NRL-bias
NRL-bias
Introduction
Everyone knows that 1+1=2. Except maybe programmers. They realize 1+1=10 can be true as well, if you’re talking binary. But imagine explaining that 1+1=10 to your 12 year old daughter that does not know programming. I know, hard to imagine such a 12 year old these days. Still, it would take you a while to convince her, being her father. But what if you were her 14 year old brother? That she does not trust in the least, because of all the pranks you’ve pulled on her over the years. Would you THEN be able to convince her that 1+1=10? No matter how convincing you are, you are not particularly likely to succeed! And this is the best metaphor I can think of for the effect of the N, R, and L variables on the way in which research is received currently by the scientific community. We all think we can accept new ideas, but it is not easy at all, because of the NRL-bias.
Pretty much the only thing that can overcome the NRL-bias in this case is high perceived authority. As in being perceived as a trustworthy father, as opposed to an untrustworthy slightly older brother. But who has high perceived Authority in any subject field? Oh yes, the people that are suffering the most from the NRL-bias, the ones that have thought about, taught, and used these theories for years and years. And how often does a seasoned scientist invent anything radically new? So you are not even fighting against the NRL-bias, you are in reality fighting against the ANRL-bias – all the variables, except for real data are working against you. And real data rarely if ever changes!
And this is the reason why. The advancement of science is not really about experimental results. It is about which theories best explain the results. This is what drives science forward. And based on these new, better alternative theories, you come up with new experiments, that would have never occurred to you before. That is how science advances. And nowadays, experiments cost millions, if not billions of dollars. And are based on existing theories, because that is how funding works. And this is why real data rarely if ever changes!
Of course in this situation, the more radically a theory differs from what is currently perceived as the truth, the more difficult it is going to be for it to be accepted. Even though this theory may be 100 times better. This should be obvious to everyone. I mean, just consider how many people still do not fully grasp the concept that Bell’s inequality does not describe reality, as it actually exists.
And this is how I constructed my experiment, which consisted of submitting papers to various publications. Now, the papers were specifically written for the subject area of economics. This subject area was selected for the simple reason that it has very little real data, especially the macro-economics discipline; a perfect illustration of theory-induced blindness in action, when real data hasn’t changed much in years.
Why do you think banks on Wall Street never hire professional financial economists? They only hire math and physics majors that hardly know anything about finance. Why? Because of the ANRL-bias; it’s one thing if you are playing around in academic fantasy-land, something else entirely when it is YOUR P&L! And of course the physics and math guys make money. Think Renaissance Capital. Unless they hire a bunch of finance professors, as in Long Term Capital Management. And we know how that turned out.
However, nobody really even bothers to look at macroeconomics. Everyone knows you can’t predict the stock market, or where the economy is going to go, or interest rates, or inflation, or anything. So nobody even tries!
Of course in such a situation, any idiot aware of the potential impact of theory-induced blindness can easily find a ton of theories that are far superior to the ANRL-bias nonsense the professional economists all agree on. But then one would have to show evidence that far superior theories are systematically being rejected by the economists. However, in order for this evidence to be overwhelming, two parameters must be met.
First of all, the rejected theories must be clearly superior mathematically, in such a way that their superiority is utterly clear to anyone not familiar with economics, which is to say, people not suffering from NRL-bias in the subject field of economics. We will call these the non-economists. And secondly, and far more importantly, the theories rejected by the NRL-biased economists must be easy for the non-economists to comprehend. So that non-economists would be able to evaluate these theories objectively, in an unbiased way.
In other words, a non-economist would actually need to understand the question that is being answered by the theory. So, for example, I wrote a paper about macro-economics that is comprehensible to a statistician who is not a top expert in macroeconomics. Such a person should be able to:
easily understand what the paper is about
comprehend the question being asked
evaluate different answers correctly, which is to say, statistically
And that’s what these two papers about portfolio optimization and inflation are. Exactly what I described above. Show this to any non-economist, and then explain to the non-economist what the economists are currently doing. They will be shocked, and so will you, by the amount of theory-induced blindness in economics.
