to beat a market benchmark, like the S&P500, on a short term (quarterly or annual) basis

3. To beat the market, fund managers must invest in stocks that diverge from the market. This means high beta stocks - aka stocks with higher systemic risk (by definition, if beta=1.0 it will match - but not outperform - the market)

4. Most fund managers display an overconfidence bias. In other words, they think they can pick the winners of the high-beta stocks. While they may be able to do so in a particular year, rarely do the same funds pick winners consistently over a long stretch of time.

So the low-volatility anomaly is the convergence of incentives (you are measured by your ability to beat the market) and the overconfidence bias of behavioral economics (you think you can pick the winners of the high-beta stocks). As large amounts of institutional funds are funneled to high-beta stocks, it leaves low-beta stocks relatively undervalued

Ask HN: What is your example of 'theory-induced blindness'?

Related Stories