Evidence – Portfolio optimization
Along these lines, our first exhibit above is a paper that contains incontrovertible direct proof of the fact that there is a huge amount of documented NRL-bias in finance, related to portfolio optimization, long-short portfolio optimization in particular.
Yes, the very thing that Harry Markowitz won the Nobel prize for in 1990. See long-short.pdf. Show this to any mathematician that does not know finance, and they will tell you the paper is obviously true. Show this to a financial economist, and they will literally not even recognize it as having anything to do with finance! I have had this paper rejected without even being looked at by numerous publications, whose names will remain anonymous. In the meanwhile, look at this paper from 2021. https://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=9606934 Why do people keep making the same mistake? Purely as a result of blindness induced by the original Markowitz 1952 paper. In other words, the NRL-bias prevents people that have learned the Markowitz formulation first, from noticing that it is inapplicable to optimizing long-short portfolios. While my long-short write up was utterly trivial mathematically, I have submitted it to “Computational and Mathematical Methods” for the sole reason that the approach described in it works amazingly well in practice. Of course they rejected it, and understandably so, as “the paper is not of sufficient level of quality and innovation to be considered for publication in the Computational and Mathematical Methods journal” – a direct quote. Not easy fighting the NRL-bias, is it?
Evidence – CAPM
Here is a short paper, less than 3 pages long, that proves CAPM is nonsense.
Now, this paper, I have informally sent to a journal editor that I have previously established a rapport with. Which is to say, somewhat bumped up my perceived A in his eyes. Let us see if that helps.
Evidence – Inflation
Inflation is the easiest and most trivial idea in all of human history. Literally, you double the money supply, and prices will double, that is all there is to it. Yet mainstream macro-economists from the most prestigious universities in the world can not agree on how it works. Another example of NRL-bias. They think about inflation in terms of preexisting theories that to them are absolutely true. Trying to explain how inflation really works to these people is like trying to convince a Catholic priest that there is no God. I have had this paper rejected by numerous journals as well. But I think I may be getting through to these people. I have attempted to use a mathematical crutch in this case, which is to say the maximum likelihood approach, to explain how inflation should be modeled. Let us see if this works.
This was last submitted for publication to “American Economic Journal: Macroeconomics”, maybe someone over there will fight through the ANRL-bias, but I doubt it. Show this paper to any statistician. And then explain to him what the mainstream macro-economists are doing. He will be shocked too!
More Evidence
To begin with, the N, R, and L variables explain the Adam Smith diamond-water paradox, properly. Of course even economists have figured out by now that value is subjective, and they have even invented a few synonyms for the word subjective, such as marginal utility. Naturally, without being able to fully explain where subjective value comes from, or why it changes. Of course it comes mostly from the N, R, and L variables. Indeed, this is how advertising, and propaganda in general works. What do you think Facebook is?
And why do you think cell phones are so addictive?
Estimating what we do not know
This is just a thought experiment, which is to say, a white paper with multiple potential bugs. The idea is to take the “Theory of the Firm” by Jensen and Meckling, and apply it to global-macro, which is to say, build a simple regression model that predicts cross-sectional real GDP for different countries. Why has no one thought to try this utterly trivial and obvious approach, that clearly has a huge amount of explanatory power? Theory-induced blindness, what other explanation is there?
Conclusion
So how do we get rid of theory-induced blindness? Well, you can of course use mathematical crutches, like the maximum likelihood approach. Of you can just reprogram your brain to work properly. Which is the topic of my next paper. I will just end this paper with an unsubstantiated claim. Theory-induced blindness is impacting all of science, including math, physics, chemistry, and biology. And not in a good way. Why do you think mathematicians can’t get their head around the meaning of the Ramanujan sum? Which is not to say that all crazy theories are right! I believe Jacques Benveniste and the water memory controversy is an example of a theory that was wrong. Yet I can not be sure without looking at real data, as I only read hearsay about what happened.